land use change, vegetation dynamics and rainfall spatio

Transcription

LAND USE CHANGE, VEGETATION DYNAMICS AND
RAINFALL SPATIO-TEMPORAL VARIABILITY OVER WEST
AFRICA
BAMBA, Adama
BSc (UAA, côte
, MSc (UAA, côte
(MET/11/7675)
A THESIS
SUBMITTED TO THE SCHOOL OF POSTGRADUATE STUDIES, THE FEDERAL
UNIVERSITY OF TECHNOLOGY, AKURE (FUTA) IN PARTNERSHIP WITH THE
WEST AFRICAN SCIENCE SERVICE CENTRE ON CLIMATE CHANGE AND
ADAPTED
LAND
USE
(WASCAL),
IN
PARTIAL FULFILLMENT OF THE
REQUIREMENTS FOR THE AWARD OF DOCTOR OF PHILOSOPHY DEGREE IN
METEOROLOGY AND CLIMATE SCIENCE.
DEPARTMENT OF METEOROLOGY AND CLIMATE SCIENCE
FEDERAL UNIVERSITY OF TECHNOLOGY AKURE,
NIGERIA
MARCH, 2015
ABSTRACT
The decadal variability of rainfall and vegetation over West Africa have been studied over the
last three decades, 1981-1990, 1991-2000 and 2001-2010 denoted as 80s, 90s and 00s
respectively. The Climate Research Unit (CRU) monthly precipitation and temperature data,
the Global Prediction Climatology Project (GPCP) monthly precipitation data, daily rainfall
data from two stations in
and Normalized Difference Vegetation Index (NDVI)
from the National Oceanic and Atmosphere Administration (NOAA), all covering the period
1981-2010 have been used in this study. The aim of the study is to ascertain how changes in
the land surface state affect the spatio-temporal distribution of rainfall over the West Africa
region. The relationship between rainfall and vegetation indices over the region was
determined
. Also the decadal comparison between rainfall and
NDVI over the region was based on the significant t-test and the Pearson
. The
impact of land use change (deforestation) on West African monsoon, particularly the rainfall
of June-July- August-September (JJAS) was simulated using the regional climate model
(RCM) RegCM version 4. The RegCM4 was coupled with the Biosphere-Atmosphere
Transfer Scheme (BATS) land surface state model and forced with ERAINT. The simulation
covered three years from 2005 to 2007. But the study focused on JJAS of 2005 and 2006, two
different years in terms of monsoon onset and sea surface temperature (SST) over the Gulf of
Guinea (GG). 2005 was characterized by early monsoon onset and cold SST while 2006 had
late monsoon onset and high SST over GG. The model performance was evaluated by
comparing the model output with GPCP and CRU observation datasets. Results show that
significant return to wetter conditions is observed between the decade 80s and decade 90s
over West Africa and it was maintained during decade 00s except over central part of Benin
and the western side of Nigeria where a decrease in annual rainfall magnitude was observed.
ii
During the same period, a re-greening of the Central Sahel and Sudano-Sahel regions was
noted. From decade 90s to 00s, this re-greening belt was observed toward the South and the
Coastal areas, mainly over the Guinea Coast, Sudano-Guinea and Western Sahel regions. A
northward movement of vegetation increase was also observed. A linear relationship was
observed between rainfall and NDVI in the savannah region. Linear relationship between
rainfall and NDVI was not observed in other regions. This may suggest that the re-growth of
vegetation in the savannah region may be linked to the availability of the rainfall. The
vegetation re-greening which was observed over the Sahel region in 90s following the
recovery of rainfall from the drought of the 80s was not sustained in the decade 00s due to a
slight reduction in rainfall. The RegCM4 simulation results indicate that, the model was able
to reproduce the early and late monsoon onsets over the Sahel in 2005 and 2006 respectively,
when no changes were made to the vegetation cover. After changes were made to the
vegetation cover, the RegCM4 simulation results show that the position of structures like the
Intertropical Discontinuity (ITD), and the zones of ascending and descending motion over the
Sahara desert were not affected much; however the changes in vegetation were observed to
have delayed the monsoon onset over Sahel. The time lag between the end of monsoon in
Guinea region and its onset in Sahel is more than one month (~45 days) in 2005 and around
one month (30 days) in 2006. However, the time lag between the end of monsoon in Guinea
region and its onset in Sahel is always less than a month in the simulation experiments
without changes in vegetation cover. The results of the study have shown the impact of
deforestation over the West African savannah zone on rainfall spatial and temporal variability
and provided maps of rainfall and vegetation index variability that can guide decision making
for policy makers to prevent further deforestation and soil degradation.
iii
RESUME
Cette thèse étudie l'influence du changement d'état de surface de la végétation sur la
variabilité spatio-temporelle des précipitations en Afrique de l'Ouest.
décen
, la variabilité
végétation sont révisées au cours des trois
dernières décennies 1981-1990, 1991-2000 et 2001-2010 notées respectivement 80s, 90s et
00s. Le jet de données utilisé est constitué des données de précipitation du Climate Research
Unit (CRU), prévision mondial de climatologie projet (GPCP) et quelques données de
station
comme observation et
ilisation de
normalisé de végétation (NDVI) du groupe Global Inventory
Monitoring and Modeling Studies (GIMMS).
impact de la déforestation sur la variabilité
des précipitations dans la région est étudié en utilisant le modèle climatique régional RegCM
version 4
International Centre for Theoretical Physics
Le principe est d'anticiper
le phénomène de la déforestation en attribuant à la zone de transition située entre la région
guinéenne et la région soudanaise (9 W-15 E et 6 N-10 N) des herbes courtes. Cette zone qui
est constituée par la savane arborée et parsemée de hautes herbes. Le modèle a été évaluée à
l'aide GPCP, CRU ensembles de données d'observation. La méthodologie
basée
sur l'utilisation du test de significativité statistique entre les différentes décennies et la
corrélation de Pearson entre les précipitations et NDVI. En ce qui concerne, l'expérience de la
déforestation, le modèle a été couplé avec le model Biosphere-Atmosphere Transfert (BATS)
modèle d'état de surface et forcé avec ERAINT. La simulation
(2005-2006-2007), mais l'accent a été particulière mise sur la période JJAS de 2006 et 2007
correspondant à la saison des pluies dans la région sahélienne. Les principaux
résultats sont indiqués comme suit: De 80 décennie à 90s de la décennie, un important retour
à des conditions plus humides est observée sur l'Afrique de l'Ouest et confirmé lors de la 00s
iv
décennie, sauf sur le Bénin centrale et tout le côté ouest du Nigeria où il ya une diminution
annuelle précipitations. Et au cours de la même période, un reverdissement des régions du
Sahel Central et soudano-sahéliens est noté. De la décennie 90 à 00, cette ceinture de
reverdissement est observée vers le les régions côtières du Sud et, principalement sur les
régions de la côte Guinéenne, soudano-guinéennes et du Sahel occidental. Le reverdissement
de la végétation est observé dans le sens Sud-Nord. Néanmoins, dans la décennie 90s les
changements positifs étaient en dessous de la latitude 10 N et pendant la décennie 00s il a
atteint la latitude 12 N plus haut dans les régions comme une frontière Mali, la Guinée et le
Sénégal. Une relation linéaire qui se manifeste par une forte corrélation a été trouvée entre les
précipitations principalement dans les régions de savane. La déforestation est un processus en
cours sur la région Afrique de l'Ouest en dépit de son effet négatif sur l'environnement et le
climat régional. Pour l'aspect simulation, le modèle a été capable de reproduire la mousson
précoce et tardive sur le Sahel respectivement en 2005 et 2006 avant le changement apporte
au couvert végétal. Après les changements de la couverture végétale, il est constaté que la
position des différentes structures telles que le FIT, les zones de convection et des zones de
subsidence ne sont pas affectés beaucoup; Cependant, les changements ont eu un impact sur
la mousson au le Sahel. En effet, il semble retarder l'apparition de la mousson au le Sahel. La
entre la fin de la mousson dans la région guinéenne et son apparition au
Sahel est plus d'un mois en 2005 et environ un mois en 2006, alors que cette période est
inferieure moins à mois avant la déforestation. Par conséquent, elle a réduit les précipitations
de JJAS à respectivement 5% et 3% en 2005 et 2006. L'analyse décennale des observations
des précipitations et de
la végétation ainsi que la simulation ont montré que le
reverdissement de la végétation dans la région diffère dans le temps
zone à
;
cependant, la déforestation induirait une baisse des précipitations dans la région.
v
ACKNOWLEDGEMENT
This Ph.D programme is fully supported by the German Ministry of education and Research
(BMBF) through the West African Science Service Centre on Climate Change and Adapted
Land Use (WASCAL). I am therefore grateful to WASCAL for granting me the financial
support for the study and research visit to the International Centre for Theoretical Physics
(ICTP) in Italy and to the Laboratoire d'Etude des Transferts en Hydrologie et Environnement
(LTHE) in France and my participation in conference to present the results of the study.
I sincerely thank the executive Director and the staff of WASCAL Head office, Accra, Ghana
and the Director, Prof. J.A. Omotosho and staff of WASCAL GRP-WACS, FUTA, Nigeria for
their strong support and encouragement throughout the period of the study. I am also grateful
to Prof. K.O. Ogunjobi, the Head, and the Staff of the Department of Meteorology and
Climate Science, FUTA, Nigeria, for their assistance and cooperation.
I am deeply thankful to my supervisor Prof. Arona Diedhiou at Institut de Recherche pour le
Développement (IRD) and LTHE and Co-Supervisor, Dr. Ahmed Balogun of FUTA, Nigeria,
as well as my advisors Prof. Abdourhamane Konaré from Université Félix Houphouet Boigny
(UFHB), Pole Scientifique et
(PSI) and Dr. Thierry Pellarin from the
Laboratoire d'Etude des Transferts en Hydrologie et Environnement (LTHE) Grenoble,
France for their assistance, guidance and immense contribution to make this thesis a reality.
I am grateful to Prof. Savané Issiaka and Prof. Kamagaté Bamory from Université Nangui
Abrogoua (UNA) former Université Abobo-Adjamé (UAA) for giving me the foundation tool
for research and their constant counsel.
Many thanks to my external examiner, Prof. T.O. Odekunle from Obafemi Awolowo
University, Ile-Ife, Osun state, who help me to improve my thesis.
vii
My sincere gratitude to my colleagues from the WASCAL GRP WACS, FUTA, Nigeria for
the good relationship we had during the training and the thesis writing.
I am also grateful to colleagues from the CNC Bingerville and from UNA in the
circumstances Ouédraogo Moussa and Ouattara Ismael with whom I shared many
experiences and also to Ismaila Diallo at ICTP for his assistance during my stay.
To my family, I am profoundly grateful to my uncles Koné Dramane and Koné Amara, my
brothers Bamba Yaya and Bamba Amadou, my sisters Bamba Salimata and Bamba Nabintou
etc. for their assistance, patience and supports.
To you all whose names are not officially mentioned in this document that we shared times,
ideas, materials etc. together during all these years for the achievement of this goal in Cote
Morocco (CRASTE), Ghana (UCC), Nigeria (FUTA), Italy (ICTP) and France
(LTHE, Grenoble) please find through these simple words my sincere acknowledgements.
A thought for my deceased father Bamba Siriki who I did not know, may your soul stay in
peace in paradise.
In God I believe, and I thank him for his omnipresence in my life
viii
DEDICATION
ix
TABLE OF CONTENTS
ABSTRACT ..............................................................................................................................ii
RESUME ................................................................................................................................. iv
CERTIFICATION .................................................................................................................. vi
ACKNOWLEDGEMENT .....................................................................................................vii
DEDICATION......................................................................................................................... ix
TABLE OF CONTENTS ........................................................................................................ x
ACRONYMS ......................................................................................................................... xiv
LIST OF FIGURES .............................................................................................................xvii
LIST OF TABLES ...............................................................................................................xxii
Chapter 1 .................................................................................................................................. 1
INTRODUCTION.................................................................................................................... 1
1.1 Overview .......................................................................................................................... 1
1.2 Statement of the Problem ................................................................................................. 4
1.3 Justification ...................................................................................................................... 7
1.4 Aim and Specific Objectives ............................................................................................ 8
Chapter 2 .................................................................................................................................. 9
LITERATURE REVIEWS ..................................................................................................... 9
2.1 Rainfall Spatio Temporal Variability over West Africa .................................................. 9
2.1.1 West African 1970s Drought Causes ........................................................................ 9
2.1.2 The Intertropical discontinuity and Intertropical convergence zone over West
Africa ............................................................................................................................... 11
2.1.3 Impact of climate change and the additional radiative forcing ............................... 16
2.2 Vegetation Dynamics and Rainfall Variability over West Africa.................................. 17
2.3 The Relationship Rainfall and Vegetation ..................................................................... 19
2.4 Vegetation Indices over West Africa ............................................................................. 20
2.5 The West African Monsoon ........................................................................................... 23
2.6 An Overview of Synoptic Scale Atmospheric Features over West Africa .................... 26
2.7 Climate Models .............................................................................................................. 28
x
2.7.1 Climate and general-circulation models ................................................................. 28
2.7.2 Regional climate models: RegCM .......................................................................... 31
Chapter 3 ................................................................................................................................ 34
RESEARCH METHOD ........................................................................................................ 34
3.1 Study Area ...................................................................................................................... 34
3.1.1 Study Area Location ............................................................................................... 34
3.1.2 The Vegetation over West Africa ............................................................................ 36
3.1.3 Regional climatology of West Africa ...................................................................... 39
3.2 Data Collection ............................................................................................................... 46
3.2.1 Rainfall from station data........................................................................................ 46
3.2.2 Climate Research Unit data .................................................................................... 46
3.2.3 Global Precipitation Climatology Project ............................................................... 47
3.2.4 Vegetation Indices Data form GIMMS ................................................................... 48
3.2.5 Forcing Parameters ................................................................................................. 49
3.3 Data Analysis ................................................................................................................. 51
3.3.1 Observation data processing ................................................................................... 51
3.3.1.1 Significance t-test of Differences ..................................................................... 51
3.3.1.2 Standardized Precipitation Index ..................................................................... 52
3.3.1.3 Correlation between rainfall and NDVI ........................................................... 53
3.3.2 Model Setting and Simulation ................................................................................ 53
3.3.2.1 Model Description and Simulation .................................................................. 53
3.3.2.2 Model Evaluation ............................................................................................. 56
3.3.2.3 Land Surface Model ......................................................................................... 56
3.3.2.4 Structure of BATS ............................................................................................ 59
3.3.2.5 Experimentation: Change in Land Surface State ............................................. 60
3.3.2.6 Case study of 2005 and 2006 ........................................................................... 65
Chapter 4 ................................................................................................................................ 67
RESULTS AND DISCUSSION ............................................................................................ 67
4.1 Mean NDVI and Rainfall over Last Three Decades ...................................................... 67
4.1.1 Vegetation and Spatial Distribution of Rainfall over West Africa .......................... 67
4.1.2
NDVI Decadal Variability over West Africa during 1981-2010 ........................ 69
4.1.2.1 Decadal Mean of NDVI ................................................................................... 69
xi
4.1.2.2 Seasonal Variability of NDVI .......................................................................... 71
4.2 Spatio-Temporal Distribution of the Rainfall over Three Last Decades ....................... 73
4.2.1 Changes compare to thirty years climatology ......................................................... 73
4.2.2 Decade to decade changes ...................................................................................... 75
4.2.3 Rainfall distribution ................................................................................................ 77
4.2.4 Upward Trend of the Rainfall over the Region ....................................................... 79
4.2.5 Changes over Seasons ............................................................................................. 81
4.3 Spatio-Temporal Distribution of the Vegetation over Three Last Decades ................... 83
4.3.1 NDVI Decadal Anomaly Variability over West Africa ........................................... 84
4.3.2 NDVI Decadal Variability over West Africa ........................................................... 86
4.3.3 Decadal Change on NDVI ...................................................................................... 88
4.3.4 NDVI seasonal variability over West Africa ........................................................... 90
4.3.5 Frequency distribution of the NDVI ....................................................................... 93
4.4 Relationship between Rainfall and NDVI over West Africa ......................................... 95
4.4.1 Intra Annual Variability of Rainfall and NDVI ....................................................... 96
.. 97
4.4.3 Rainfall Intra-seasonal Variability ........................................................................ 101
4.4.4 NDVI Intra-seasonal variability............................................................................ 103
4.4.5 Rainfall and NDVI Monthly Climatology ............................................................ 106
4.4.6 Relationship between Rainfall and NDVI over West Africa ................................ 109
4.5 Changes in Atmospheric Parameters............................................................................ 114
4.5.1 Upper, middle and lower levels tropospheric winds ............................................. 118
4.5.2 Tropical Easterly Jet .............................................................................................. 118
4.5.3 African Easterly Jet ............................................................................................... 120
4.5.4 Monsoon fluxes..................................................................................................... 122
4.5.5 Zonal Wind, Convection and Wind Velocities ...................................................... 125
4.5.6 Relative humidity and Convection........................................................................ 131
4.5.7 Change in Surface Temperature, Evapotranspiration Flux and Albedo ................ 135
4.5.8 Rainfall Seasonal Variability................................................................................. 136
Chapter 5 .............................................................................................................................. 147
CONCLUSION AND RECOMMANDATIONS ............................................................... 147
5.1 Conclusion.................................................................................................................... 147
5.2 Recommendations ........................................................................................................ 149
xii
5.3 Limitations of the Study ............................................................................................... 149
REFERENCES ..................................................................................................................... 151
xiii
ACRONYMS
AEJ:
African Easterly Jet
AEWs:
African Easterly Waves
AGRHYMET:
Centre Régional de Formation et d'Application en Agrométéorologie et
Hydrologie Opérationnelle
AHVRR:
Advanced Very High Resolution Radiometer
AMMA:
Analyse Multidisciplinaire Mousson Africain
ACM2:
Atmospheric convective Model Version2
BATS:
Biosphere-Atmosphere Transfer Scheme
BMBF:
German Ministry of education and Research
CIAT:
International Center for Tropical Agriculture
CRU:
Climate research Unit
DJF:
December-January-February
ECMWF:
European Centre for Medium-Range Weather Forecasts
ENSO:
El Nino North South Oscillation
EOP:
Enhanced Observing Period
FAO:
Food and Agriculture Organization of the United Nation
FGGE:
First GARP Global Experiment
GDP:
Gross Domestic Product
GG:
Gulf of Guinea
GIMMS:
Global Inventory Monitoring and Modeling Studies
GPCP:
Global Prediction Climatology Project
GTS:
Global Telecommunications System
ICTP:
International Centre for Theoretical Physics
IPCC:
Intergovernmental Panel on Climate Change
IRD:
Institut de Recherche pour le Développement
xiv
ITD:
Inter-Tropical Discontinuity
ITF:
Intertropical front
JJA:
June-July-August
JJAS:
June-July-August-September
LTHE:
Laboratoire d'Etude des Transferts en Hydrologie et Environnement
MAM:
March-April-May
MCS:
Meso-scale Convective System
NCAR:
National Center for Atmospheric Research
NCDC:
National Climatic Data Center
NCEP:
National Centers for Environmental Prediction
NDVI:
Normalized Difference Vegetation Index
NOAA:
National Oceanic and Atmosphere Administration
NORMER:
Normal Mercator
RCMs:
Regional Climate models
RegCM:
Community Regional Climate Model of ICTP
SON:
September-October-November
SPI:
Standardized Precipitation Index
SSR:
Sudano-Sahel Region
SST:
Sea Surface Temperature
SUBEX:
Subgrid Explicit moisture scheme
TEJ:
Tropical Easterly Jet
TRMM:
Tropical Rainfall Measuring Mission
UER:
Upper East Region
WAM:
West African Monsoon
WASCAL:
West African Science Service Centre on Climate Change and Adapted
Land Use
WMO:
World Meteorology Organisation
xv
WRF:
Weather Research and Forecasting model
xvi
LIST OF FIGURES
Figure 1. 1: Flood of September 2009 in Burkina Faso [a]; and cultivated land in wooded
6
Figure 2. 2: Schematic depicts relative latitudinal positions of the ITCZ/ITD, TEJ, AEJ-N,
AEJ-S and the WAJ during wet and dry years for 0 10 N (top) and 0 -12 S (bottom) (Refer
online version for color images) from Williams and Kniveton, (2011).
13
Figure 2. 3: Schematic of the atmospheric circulation in the West African monsoon system
during the boreal summer. Closed solid lines represent the isotachs of the African Easterly Jet
(AEJ), which lies around 600 hPa. The red arrows show the thermally direct meridional
monsoon circulation, and are typical of the time-mean winds in the peak monsoon season
(Lebel et al., 2010).
25
Figure 3. 1: Map of Africa showing the Study area over West Africa with six selected sites
over Sahel (Burkina Faso and Niger), Sudanian Savannah (Mali and Benin) and Guinea
region (Ivory Coast and Ghana) for propose of the intra annual variability studies.
35
Figure 3. 2: Wooded savannah landscape in the Lamto reserve located in central part of Cote
Africa.
38
Figure 3. 3: Temperature and rainfall monthly mean in Bawku, Upper East Region of Ghana
41
(1993 2011).
Figure 3. 4: Lamto Geophysical station
where many
climate and seismic parameters are measured; The station was created in 1962 by Maxime
Lamotte and Jean-Luc Tournier.
43
Figure 3. 5: Temperature and rainfall monthly mean at Lamto station, Guinea region in Cote
2004).
43
Figure 3. 6: Intra seasonal Temperature mean at the Lamto station averaged over 1971-2000.
44
Figure 3. 7: Temperature Anomaly at the Lamto Station plotted with regard to climatology of
1971-2000.
45
xvii
Figure 3. 8: Model simulation domain with the topography (in contour) the zone of interest in
red box.
54
Figure 3. 9: Schematic of individual physical process (Dickinson, 1993)
58
Figure 3. 10: Initial land cover obtained after RegCM4 domain simulation before changes in
land cover (left) and after making changes in land cover (right); more details about the legend
in Table 3.2.
62
Figure 3. 11: Diagram showing the differences on SST in Gulf of Guinea and monsoon onset
between 2005 and 2006 over West Africa.
66
Figure 4. 1: Map of West Africa showing the rainfall climatology (1971-2000) based on CRU
observation data (contour); NDVI climatology of yearly sum (shaded); the filled triangles
represent the site where rainfall and NDVI have been selected for the intra variability study.
68
Figure 4. 2: Mean annual Rainfall (mm yr-1) shown in [a]; [b] and [c] for respectively decades
80s 90s and 00s and mean annual NDVI shown in [d]; [e]; [f] for respectively decades 80s;
90s and 00s.
70
Figure 4. 3: NDVI decadal mean showing changes in vegetation cover with progressive
southward increase in decade 80s (a, b, c and d); decade 90s (e, f, g and h) and decade 00s (i,
j, k and l).
72
Figure 4. 4: Spatial distribution of rainfall significant changes in decade [a] 80s; [b] decade
90s and [c] decade 00s compare to 30 years average (1981-2010) over West Africa at a level
of 95%.
74
Figure 4. 5: Decadal changes in rainfall seasonal spatial distribution over West Africa at a
level of 95%. Blue lines are areas with significant changes.
76
Figure 4. 6: Time latitudinal diagrams of rainfall seasonal mean (a, b and c) in decade 80s;
decade 90s and decade 00s respectively and seasonal anomaly (d, e and f) for decade 80s;
decade 90s and decade 00s respectively.
78
Figure 4. 7: SPI at six locations in Burkina Faso and Niger (Sahel); Mali and Benin (Sudan)
and Ivory Coast and Ghana (Guinea Coast) over West Africa showing more wet condition
mainly apart from decade 1990. The climatology is based on 1981-2000 rainfall mean.
80
Figure 4. 8: Decadal changes in rainfall seasonal spatial distribution between decade 80s-90s
(a, b, c and d), decade 90s-00 (e, f, g and h) and decade 80s-00s (i, j, k and l) over West
Africa at a level of 95%; Blue lines are areas with significant changes.
82
xviii
Figure 4. 9: Annual significance t-test computed between decades showing significant
positive and negative changes in vegetation cover over West Africa in a, b and c for decades
80s; 90s and 00s respectively compare to 30 years average (1981-2010) at a level of 95%. 85
Figure 4. 10: Annual significance t-test computed between decades showing the positive and
negative changes in vegetation cover over West Africa between decade 80s-90s [a]; decade
80-00s [b] and decade 00s-90s [c] at a level of 95%.
87
Figure 4. 11: Time latitudinal diagrams of seasonal NDVI shown in a, b and c for decade 80s,
90s and 00s respectively and NDVI seasonal anomaly
own in d,
e and f for decades 80s, 90s and 00s.
89
Figure 4. 12: Positive and negative changes in vegetation cover over West Africa between
seasons in decade 80s-90s (a, b, c and d); seasons in decade 80-00s (e, f, g and h) and seasons
in decade 00s-90s (i, j, k and l) at a level of 95%.
91
Figure 4. 13: Annual frequencies and distribution of NDVI at Niger (left) and Burkina Faso
(right) sites over Sahel region.
93
Figure 4. 14: Annual frequencies and distribution of NDVI at Mali (left) and Benin (right)
sites over Sudanian region.
94
(right) sites over Guinea region.
95
Figure 4. 16: Decadal rainfall at the monthly timescale plotted for the six selected sites in the
Sahel region ([a] Niger and [b] Burkina Faso), the Sudanian region ([c] Mali and [f] Benin)
102
Figure 4. 17: NDVI decadal mean averaged over months at six different points over Sahel
region [a] Niger and [b] Burkina Faso, Sudanian regions [c] Mali and [f] Nigeria and Guinea
105
Figure 4. 18: NDVI and rainfall monthly mean averaged respectively over 1981-2012 and
1981-2006 at six different points over Sahel region [a] Niger and [b] Burkina Faso, Sudanian
region [c] Mali and [f] Benin and Guine
107
Figure 4. 19: Spatial correlation between NDVI and rainfall over [a] decade 1980s, [b]
decade 1990s and [c] decade 2000s significance areas at 95% confident level.
110
Figure 4. 20: Scatter plot showing correlation and linear equation between rainfall and NDVI
over 1981-2010 at six different points over Sahel region [a] Niger and [b] Burkina Faso,
xix
112
Figure 4. 21: Rainfall monthly mean biases computed over June-July-August-September
(JJAS) based on CRU a) and b) and GPCP c) and d).
115
Figure 4. 22: Temperature monthly mean biases computed over June-July-August-September
(JJAS) based on CRU a) 2005 and b) 2006.
116
Figure 4. 23: JJAS mean Tropical Easterly Jet at 200 hPa before changes [a] and [c]; and after
changes in vegetation cover [b] and [d] in 2005 and 2006.
119
Figure 4. 24: JJAS mean Africa Easterly Jet at 700 hPa before changes [a] and [c]; and after
121
Figure 4. 25: JJAS mean Monsoon fluxes at 850 hPa before changes [a] and [c]; and after
124
Figure 4. 26: Time latitudinal variability of JJAS zonal wind mean at 850 hPa averaged along
vegetation cover (c and d) in 2005 and 2006.
126
Figure 4. 27: Vertical cross-section of the wind vel
of 2005 showing convergence and divergence zones a) before the change and b) after the
change.
128
Figure 4. 28: Vertical cross-section of the wind velocity
of 2006 showing convergence and divergence zones a) before the change and b) after the
change.
128
Figure 4. 29: Time latitudinal variability of JJAS wind velocity mean at 925 hPa averaged
vegetation cover [c] and [d] in 2005 and 2006.
130
Figure 4. 30: Vertical cross-section of relative humidity in percentage (shaded) and the wind
2005 [a] and in JJAS of 2006 [b].
132
2005 [a] and in JJAS of 2006 [b].
134
xx
Figure 4. 32: Time latitudinal diagram of daily and monthly mean rainfall (mm day-1)
b) and d) with change in surface in JJAS 2005.
138
140
Figure 4. 34: JJAS rainfall biases between the experiment without changes in vegetation
cover and the experiment with changes in vegetation cover in 2005 [a] and 2006 [b].
141
Figure 4. 35: Rainfall averaged over sub Sudanian band before changes (CTL) and after
changes (Sens) in vegetation cover in 2005 [a] and in 2006 [b].
143
Figure 4. 36: Rainfall averaged over sub Sudanian band before changes and after changes in
vegetation cover in 2005 and 2006.
145
xxi
LIST OF TABLES
Table 3. 1 Drought categories from SPI (McKee et al., 1993) ................................................ 52
Table 3.2: Summary of the model configuration ..................................................................... 55
Table 3.3: Land cover/vegetation classes................................................................................. 62
Table 3.4: BATS vegetation/land-cover (Dickinson, 1993) ..................................................... 64
Table 4.1: Descriptive Statistics for the rainfall and NDVI time series ...... ..............................96
Table 4.2: Monthly Cross-correlation between AHVRR NDVI and Rainfall at Lamto station
(1981-2000); Correlation is significant at 0.05 level and (**) correlation is significant at 0.01
level. The indices p and n are respectively rainfall and NDVI. ............................................... 98
Table 4.3: Monthly Cross-correlation between AHVRR NDVI and Rainfall at Daloa (19812000); Correlation is significant at 0.05 level and (**) correlation is significant at 0.01 level.
The indices p and n are respectively rainfall and NDVI. ....................................................... 100
Table 4.4: Brief description of observational datasets and model used to set the simulation of
RegCM4.4 .............................................................................................................................. 117
Table 4.5: Changes in Evapotranspiration, Temperature and Albedo due to vegetation cover
change in JJAS of 2005 and 2006 over changed band........................................................... 135
xxii
Chapter 1
INTRODUCTION
1.1 Overview
In many African regions, the land surface has faced some considerable amount of pressure
over decades due to high rate of deforestation and exploitation. To compound matters, West
African countries have experienced some drought sequences between 70s and 80s. During
these periods, the Sahel has experienced the most substantial and sustained decline in rainfall
recorded anywhere in the world (Hulme and Kelly, 1997). These have induced some changes
in vegetation cover over the region due to a close link between vegetation and rainfall. Many
studies have shown the influence of precipitation on vegetation production (e.g. Herrmann et
al., 2005; Che et al., 2014), which in turn controls the spatial and temporal occurrence of
grazing and favours nomadic lifestyle (Sivakumar, 2007). Furthermore, the rate of population
increase over the world over the region is one of the highest in the world (Yuen and Kumssa,
2011). Also people live in poor conditions, so natural resources in general and particularly the
natural vegetation cover are
Deforestation has
become one of the major phenomena impacting the climate over West Africa. However,
desert conditions are induced by gradual and prolonged loss of vegetation cover over
extensive land areas in a country, and across two or more countries. With time, desertification
over West African regions has transformed extensive land areas into arid- and semi-arid
zones and has therefore created a vast land area that has no sufficient quality life-supporting
natural resource base and weather conditions. Furthermore, this permanent loss of vegetation
cover leads to reduction in soil moisture that curtails biodiversity productivity and permit
1
drought conditions to persist (Charney, 1975; Hulme and Kelly, 1997). In this regard, the
capacity of the original vegetation land cover to regenerate is severely impaired because of
near-total absence of rainfall. However, ambient temperatures are typically very high, a factor
that sustains continuous huge losses of soil moisture and the few available water bodies
through direct evaporation. Unusually high wind speeds, low and high atmospheric pressure
spots, etc. sustain these adverse climatic conditions.
Thus, poor land utilization practices, especially in subsistence farming and nomadic pastoral
economies in the majority of the African countries have accelerated the loss of natural
vegetation and exacerbated the problem of climate change.
The Guinea zone of West Africa is the area with intense economic activities which are mainly
agriculture, wood trade, and mine exploitation, particularly in
, where the
economic growth is supposed to be based on agriculture. This region is drastically affected by
the above activities. With regard to the importance of the land surface state in the water cycle,
many researches focused on the relationship between rainfall and vegetation. Since the long
drought period of the 70s and 80s, the relationship between land surface state and atmospheric
parameters has been the focal point of important debates, and subject of many studies
(Nicholson et al., 1990; Diedhiou et al., 1999; Dickinson, 2003 and Laux et al., 2007).
According to Herrmann et al. (2005), in semiarid environments, the relationship between
rainfall and vegetation has received a great deal of interest. Indeed, changes in land surface
condition are suspected to be the precursor to changes of rainfall spatial and temporal
distribution, precisely during drought years over Sahel region. During these periods, changes
in land surface condition have been clearly linked to the deforestation in the Guinea region of
West Africa (Charney, 1975). Charney (1975) linked the Sahel 70s drought to the change in
the surface state based on the relationship between rainfall and the land surface state: the
human activities have decreased the forest cover which in turn has increased the surface
2
albedo to create a subsidence over the region and decreased the precipitation which finally
have a negative impact on vegetation cover. This point of view is shared later by Sivakumar
(1992) and Herrmann et al. (2005) who linked this drought to anthropogenic land use
changes.
As a result of some poorly implemented environmental policies adverse anthropogenic
activities on land cover are still going on despite the warnings from scientists and. In contrast
to previous findings recent studies have shown some recovery of rainfall and re-greening of
vegetation over these regions (Rasmussen et al., 2001). The vegetation cover measured
through the chlorophyllous activity of the plants indicates a vegetation recovery. Meanwhile,
the regional rainfall fluctuations responded to the strong seasonal influence of the West
African Monsoon and associated deep convections processes. This response is related to the
thermal gradient between the ocean and continental zone (Eltahir and Gong, 1996). This is in
agreement with Omotosho (1990); Rodwell and Hoskins, (1996) and Mathon and Laurent
(2001).
In order to enhance the knowledge on West African climate and fill data gaps in Africa,
capacity building is required on methods and skills. Areas needing immediate attention are
remote sensing applications on land use change and climate systems. The remote sensing was
firstly based on coarse resolution, and subsequently improved to higher resolution. Also,
interest in regional climate modelling has steadily increased in the last two decades (Giorgi,
2006). As a result, a number of regional climate models (RCMs) have been developed, with a
wide base of model users. One such RCM is the RegCM system, which has evolved from the
first version developed in the late 1980s (RegCM1; Dickinson et al., 1989; Giorgi, 1990) to
later versions in the early 1990s (RegCM2; Giorgi et al., 1993), late 1990s (RegCM2.5;
Giorgi and Mearns, 1999), and 2000s (RegCM3 Sylla et al., 2010; and RegCM4 Pal et al.,
2007; Giorgi et al., 2012). The RegCM model was the first limited area model developed for
3
long-term regional climate simulation: it has been used in numerous regional model
intercomparison projects, and it has been applied by a large community for a wide range of
regional climate studies, from process studies to paleo-climate and future climate projections
(Giorgi and Mearns, 1999 and Giorgi et al., 2006).
The need to understand the West African climate system led to the establishment of important
research programmes over the region among them are the African Monsoon Multidisciplinary
Analysis (AMMA) (Janicot et al., 2008; Lebel et al., 2009 and Lebel et al., 2010) and, the
West African Science Service Center on Climate Change and Adapted Land Use (WASCAL)
(www.wascal.org; Bliefernicht et al., 2012). Our attention was drawn to the particular area in
between Sahel and the Guinea zone where in addition to wood exportation trade and
agriculture, the bush fire is also one the major cause of the loss of vegetation cover. Scientists
and decision makers have a crucial role to play for in reducing the rate of deforestation and
damage within the ecosystem by formulating policies that are guided by scientific evidence.
For the purposes of this study, attention was on parameters such as Normalized Difference
Vegetation Index (NDVI) and the gridded Climate Research Unit rainfall (CRU) data. The
statistical relationship between the recent changes in the NDVI and rainfall was evaluated.
Then, the RegCM model is used for the modeling aspect of the work which concerns
simulating rainfall and atmospheric features under two different experiments.
1.2 Statement of the Problem
According to IPCC (2013), Africa is one of the most vulnerable continents to the impact of
climate change and climate variability. This situation is aggravated by the interaction of
coupled with the low adaptive capacity of the
continent. Furthermore, changes in a variety of ecosystems are already being detected,
particularly in African ecosystems, at a faster rate than anticipated (Boko et al., 2007).
4
Specifically, the West African region is going through some extreme events were mainly
flood these last years. For instance some dramatic extreme events are observed respectively
1st September 2009 in Ouagadougou (Burkina Faso) (Fig. 1.1[a]), in Dakar (Senegal), 2010
and 2012 in Lagos and Abuja (Nigeria) and recently in June 2014 in Abidjan (
)
that led to wanton loss of lives and property. Land surface degradation is a continuous
process going on at a disturbing rate through agricultural and other activities (Fig. 1.1[b])
The consequences of these are socio-economicchallenges that inconvenient the population.
Earlier studies have not established the relationship between land use change and rainfall
spatio-temporal variability in the West African region.
This study will investigate the impact of deforestation over the West African savannah zone
on rainfall spatial and temporal variability as well as the trend of the distribution of
vegetation and rainfall between 1980 and 2010 following the long drought period of the
1970s in West Africa.
5
Figure 1. 1: Flood of September 2009 in Burkina Faso [a]; and cultivated land in wooded
savannah in Niarala village of
[b].
6
So many countries over West Africa have experienced the phenomenon with its associated
distress. The risks of flooding and drought have been projected to increase in many areas, the
frequency of heavy precipitation events (or proportion of total rainfall from heavy falls) will
be very likely to increase over most areas during the 21st century, with consequences for the
risk of rain-generated floods (IPCC, 2007; Bates et al., 2008; Quevauviller, 2011). At the
same time, the proportion of land surface in extreme drought at any one time is projected to
increase (likely), in addition to a tendency for drying in continental interiors during summer,
especially in the sub-tropics, low and mid-latitudes. Thus, the change in the surface state
deeply affects the rainfall regime through the modification of the onset and cessation of the
agricultural process, and the seasonal forecasting. Hence, this study wishes to address the
following questions:
(i) what is relationship between the rainfall and the land use over this particular region of
West Africa?
(ii) what are the spatial distribution of the recent trends of rainfall and the vegetation over
West Africa region?
(iii) what could be the model simulation of the impact of deforestation over savannah
zone of West Africa on the rainfall spatio-temporal variability over the region?
1.3 Justification
The land surface cover plays a key role in moderating the climate. However, the vegetation
cover has been depleted these last few years over the region where strong variability is going
on, characterized by the recurrence of the extreme rainfall and long drought of early 70s up to
mid 80s. The anthropogenic effects on changes of the vegetation cover are also a continuous
phenomenon. However, according to the FAO report, Africa has recorded the highest
deforestation rate (0.7%) throughout the world (FAO, 2000). The deforestation will induce
7
the land degradation which is an obstacle for food security. Furthermore, the population over
West Africa is one the most vulnerable throughout the world due to the high rate of poverty;
in fact the Gross Domestic Product (GDP) is one of the lowest through the world.
Understanding the relationship between rainfall and vegetation cover changes could allow
problems associated with deforestation on their day-to-day
life. Improving knowledge on the relationship between rainfall and vegetation could be a
sensitisation tool for vegetation and climate management.
1.4 Aim and Specific Objectives
The aim of this study is to ascertain how the changes in the land surface state can affect the
spatio-temporal distribution of the rainfall over West Africa region.
The specific objectives are to:
i.
assess the evolution of land cover change and rainfall spatio-temporal distribution over
the last three decades;
ii.
determine the relationship between rainfall and vegetation indices over West African
region;
iii.
simulate the rainfall over West Africa during the rainy season using regional climate
mode (RegCM) and;
iv.
evaluate the impact of the land use change on West African monsoon using RegCM.
8
Chapter 2
LITERATURE REVIEWS
2.1 Rainfall Spatio Temporal Variability over West Africa
2.1.1 West African 1970s Drought Causes
Land surface is an important component of the climate system. It is one of the major driving
forces of the regional climate. Therefore, Changes in surface energy budgets resulting from
land cover changes can have a profound influence
(Sivakumar, 2007).
This can be through the modification of the evapotranspiration, sensible heat flux, the flux of
moisture to the atmosphere etc. However, reports on the physical causes of the long drought
period in the Sahel region in West Africa during 1970s up to 1980s has progressed along two
parallel directions (Giannini et al., 2003; Sanni et al., 2012).
The first group was motivated by the belief that humanity was irreversibly impacting the
environment and climate. That, by way of land cover and/or land use changes associated with
the expansion of farming and livestock herding into marginal areas, ultimately linked to the
pressure of rapid population growth. This will emphasized the role of the feedback between
the atmospheric circulation and land surface processes. The fa
(Charney, 1975) some years ago linked the long drought period which occurred during 1970s
to 1980s to the deforestation over the West African region (Nicholson et al., 1998; Giannini et
al., 2003; Zeng, 2003; Dai et al., 2004; Lebel and Ali, 2009 and Lebel et al., 2010). This
phenomenon has had an impact on the rainfall spatio-temporal variability over the region.
This argument is sustained by the positive feedback effects induced by the relation between
9
rainfall and albedo. The human activities have decreased the forest cover. The absence of
vegetation increases the surface albedo therefore, the surface energy budget which creates a
subsidence over the region and decreased the precipitation. Finally to close the cycle, the
process will affect the vegetation cover (Charney, 1975; Hernandez et al., 2000) as there is no
rainfall to support growth.
The second group was revitalized by the initial successes in dynamical seasonal prediction of
the mid-1980s, and pointed to the atmospheric response to temperature changes in the global
oceans as the leading cause of African climate variability. They found a link between the
rainfall spatio-temporal variability over Sahel and the sea surface temperature (SST) (Vizy
and cook, 2001; Balas et al., 2007; Odekunle and Eludoyin, 2008 and Oueslati and Bellon,
2015). They concluded a link between the mentioned drought and change in SSTs over
Tropical Atlantic Ocean. Their findings were mainly based on case studies. However, a full
demonstration of that has not been done (Giannini et al., 2003; Hamatan et al., 2004). The
convergence of precipitation axis is often aligned close to the zone of maximum SST, but is
not anchored to it. Indeed, the maximum SST located within the equatorial counter-current is
a result of the interactions between the trade winds and horizontal and vertical motions in the
ocean surface layer (Giannini et al., 2003).
photosynthetic activities at seasonal and interannual scales (Anyamba, 2005). Nicholson et
al. (1990) found strong relationship between NDVI and rainfall in areas with rainfall amount
ranging between 200 and 1200 mm. Giannini et al. (2003) opined that the land-atmosphere
feedback acts to amplify the ocean forcing of the Sahel precipitation signal.
10
2.1.2 The Intertropical discontinuity and Intertropical convergence zone over West
Africa
The Intertropical convergence zone (ITCZ) is a zonal band of low atmospheric pressure and
thunderstorms caused by converging Trade Winds, rising air and intense thermal heating
oscillating on both sides of the equator (Williams and Kniveton, 2011); the location of the
ITCZ shifts throughout the year defining the wet and dry seasons in countries located in the
tropics hence, numerous interest for its study. Barry and Chorley (2003) defined it as the
tendency for the trade wind systems of the two hemispheres to converge in the equatorial
(low-pressure) zone and referred to it as referred as the ITCZ over Ocean and ITD over land.
It is one of the direct drivers of rainfall variability in West Africa through perturbations in its
strength and position (Odekunle, 2010). The ITD forms the ascending branch of the Hadley
cell, thus what affects the strength of the trade winds from either hemisphere will also affect
the position and strength of the ITD (Fig. 2.1). The North-South seasonal march of the ITD
can bring two rainy seasons to the South in West Africa; however the regions far from the
equator to the North will only experience one pronounced dry and wet season (Williams and
Kniveton, 2011). It should be noted that the position and intensity of the ITD does not always
imply the position of the rainbelt (East-West band of intense localised rainfall), as observed
by Nicholson (2009). However the strong moist ascent associated with the ITD means its
position and intensity will be an important factor in rainfall variability. It is therefore
important to investigate how the mechanisms could perturb the large-scale circulation
patterns in the region. Nevertheless, West Africa is a climatologically diverse region shown
to vary extensively in both its seasonality as well as its inter-annual and inter-decadal rainfall.
Furthermore, the rainfall is dictated by three rain-bearing processes (MCS, ITD and
monsoon). The onset stage of the West African summer monsoon is linked to an abrupt
latitudinal shift of the ITD from a quasi-stationary location at 5 N in May-June to a second
11
quasi-stationary location at 10
-August. This stage corresponds to major
changes in the atmospheric circulation over West Africa linked to the full development of the
summer monsoon system (Sultan and Janicot, 2003). This abrupt shift occurs mostly between
10 W and 5 E where a meridional land-sea contrast exists and it is also generally
characterized by a temporary decrease of convection over the whole of West Africa. During
the 1968-2005 period, the mean date of the monsoon onset was 24 June with a standard
deviation of 8 days. This has been focal point of numerous studies applying different methods
to determine the rainfall onset over the region (Omotosho, 1990; 1992).
12
Figure 2. 1: Schematic depicts relative latitudinal positions of the ITCZ/ITD, TEJ, AEJ-N,
AEJ-S and the WAJ during wet and dry years for 0 10 N (top) and 0 -12 S (bottom) (Refer
online version for color images) from Williams and Kniveton, (2011).
13
After rather weak convective activity in winter, the first rainy season over the Guinea Coast
(the coast line is located at 5 N) begins around mid-April with an intensification of
convective activity during the second half of April and most of May. This is followed by a
temporary weakening around the end of May and a recovery of convective activity in the first
half of June. Convection weakens again from approximately 25 June to 10 July, which
corresponds to the typical transition phase of the monsoon onset (Sultan and Janicot, 2003),
followed by the installation of the monsoon and convection over the Sahelian latitudes with
the centre of gravity of the ITD located between 10 N and 12 N during the whole summer.
This transition phase is said to be centred on 3 July, ten days after the mean onset date, and
corresponds to a cumulative probability of occurrence of 10% meaning that 10% on the
onsets occurred on and after 3 July.
Views on the exact nature of the ITD have been subject to continual revision (Barry and
Chorley, 2003). From the 1920s to the 1940s, the frontal concepts developed in mid-latitudes
were applied in the tropics, and the streamline confluence of the northeast and southeast
trades was identified as the Intertropical front (ITF). Over continental areas such as West
Africa and South Asia, where in summer hot, dry continental tropical air meets cooler, humid
equatorial air, this term has some limited applicability. Sharp temperature and moisture
gradients may occur, but the front is seldom a weather-producing mechanism of the midlatitude type. Elsewhere in low latitudes, true fronts (with a marked density contrast) are rare.
Recognition of the significance of wind field convergence in tropical weather production in
the 1940s and 1950s led to the designation of the trade wind convergence as the Intertropical
Discontinuity (ITD). This feature is apparent on a mean streamline map, but areas of
convergence grow and decay, either in situ or within disturbances moving westward, over
periods of a few days. Moreover, convergence is infrequent even as a climatic feature in the
doldrum zones. Satellite data show that over the oceans the position and intensity of the ITCZ
varies greatly from day to day. The ITCZ is predominantly an oceanic feature where it tends
14
to be located over the warmest surface waters. Hence, small differences of sea surface
temperature may cause considerable changes in the location of the ITCZ. A sea surface
temperature of at least 27.5°C seems to provide a threshold for organized convective activity;
above this temperature organized convection is essentially competitive between different
regions potentially available to form part of a continuous ITCZ. The convective rainfall belt
of the ITD has very sharply defined latitudinal limits. For example, along the West African
coast the following mean annual rainfalls are recorded: 12°N 1939 mm, 15°N 542 mm, 18°N
123 mm.
In other words, moving southwards into the ITD, precipitation increases by 440% in a
meridional distance of only 330 km (Barry and Chorley, 2003). As climatic features, the
equatorial trough and the ITD are asymmetric about the equator, lying on average to the
north. They also move seasonally away from the equator in association with the thermal
equator (zone of seasonal maximum temperature). The location of the thermal equator is
related directly to solar heating, and there is an obvious link between this and the equatorial
trough in terms of thermal lows. However, if the ITD were to coincide with the equatorial
trough then this zone of cloudiness would decrease incoming solar radiation, reducing the
surface heating needed to maintain the low-pressure trough. In fact, this does not happen.
Solar energy is available to heat the surface because the maximum surface wind convergence,
uplift and cloud cover is commonly located several degrees equatorward of the trough. In the
Atlantic, for example, the cloudiness maximum is distinct from the equatorial trough in
August. Convergence of two trade wind systems occurs over the central North Atlantic in
August and the eastern North Pacific in February. In contrast, the equatorial trough is defined
by easterlies on its poleward side and westerlies on its equatorward side over West Africa in
August and over New Guinea in February (Barry and Chorley, 2003). The dynamics of lowlatitude atmosphere ocean circulations are also involved. The convergence zone in the
15
central equatorial Pacific moves seasonally between about 4°N in March to April and 8°N in
September, giving a single pronounced rainfall maximum in March to April. This appears to
be a response to the relative strengths of the northeast and southeast trades.
2.1.3 Impact of climate change and the additional radiative forcing
Since 1990, every 5 years a group of approximately 2000 natural and physical scientists,
economists, social scientists, and technologists assemble under the auspices of the United
Nations-sponsored Intergovernmental Panel on Climate Change (IPCC). These scientists
spend 3 years reviewing all of the information on climate change and produce a voluminous
report following a public review by others in the scientific community and by governments.
Climate is defined as the 30 to 40 year average of weather measurements, such as average
seasonal and annual temperature; day and night temperatures; daily highs and lows;
precipitation averages and variability; drought frequency and intensity; wind velocity and
direction; humidity; solar intensity on the ground and its variability due to cloudiness or
pollution; and storm type, frequency, and intensity. Understanding the complex planetary
processes and their interaction requires the effort of a wide range of scientists from many
fluxuations. The current average rate at which solar energy strikes the earth is 342 watts per
square meter (W m-2). It is found that 168 W m-2reaches the earth
light: 67 W m-2 is absorbed directly by the atmosphere, 77 W m-2 is reflected by clouds, and
30 W m-2
to emit back into space. The net effect is that instead of being a frozen ball averaging -19 C,
Earth is a relatively comfortable 14 C. This difference of 33 C arises from the natural
greenhouse effect. Human additions of greenhouse gases appear to have increased the
16
temperature an additional 0.6±0.2 C during the 20th century. The transmission of visible light
from the sun and the trapping of radiant heat from the earth by gases in the atmosphere occur
in much the same way as the windows of a greenhouse or an automobile raise the temperature
by letting visible light in but trap outgoing radiant heat. The analogy is somewhat imperfect
since glass also keeps the warm inside air from mixing with the cooler outside air. The
called radiative forcing. The units
radiative forcing from human additions of carbon dioxide since the beginning of the
industrial revolution is 1.46 W m-2. Methane and nitrous oxide addition have provided
relative radiative forcings of 0.48 and 0.15 W m2, respectively. Other gases have individual
radiative forcings less than 0.1 W m-2. The total radiative forcing of all greenhouse gases
added by human activity to the atmosphere is estimated to be 2.43 W m-2. This should be
compared to the 342 W m-2that reaches the earth from the sun. Hence, the greenhouse gases
which is enough to cause the global temperature to increa
.
2.2 Vegetation Dynamics and Rainfall Variability over West Africa
The early 1970s to middle 1980s period has been considered generally as the long drought
period over the West African regions, specifically the Sahel has been the most affected among
the regions (i.e. agricultural, hydrological, meteorological droughts etc.). Many debates have
focused on the impact on rainfall spatio-temporal variability by changes in vegetation cover.
So, this section reviews the synthesis of research based on respectively rainfall recovery and
vegetation re-greening over the two last decades in West Africa.
Many studies have underlined the rainfall recovery in the 90s-00s over West African region.
For instance, Nicholson (2005) used rainfall estimates from the Tropical Rainfall Measuring
17
Mission (TRMM) to deduce a rainfall recovery mainly over western Sahel and dry conditions
over northern Sahel. Nowadays, researches are bringing out the causes of this recovery. Some
have linked it to the increase of SST and changes in atmospherics parameters. More recently,
the rainfall recovery over Sahel has been linked to the Saharan Heat Low (SHL) by Evan et
al. (2015). They found an upward trend of the SHL due to greenhouse warming which is
caused by the presence of water vapor. Thus, they conclude a consistent recovery form
drought with the warming process of the SHL.
Regarding the vegetation re-greening, desertification has been of major concern that affects
the land cover changes. Desertification is defined by the United Nations as land degradation
occurring in dry land caused by a range of factors including climate variations and human
management. Some years ago numerous researches were focusing on it over West Africa
Sahel region. The world has witnessed unprecedented changes in the pace, magnitude and
spatial extent of changes in the land surface use (Sivakumar, 2007). So that extended
droughts in certain arid lands have initiated or exacerbated desertification. In the past 25
years, the Sahel has experienced the most substantial and sustained decline in rainfall
recorded anywhere in the world observed from available records (Hulme and Kelly, 1997).
Some findings have indicated a re-greening of the vegetation over some specific areas of the
region. Thus based on the observation that the desertification and revitalization of dunes were
phenomena associated with the period between the early 70s and the mid-80s as observed by
Rasmussen et al. (2001) from satellite data Herrmann (2005) also associated a recovery from
the great Sahelian droughts to the recent increase in seasonal greenness over large areas of
West Africa. Furthermore, Fenshol and Rasmussen (2011) investigated the relationship
between the vegetation productivity and rainfall Sahel-Sudanian zone of Africa and showed
an increase in NDVI over Sahel-Sudanian zone during the period 1982-2007. Some other
authors have used sequential approach to investigate this. Thus, Anyamba and Turcker (2005)
divided the time series of rainfall and NDVI in two periods 1982-1993 and 1994-2003 by
18
defining respectively the persistence of drought on NDVI and the
conditions with region-wide above normal NDVI conditions. More recently, Dardel et al.
(2014) analysed the re-greening of Sahel after some experimentation on two sites in Mali and
Niger using both satellite and observation data. Apart Anyamba and Turcker (2005) who
made sequential study to separate the different periods in the NDVI trends, over studies have
used continuous approach to investigate the phenomenon of rainfall recovery.
2.3 The Relationship Rainfall and Vegetation
Surface water balances reflect the availability of both water and energy. In regions where
water availability is high, evapotranspiration is controlled by the properties of both the
atmospheric boundary layer and surface vegetation cover (Bates et al., 2008). Changes in the
surface water balance can feed back on the climate system by recycling water into the
boundary layer (instead of allowing it to run off or penetrate to deep soil levels). The sign and
magnitude of such effects are often highly variable, depending on the details of the local
environment. Hence, while in some cases these feedbacks may be relatively small on a global
scale, they may become extremely important at smaller space or time-scales, leading to
regional/local changes in variability or extremes. The impacts of deforestation on climate
illustrate this complexity. Zheng and Eltahir (1998) showed that the meridional distribution
of vegetation plays a significant role in the dynamics of West African monsoons. The
response of the atmosphere to any perturbation in the distribution of vegetation depends
critically on the location of this perturbation. Some studies indicate that deforestation could
lead to reduced daytime temperatures and increases in boundary layer cloud as a consequence
of rising albedo, transpiration and latent heat loss. However, these effects are dependent on
the properties of both the replacement vegetation and the underlying soil surface and in some
cases the opposite effects have been suggested. The effects of deforestation on precipitation
are likewise complex, with both negative and positive impacts being found, dependent on
19
land surface and vegetation characteristics (Bates et al., 2008). A number of studies have
suggested that, in semi-arid regions such as the Sahel, the presence of vegetation can enhance
conditions for its own growth by recycling soil water into the atmosphere, from where it can
be precipitated again. This can result in the possibility of multiple equilibriums for such
regions, either with or without precipitation and vegetation, and also suggests the possibility
of abrupt regime transitions, as may have happened in the change from mid-Holocene to
modern conditions. Soil moisture is a source of thermal inertia due to its heat capacity and the
latent heat required for evaporation. For this reason, soil moisture has been proposed as an
important control on, for example, summer temperature and precipitation. Feedbacks between
soil moisture, precipitation and temperature are particularly important in transition regions
between dry and humid areas, but the strength of the coupling between soil moisture and
precipitation varies by an order of magnitude between different climate models, and
observational constraints are not currently available to narrow this uncertainty (Bates et al.,
2008).
2.4 Vegetation Indices over West Africa
The utility of remote sensing data especially satellite images have been proven in climate
monitoring and prediction. Also historical baselines of forest cover are needed to understand
the causes and consequences of recent changes and to assess the effectiveness of land-use
policies (Kim et al., 2014). McGuffie (1994) showed that through the use of satellites it is
possible to monitor many aspects of the surface and atmosphere of the Earth. Meteorological
satellites have enhanced our understanding of the synoptic processes and now form a routine
part of weather information which is distributed to the general public. These satellites have
also provided increased understanding of many smaller-scale processes which were not
resolved by the surface synoptic network. Indeed, satellite observations provide more
spatially and timelier continuous input data coverage sources than ground gauge station
20
observations. Because of the characteristics of the spatial coverage of climate, satellite
images enable us to understand manifestation of drought in larger area in less time consuming
way than conventional method. When drought exists, due to reduction in precipitation, the
capacity to carry out the chlorophyllian by the vegetation is notably reduced (Kim et al.,
2014; Reiche et al., 2015). The response of the green vegetation is characterized by
maximum absorption radiation in the red region and large reflection in the neighbouring
infrared region. it also has been observed that in unhealthy, ageing or subject to condition of
vegetation stress, the reflectance in red region increases while in near infrared region
decreases. The normalized difference vegetation index (NDVI), developed by Trucker in
1979 is the most popular vegetation index used to monitor vegetation at regional to global
scales. It is calculated as in the following equation.
NDVI
where
NIR
and
RED
NIR
RED
NIR
RED
(2.1)
are the reflectance in the near infra red (NIR) and in the Red bands
respectively, their values vary between [-1, +1].
As NDVI is not sensitive to influences of soil background reflectance at low vegetation cover
and log vegetation response to precipitation deficient, NDVI itself does not reflect drought or
non-drought conditions. But severity of drought may be defined as NDVI anomaly from its
long-term. The anomaly NDVI of drought may be defined as NDVI at current time step, such
as month, and a long-term mean NDVI of the same time step for each pixel. When the
NDVIanomaly is negative, it indicates the below-normal vegetation condition and maybe a
drought situation. The larger the negative departure, the greater drought severity may be
suggested.
21
(2.2)
where NDVIi is the value for time step i, NDVIi,mean is the long-term mean value of same time
step i
Remote sensing has been widely applied in many studies (Browning and Roberts, 1994; Marx
et al., 2008; Henke et al., 2012; Mito et al., 2012; Stow et al., 2014 and Yuan et al., 2015).
Therefore, when compared with other methods, remote sensing has obvious superiority in
estimating a real sensible heat flux over different surface conditions. However, a major
challenge is in retrieving accurate and reliable values of sensible heat flux by this technique.
The challenge consists of two tasks: the first is to retrieve accurate estimates of surface
temperatures from satellite data and to extrapolate them temporally and the second is to relate
estimates of sensible heat flux obtained over relatively large surfaces from satellite data, 1-16
km2, to ground measurements usually representative of less than 0.1 km2.
Mangiarotti et al. (2012) have used AVHRR-NDVI data to study the predictability of
vegetation cycles over the semi-arid region of Gouma in Mali. For that study a model based
on a reconstruction approach in which the NDVI signal is taken as a proxy of the system's
dynamics was used and also as a model will be for this reason, the corresponding models will
be referred to as proxy models here. Forecasts are obtained from these models. And it will
base of on the growth of the forecast error. As a result, it indicated a rapid increase in error
with regard to the horizon of prediction and shows large interannual variability. And the
degree of forecasting error clearly decreases as the aggregation scale increases, revealing the
higher predictability of the behaviour of vegetation at the scale of large regions.
Unfortunately, most of the existing land cover products were produced based on remotely
sensed data that were acquired during a single year or non-consecutive years, which has
largely constrained their contribution to long-term or up-to-date applications. Therefore,
22
timely updating of the existing land cover products based on the large and growing body of
satellite data is still urgently required for numerous applications (Chen et al., 2015).
2.5 The West African Monsoon
The basic drive for the monsoon circulation is provided by the contrast in the thermal
properties of the land and sea surfaces (James and Gregory, 2012). Because the thin layer of
soil that responds to the seasonal changes in surface temperature has a small heat capacity
compared to the heat capacity of the upper layer of the ocean that responds on a similar time
scale, the absorption of solar radiation raises the surface temperature over land much more
rapidly than over the ocean. The warming of the land relative to the ocean leads to enhanced
cumulus convection, and hence to latent heat release, which produces warm temperatures
throughout the troposphere. The basis of all transport of air masses in the atmosphere is
movements of air from areas with high pressure to areas with lower pressure. Unequal
flux is much higher in the tropics compared to the Polar Regions. Therefore, heat is
transferred by air movement as well as ocean currents (such as the warm Gulf Stream) from
the equator to the poles. Also, the characteristics of the surface (sand is heated very quickly,
whereas ocean surfaces are not) are very important and lead to phenomena that differ very
much in scale, from a local see breeze to the distribution of high and low pressures over
the equator
prevents the straightforward movement of air from low- to high pressure areas and induces
the so-called Coriolis force, resulting in circular movements of air around high-pressure and
low-pressure areas. The friction induced by mountains and smaller objects, such as cities or
forests, is another factor that influences wind direction and speed.
The annual climatic regime over West Africa has many similarities to that over South Asia,
the surface airflow being determined by the position of the leading edge of a monsoon trough.
23
This airflow is southwesterly to the South of the trough and easterly to northeasterly to its
North. The major difference between the circulations of the two regions is due largely to the
differing geography of the land to sea distribution and to the lack of a large mountain range to
the North of West Africa. This allows the monsoon trough to migrate regularly with the
seasons. In general, the West African monsoon trough oscillates between annual extreme
locations of about 2°N and 25°N. In 1956, for example, these extreme positions were 5°N on
1 January and 23°N in August. The leading edge of the monsoon trough is complex in
structure and its position may oscillate greatly from day to day through several degrees of
latitude. The classical model of a steady northward advance of the monsoon has recently been
called into question. The rainy season onset in February at the coast does propagate
northward to 13°N in May, but then in mid-June there is a sudden synchronous onset of rains
between about 9°N and 13°N. The mechanism is not yet firmly established, but it involves a
shift of the lower tropospheric African Easterly Jet (AEJ).
24
Figure 2. 2: Schematic of the atmospheric circulation in the West African monsoon system
during the boreal summer. Closed solid lines represent the isotachs of the African Easterly Jet
(AEJ), which lies around 600 hPa. The red arrows show the thermally direct meridional
monsoon circulation, and are typical of the time-mean winds in the peak monsoon season
(Lebel et al., 2010).
25
In winter, the southwesterly monsoon airflow over the coasts of West Africa is very shallow
(1000 m) with 3000 m of overriding easterly winds, which are themselves overlain by strong
(>20 m s 1). North of the monsoon trough, the surface northeasterlies (i.e. the 2000 m deep
Harmattan flow) blow clockwise outward from the subtropical high pressure centre. They are
compensated above 5000 m by an anticlockwise westerly airflow that, at about 12,000 m and
20 to 30°N, is concentrated into a subtropical westerly jet stream of average speed of 45 m s
1
. Mean January surface temperatures decrease from about 26°C along the southern coast to
14°C in southern Algeria.
With the approach of the northern summer, the strengthening of the South Atlantic
subtropical high pressure cell, combined with the increased continental temperatures,
establishes a strong southwesterly airflow at the surface that spreads northward behind the
monsoon trough, lagging about six weeks behind the progress of the overhead sun. The
northward migration of the trough oscillates diurnally with a northward progress of up to 200
km in the afternoons following a smaller southward retreat in the mornings. The northward
spread of moist, unstable and relatively cool southwesterly airflow from the Gulf of Guinea
brings rain in differing amounts to extensive areas of West Africa. Aloft, easterly winds spiral
clockwise outward.
2.6 An Overview of Synoptic Scale Atmospheric Features over West Africa
Burpee (1972), one of the AEWs study pioneer tried to improve the knowledge about the
origin and the structure of AEWs. Since then, studies have been going on to mastery the
contour of this phenomenon. To continuous in the AEWs origin, Albignat (1980) gave more
details about it. He used the spectrum and cross-spectrum analysis to locate Waves observed
on a time period covering 23 August to 19 September 1974. He found like his predecessor
stward. He found that the
mid-tropospheric easterly jet stream plays an important role in the development of the waves
26
but a possible additional source of energy should come from the release of latent heat by
organized cumulus convection. Mekonnen et al. (2006) used the brightness temperature to
examine the association between convection and AEWs. Adedoyin (1997) has shown that the
warming up of the Indian, Pacific and South Atlantic Oceans strengthens the AEJ. A stronger
AEJ, on the other hand, leads to a southward shift of the zone of squally activities in tropical
North Africa thereby resulting in rainfall deficits north of latitude 12. The change in climate
ributable
to fact that the SSTs of the three influencing oceans have persistently warmed up in July,
August and September of the 18year period 1969-1986. Omotosho et al. (2000) used the
upper wind to predict the rainfall onset and cessation two months ahead using. He found that
agriculturally reliable rainfall commences about 70 days after the first sudden changes in
wind direction from westerly to easterly at the 400 hPa level and above. For that he used
simple empirical schemes for predicting. Some years later Diedhiou et al. (2002) used the
reanalysis data from NCEP/NCAR have studied the energetic of 3-5-day and 6-9-day African
Easterly wave regimes. He confirmed that these energies are due to baroclinic and barotropic
instability. According to the source of energy which gives them birth and maintains them, the
3-5-day waves grow between
Paeth et al. (2005) focus on the mechanisms of
interannual rainfall fluctuations at the synoptic scale during a 25 year hind cast period
extending from 1979 to 2003. They used SSTs and large-scale atmospheric circulation as
prescribed at the lateral boundaries of the regional climate model sector as key factors in
precipitation variability.
27
2.7 Climate Models
2.7.1 Climate and general-circulation models
Climate models have been put to a variety of scientific and regulatory uses. Primarily the
models are used to predict meteorological events, to simulate long term climatic parameters
and to estimate the atmospheric concentration field in the absence of climate monitored data.
In this case, the model can be a part of an alert system serving to signal when atmospheric
pattern potential is high and requiring interaction between control agencies and emitters. The
models can serve to locate areas of expected high concentration for correlation with health
effects. Real-time models can also serve as official guides in cases of nuclear or industrial
accidents or chemical spills. Here the direction of the spreading cloud and areas of critical
concentration can be calculated.
A current popular use for atmospheric diffusion models is in air quality impact analysis. The
models serve as the heart of the plan for new source reviews and the prevention of significant
deterioration of air quality. Here the models are used to calculate the amount of emission
control required to meet ambient air quality standards. The models can be employed in
preconstruction evaluation of sites for the location of new industries. Models have also been
used in monitoring network design and control technology evaluation.
The basic purpose of climate models originally was to explain features of current climate and
general circulation in terms of the geometry of the earth-sun systems and basic physical
principles. Once this was accomplished, it was possible to experiment with climate models by
changing various input parameters to account for the features of past climates and to
with explaining the temperature distribution; the effect of circulation, if included at all, is
prescribed in terms of the temperature distribution. Such climate models, range from simple
zero dimensional (averaged over the whole atmosphere) to models allowing for vertical and
28
horizontal variations of input parameters. The models differ in the way boundary conditions
and physical processes are handled. Sometimes clouds are prescribed, or they may be
generated by the models. Ocean temperatures may be given, or the atmospheric model may
be combined with an oceanic model. One of the important effects in many climate models is
the ice-albedo feedback: If the model predicts a cooling, the ice sheets expand, causing
increased albedo and more cooling-
s
so severe that only a small cooling led to an ice covered earth. In spite of such excesses,
climate models have explained most features of the vertical temperature distributions and
effects of land-water differences and topography. General-circulation models (GCMs)
attempt to duplicate the distribution of wind as well as of temperature and moisture. They are
usually three dimensional; however, the earliest models had little resolution in the vertical;
for example, the pioneering model by Norman (1956) consisted of only two layers. In the
meantime, much more complete models have been developed, many requiring the fastest
computers available.
In the models, generally seven basic equations are solved for the seven basic variables of
meteorology: pressure, density, temperature, moisture, and three velocity components. The
seven equations are the gas law, the first Law of thermodynamics, equations of continuity for
modelled explicitly, ozone concentration must be added as a variable and at least one
equation must be included to describe the ozone budget. Since the ozone budget depends on
concentrations of many other trace gases, many more variables and equations are sometimes
added. The location of the Inter-Tropical Discontinuity (ITD) represents the confluence zone
of the south-westerly monsoon winds with the north-easterly dry Harmattan winds. The
monsoon winds are controlled by the pressure gradient between the low pressures of the
Saharan heat low centred along the ITD and the oceanic high pressures of the Santa Helena
anticyclone. The Harmattan winds are controlled by the pressure gradient between the
29
Saharan heat low and the Libyan and Azores anticyclones. In June the ITD is centred over the
southern coast of West Africa, South of 10 N and corresponds to the last part of the first rainy
season over the Guinea coast region. At this stage the spatial extension of the monsoon winds
is limited and the ITD is positioned between 15 N and 20 N with its northernmost latitude
between 0 W and 5 E. In July the ITD is shifted to the north reaching a quasi-stable state
around 10 N and the area of westerly winds extend over land and over the tropical Atlantic
between 5 N and 15 N, the ITD reaching 20 N, its northernmost latitude. The westerly wind
speed increases over West Africa. In August the monsoon is fully developed consistent with
the highest pressures in the southern tropical Atlantic and the northernmost location of the
Saharan heat low over land. The ITD is still around 10 N but with increased precipitation.
The westerly wind area has its largest extension, especially over the northern tropical Atlantic
where westerly moisture advection inland reaches its seasonal maximum. In September the
westerly wind area does not change significantly but the westerly wind speed decreases
drastically. This is the last part of the fully-developed monsoon season.
The AEJ maintenance is mainly controlled by the low/mid-levels transverse circulation
induced by the heat low (Thorncroft and Blackburn, 1999). This circulation, as well as the
AEJ, has its highest intensity and spatial extension in June before the monsoon onset (Sultan
and Janicot, 2003). The core of the jet is located between 5 N-10 N and 10 W-10 E with a
mean highest speed of 14 m s 1. In July and August the AEJ moves to the north, and is
oriented along a southeast-northwest axis around 15 N over the western coast of West Africa.
Its core speed at this time decreases to 10 m s 1. It retreats southward in September and its
speed increases to 12 m s 1. Another core is noticeable south of the equator in September
independent of the AEJ and has been described previously by Grist and Nicholson (2001).
At 200 hPa the high-level anticyclone structure, which is the sign of the Indian and African
monsoons induces an easterly wind field on its southern flank with a core speed greater than
30
20 m s 1 centred over the Indian Ocean. This jet, the Tropical Easterly Jet (TEJ), enhances in
July and August in parallel with the monsoon activity. In particular its westward extension
over Africa is reactivated and a maximum of 14 m s 1 is evident between 10 W and 30 W. It
decreases significantly in September over its whole domain.
In 2006 the TEJ was weaker over the Indian sector during the whole summer compared to the
mean (the Indian monsoon was a bit more active over north-eastern India but a bit weaker
over central India and highly weaker over the equatorial Indian Ocean and Indonesia
compared to the 1979-1999 mean). Over West Africa, after a period of weaker wind speeds
in June, the TEJ extended further to the West in July and displayed higher speeds in August
and September, consistently with a bit higher active monsoon season compared to the 19791999 mean.
2.7.2 Regional climate models: RegCM
Global modeling can provide (spatially, temporally, and spectrally) complete and consistent
datasets for all aerosol properties. Concerns exist, however, as to the accuracy of the
underlying assumptions (emissions, transport, and water uptake) and parameterizations
(aerosol processing, interactions with clouds). With rather general constraints (column and
component-integrated data from remote sensing), there is significant diversity in aerosol
global modeling, especially at modeling sub steps (Textor et al., 2006). To counteract this
diversity and to establish characteristic particle properties from global modeling, aerosol
simulations of more than twenty different models were considered. All of these models
employed advanced aerosol modules, which distinguished between aerosol components of
dust, sulphate, sea salt, organic carbon, and black carbon (Kinne et al., 2006). Simulated
monthly averages were re-gridded to a common 1°×1° latitude/longitude horizontal
resolution. The local median value of monthly averages suggested by all models at any grid
31
point was picked; thereafter, these median values were combined to define monthly fields
from global modeling. These median fields have the advantage that extreme behaviour
(outliers) of individual models is suppressed. In addition, these median model fields tend to
score better when evaluated than individual models (Schulz et al., 2006).
Regional Climate Models (RCMs) (Giorgi and Mearns, 1999) are now extensively used to
downscale large scale climate information to regional scales in order to account for fine-scale
processes that regulate the spatial structure of climate variables.
In this study the RegCM 4 will be used. It has been wisely applied through the world for both
dynamical and statistical downscaling. Sylla et al. (2012) have used the ICTP regional
climate model, RegCM3, nested in NCEP and ERA-Interim reanalyses (NC-RegCM and
ERA-RegCM, respectively) to explore the effect of large-scale forcings on the model biases
over a southern Africa domain at 25 km grid spacing. It was discovered that the RegCM3
shows a generally good performance in simulating the location of the main rainfall features,
temperature and synoptic scale circulation patterns, along with cloud cover and surface
radiation fluxes; it also has some wet and dry biases.
used ICTP-RegCM3 to
examine the coastal effects over the Eastern Mediterranean region has downscaled to a 10 km
resolution over the EM with a 50 km driving nest. As result, the high-resolution simulation
captures strong temperature gradients of resolving the steep topography over the eastern
Black Sea and Mediterranean coasts of Turkey, as well as the Ionian coast of Greece.
It indicates that the seasonal temperature biases for 10 km simulation are <1°C and the
frequency of dry and wet spells is well reproduced by the model. It must be noted that though
it has been wisely applied for temperature and rainfall variability and trend, some also couple
it with others atmospheric parameters. This the case of Zanis et al. (2012) who coupled
RegCM3 with aerosols to investigate the direct shortwave effect of anthropogenic aerosols on
the regional European climate over a 12 year period (1996-2007). Aerosol feedback induced
32
small changes in the yearly averaged near-surface temperature over Europe during this period
and the greatest negative temperature difference of -0.2°C was observed over the Balkan
Peninsula.
33
Chapter 3
RESEARCH METHOD
3.1 Study Area
3.1.1 Study Area Location
The investigation was done over West African region shown in Figure 3.1. The climate of the
region is controlled largely by two dominant air masses. These are the dry, dusty, continental
air mass (which originates from the Sahara desert), and the warm, maritime air mass (which
originates from the Atlantic Ocean) (Imo and Ekpenyong, 2011). The influences of both air
masses are determined by the movement of the ITCZ over ocean and the ITD over land. The
interplay of these two air masses gives rise to two distinct seasons within the sub-region,
namely: the dry and wet season. The wet season is associated with the tropical maritime air
mass, while the dry season is a product of the tropical continental air mass. The influence and
intensity of the wet season decreases from West Africa coastal region towards the North.
Based on the rainfall seasonal distribution, two main sub-regions are observed. South of 8°N,
rainfall in the Guinea region is bimodal, the region also presents four seasons (Le Barbé et
al., 2002 and Konaté and Kampman, 2012): a long dry season occurs between December and
February, and two rainy seasons from March to July and from September to November are
separated by a little dry season in August. Over the Sahel region, rainfall pattern is unimodal
with amounts decreasing northwards with a gradient of 1 mm km-1 (Lebel et al., 2003), and is
characterised by a long dry season between November and March and a wet season between
April and September. The precipitation of the rainy season is associated with thunderstorm
activity which occurs along disturbance
Thus, about 80 % of the
34
total annual rainfall for most places is associated with line squall activities which are
prevalent between June and September (Adefolalu, 1986 and Mathon and Laurent, 2001).
Figure 3. 1: Map of Africa showing the Study area over West Africa with six selected sites
over Sahel (Burkina Faso and Niger), Sudanian Savannah (Mali and Benin) and Guinea
region (Ivory Coast and Ghana) for propose of the intra annual variability studies.
35
Dietz et al. (2004) have shown that over West African drylands, rainfall data for the period
1960-1990 reveals a decline in average rainfall indicating changes in aridity between 1930-60
and 1960-1990. Some of the regions in the northern zone with semi-arid conditions in 193060 had clearly become arid (on average) in the 1960-1990 period, with unsuitable conditions
for millet or sorghum production in most years. A considerable part of the sub-humid zone in
the period 1930-1960 had become semi-arid in 1960-1990 with considerable drought risks,
certainly for crops which are less adaptable to drought stress (maize, cotton).
The ITD, also known as the monsoon trough or the doldrums, is formed near the equator by
the meeting of the north-east and south-east trade winds. These winds force moist air
upwards, causing water vapour to condense out as the air rises and cools. The ITD follows
north and south through the year, as the earth tilts on its axis, relative to the sun. In West
Africa, the start of the monsoon depends on the northward progression of the ITD over the
period June to August, when the Sahel and the southern Sahara receive most of their rainfall.
Even small shifts in the position of the ITD rain belts result in large local changes in rainfall,
bringing severe droughts or flooding (Camilla, 2009).
3.1.2 The Vegetation over West Africa
West Africa hosts a rich variety of forests and each type of forest and woodland plays an
essential role in supporting and regulating the ecosystems on which people and plants
depend. The vegetation of West Africa presents a simple picture compared to other part of
tropical Africa (Konaté and Kampmann, 2012). Due to its low-lying terrain the zones of
vegetation largely reflect both the basic climatic zones and soil, essentially exhibiting the
same longitude as rainfall. This results in a series of vegetation zone running in roughly
parallel bands from the southern Guinea coast with high and evenly disturbed rainfall
36
throughout the year to zones of increasingly drier vegetation until the Sahara desert is reached
in the North.
For the delimitation and description of these vegetation zones various classification
approaches exist based on climatic and/or phytogeographic parameters. The most applied and
accepted vegetation classification for the whole Africa (White, 1983) is taken into account.
His delimitation of vegetation zones is principally based on patterns of species distribution
and distinguishes regional centres of endemism (with>50% of their flora being endemic) and
transition zones between them. For each vegetation zone several main vegetation types are
recognized, which are characterized by their physiognomy from the North to South the
structure of the vegetation change progressively desert in the North, savannah in the Centre
(Fig. 1.3) and forest over coastal regions. Specifically, based on BIOTA WEST programme
four vegetation types zones, are mainly identified namely Guinea-Congolian, GuineaCongolia/Sudanian zone, Sudanian zone and Sahel zone.
Over 75%
where it is filtered and purified. Forests provide the habitats for many thousands of plant and
anim
Forests influence the climate through a range of physical, chemical and biological processes
that affect the atmosphere, the water cycle and global energy balance (Bonan, 2008), and play
two very different, but equally important roles on the global climate stage. Their first role is
that of carbon storage. Forests buffer the planet against global warming by absorbing carbon
dioxide, thereby helping to stabilize the atmospheric levels of this greenhouse gas. Second,
they regulate local and global weather patterns by storing and releasing moisture.
37
Figure 3. 2: Wooded savannah landscape in the Lamto reserve located in central part of Cote
It is the type of vegetation met generally over Sudanian savannah regions of West
Africa.
38
As well as the risk to forest life posed by climate change, forests are also under attack from
humans. Around four million hectares of forest are felled or burnt in Africa each year, an area
equivalent to roughly twice the size of Rwanda. There are large regional differences in
deforestation, with Togo having one of the highest rates, not just in Africa, but in the world,
having lost 44 % of its forests since 1990. At a global level, average annual rates of
deforestation were around 8.9 million hectares per year in the 90s.
In West Africa, as in other parts of the world, forests are cut in favour of pasture, crops,
settlements and infrastructure, and for extraction of fuel and timber, much of which is
uncontrolled or under-regulated in the region.
3.1.3 Regional climatology of West Africa
The West African savannah belt with its tropical location and geomorphology, incoming
solar radiation is relatively constant as are the temperatures. The southern Sudanian savannah
zone is characterised by a night day variation of 20°C (Bagayoko, 2006 and Schindler, 2009),
but northwards in central Burkina Faso and near the Sahel, average temperatures of 25°C in
January and 32°C in April and relative humidity of 6% during the dry season and 95% in the
rainy season are common. It is found that the temperatures can oscillate strongly, from 15°C
during the night to more than 40°C during the day (Sandwidi, 2007). Recent analysis have
found a rise in the average temperature at about 1°C between 1960 and 1990 (Ouédraogo,
2004 and Sandwidi, 2007). In general, the region is poor in water resources. The main
constraints are the distance to the sea, the unimodal rainfall regime (Fig. 3.3), and
groundwater table of crystalline rock with poor aquifer conditions, therefore groundwater
levels vary greatly. For example, in the Atankwidi basin in the Upper East Region (UER) in
Ghana, variations between 1 m and 29 m have been reported, which strongly influence the
Schindler, 2009). Similarly, in south eastern
39
Burkina Faso the groundwater table reduces on average 0.6 mm day-1 in the cropping season,
thus the water withdrawal (76 l per capita per day (l/c/d)) greatly exceeds the provision
(20l/c/d), and according to the calculated recharge capacity of the aquifers (2% per year), the
projected demand will overtake the supply in 2030 (Sandwidi, 2007). Also the rainfall
follows a decreasing gradient from the South to the North (see Fig. 4.1). In the UER, Ghana
the monomodal rainfall regime of 3 to 5 months is from April to October, with between 900
and 1000 mm; the remaining seven months are dry (Kpongor, 2007 and Sanwidi, 2007). The
onset of the rainy season is generally stormy, but the effective rainfall for agriculture is low,
especially due to the high run off and evaporation. The latter can be exacerbated through the
Harmattan (Ouédraogo, 2004; Yilma, 2006 and Kpongor, 2007). Recurrent dry spells are also
observed, which are especially harmful during the planting season (June and July), as well as
recurrent droughts (Sanwidi, 2007). In Sudanian zone, annual rainfall ranges from 400 to
1100 mm from North to South with high spatial and temporal variability. For instance, in the
Kompiega basin in south eastern Burkina Faso, average rainfall 1959 2005 reaches 830 mm
year-1. In general, evaporation exceeds rainfall except during the rainy season when the basin
is recharged (Ouédraogo, 2004; Bagayoko, 2006 and Sandwidi, 2007).
40
Figure 3. 3: Temperature and rainfall monthly mean in Bawku, Upper East Region of Ghana
(1993 2011).
41
Lamto (5.02 W; 6.13
Baoul
Ivoire (Fig. 3.4).
It is surrounded by forest and hills and borders the southern rainforest. It covers an area of 27
km2 comprising 80% of moist savanna and 20% of forest. These characteristics make Lamto
not belonging to the northern Sudanian savannah, but to an intermediate vegetation type
called Guinea savannah (Diawara et al., 2014). Four seasons characterize the climate of
Lamto: a long dry season occurring in December-February, a long wet season during MarchJuly, a short dry season in August, and a short wet season in September-November. The mean
annual rainfall amount (~1212 mm) is less than that of the neighbouring synoptic stations
mainly located in the southern rainforest or at the same latitude, as, for instance, Gagnoa
where it reaches 1382 mm. Several variables (rainfall, temperature, sunshine, etc.) that
influence the climate are recorded daily at Lamto. The consequences that could involve
deforestation and bush-fire occurrences on its climate require particular attention and
assessment.
42
Figure 3. 4: Lamto Geophysical station
where many
climate and seismic parameters are measured; The station was created in 1962 by Maxime
Lamotte and Jean-Luc Tournier.
Figure 3. 5: Temperature and rainfall monthly mean at Lamto station, Guinea region in Cote
(1965 2004).
43
February, March and April are the hottest months at Lamto station. The mean temperature is
included
). The period corresponds to bush fire period
are part of the rainier which are April, May, June, July, and August. The area is characterized
by bimodal evolution of the temperature over the year.
Figure 3. 6: Intra seasonal Temperature mean at the Lamto station averaged over 1971-2000.
An increase in temperature is clearly seen through the temperature anomaly at Lamto station
in addition to the interannual variation of the temperature. The linear trend calculation is
giving a yearly increase of temperature abou
decade (Fig. 3.7). So as it
has been observed in the Burkina Faso, from 1964 to 2004 the temperature has increased at
observation made over the globe. The anomalies are computed based on 1971-2000 average.
Thus compare to the considered normal (1971-2000) the period from 1964 to 1985 the
anomaly is mainly negative this period correspond to the coldest of the series. Then from
1986 up to 2004 the anomaly is positive corresponding to the warmth period. Beside global
and regional climate impacts on temperature over the region some local phenomenon such as
44
deforestation, bush fire and aerosol emission could explain the increase in temperature. As it
is shown by Jones et al. (1999)
mean).
Figure 3. 7: Temperature Anomaly at the Lamto Station plotted with regard to climatology of
1971-2000.
45
3.2 Data Collection
3.2.1 Rainfall from station data
Rainfall measurements from rain gauge stations are conventionally considered the most
accurate and reliable source of rainfall data (Herrmann et al., 2005). However, this is only
true for point measurements or areas with a sufficiently dense network of rain gauges. So to
perform this study daily rainfall data have been collected data from two stations over Cote
which are Daloa and Lamto station covering the period 1964-2011and collected
respectively at the Societe de Developpement et Exploitation Aeroportuaire et Maritime
(SODEXAM) and at the Lamto station.
3.2.2 Climate Research Unit data
The choice was drown on the updated gridded climate dataset from Climate Research Unit
(CRU) version TS3.10.1, which presents a larger time cover with a wide spatial
representation over West Africa. In the worries of accurate representation of the mean state
and variability of the present climate which is important for a number of purposes in global
change research, coarse resolution datasets such as temperature and precipitation have been
adequate for monitoring and detection of climate change and GCM evaluation. Capturing
temporal variability is as important as the representation of spatial detail (New et al., 1999).
These include monitoring and detection of climate change; evaluation of General Circulation
Models (GCMs) and regional climate simulations; ground truthing, calibration, or merging
with satellite climatology; understanding the role of climate in biogeochemical cycling; and
construction of climate change scenarios (Carter et al., 1994).
The main sources of data are: National meteorological agencies, WMO 1961-90 global
standard normal, CRU global dataset of station time series, CIAT South America database
46
The principal sources used for the routine updating of the Climatic Research Unit (CRU)
monthly climate archives come through the auspices of the World Meteorological
Organisation (WMO) in collaboration with the US National Oceanographic and Atmospheric
Administration (NOAA, via its National Climatic Data Center, NCDC). The monthly
products were accessed through the Met Office Hadley Centre in the UK and NCDC in the
USA. Web links to these sources can be found in the supporting information (Harris, 2014).
These data comprise 1224 grids of observed climate, for the period 1901-2009, and cover the
global land surface at 0.5×0.5 degree resolution. The precipitation data (Eischeid et al., 1991;
Hulme, 1994) have been compiled by the CRU over the last 20 year. The original data have
been subjected to comprehensive quality control over the years. Updates for more recent
years and additional station data collated by the CRU have also been checked for
homogeneity and outliers. The correction of individual records requires detailed local
meteorological and station met information, which are not readily available (New et al.,
2000).
3.2.3 Global Precipitation Climatology Project
The Global Precipitation Climatology Project (GPCP) monthly precipitation analysis is a
globally complete, monthly estimate of surface precipitation at 2.5° x 2.5° latitude longitude
resolution that spans the period 1979 to the present (Adler et al., 2003; Huffman, et al.,
2009). However, the covered period for the model settlement was 2005 to 2007 It is a
merged, monthly analysis that employs precipitation estimates from low-orbit satellite SSM/I
and SSMIS microwave data to perform a calibration, that varies by month and location, of
geosynchronous-orbit satellite infrared (IR) data in the latitude band 40°N-S. These multisatellite estimates are combined with rain-gauge analyses (over land) in a two-step process
that adjusts the satellite estimates to the large-scale bias of the gauges and then combines the
47
adjusted satellite and gauge fields with weighting by inverse error variance. The monthly
product is typically produced about two months after the end of the observation month.
3.2.4 Vegetation Indices Data form GIMMS
The vegetation indices implanted for this study is the NASA AVHRR NDVI, covering the
period 1981-2012 with horizontal resolution of 8 km. It is derived from National Oceanic and
Atmospheric Administration (NOAA) satellites, and processed by the Global Inventory
Monitoring and Modeling Studies group (GIMMS) (Zhu et al., 2013; Donghai et al., 2014
and Jamali et al., 2015) at the National Aeronautical and Space Administration (NASA).
Spectral vegetation indices are usually composed of red and near-infrared radiances or
reflectance (Tucker, 1979), sometimes with additional channels included. According to
Cracknell (2001), these indices are one of the most widely used remote sensing
measurements. They are highly correlated with the photosynthetically active biomass,
chlorophyll abundance, and energy absorption (Myneni et al., 1995). The use of spectral
vegetation indices derived from AVHRR satellite data followed the launch of NOAA-6 in
June 1979 and NOAA-7 in July 1981 (Gray and McCrary, 1981). The AVHRR instruments
on NOAA-6 and NOAA-7 were the first in the TIROS-N series of satellites to have nonoverlapping channel 1 and channel 2 spectral bands. Overlapping red and near infrared
spectral bands precludes calculating a NDVI. The NDVI is calculated as NDVI5 (channel
22channel 1)/(channel 2 + channel 1) (Tucker, 2005). The Normalized Difference Vegetation
Index (NDVI) has become the most used product derived from NOAA AVHRR data
(Cracknell, 2001), largely from the use of NDVI datasets formed via maximum value
compositing (Holben, 1986).
The first generation NDVI data (NDVIg) from AVHRR sensors onboard the National
Oceanic and Atmospheric Administration (NOAA) 7 to 14 series of satellites have been
48
processed by the Global Inventory Modeling and Mapping Studies (GIMMS) group to a
consistent time series of NDVI and is made available to the research community. The latest
version, termed the third generation NDVI data set (GIMMS NDVI3g) has been recently
produced for the period July 1981 to December 2011 with AVHRR sensor data from NOAA 7
to 18 satellites. This data set specifically aims to improved data quality in the high latitudes
where the growing season is shorter than 2 months. It has also improved calibration that is
tied to the Sea-Viewing Wide-Field-of-View Sensor, as opposed to earlier versions of
GIMMS NDVI data sets that were based on inter-calibration with the SPOT sensor. The
availability of this new improved NDVI3g data set and its overlap with the Terra MODIS
LAI and FPAR products for the period 2000 to 2009 provides an opportunity to design and
implement a neural network algorithm to generate and evaluate the corresponding LAI and
FPAR data sets. These data sets will be termed LAI3g and FPAR3g henceforth and have the
following attributes: 15-day temporal frequency, 1/12 degree spatial resolution and temporal
span of July 1981 to December 2011 (Zhu et al., 2013).
3.2.5 Forcing Parameters
The experiment is based on the use of ICTP Regional Climate Model RegCM4 coupled with
land surface state model BATS. As vegetation degradation is an ongoing process, the
principle of this study is to anticipate the phenomenon by attributing a short grass to the
transition zone stated above which is normally arboretum savannah. The model will be set
base on rainfall from GPCP, CRU (New et al., 1999) and RegCM3 simulated for the
CORDEX experimentation and the temperature from CRU and RegCM3. The second
challenge is the period covered by the simulation which is four months (June-July-AugustSeptember) over two years (2005 and 2006). Can RegCM4 be able to capture the impact of
land surface change on rainfall and atmospheric patterns (TEJ, AEJ and West Africa monsoon
flux) at this range of time? For some authors, the intensity of these systems and their
49
latitudinal location influence not only the amount of rainfall but also its variability over West
Africa (Sylla et al., 2010). Some particularities of the region lead to make events difficult to
forecast. The instability the interconnection between different patterns, land-sea, surface state
atmosphere however, many studies have been realised with the previous version of RegCM
over the region. The model have been subject to some progresses and updates thus the choice
of the last version RegCM4 (Tompkins et al., 2005) the African Easterly Jet (AEJ) is one of
the key elements of the West African Monsoon (WAM) system, a major climatic feature of
sub-Saharan North Africa that also has important impacts on the tropical North Atlantic
(Sultan et al., 2003 and Sultan and Janicot, 2003).
The ECMWF Archive contains all observational data acquired in real time from the World
Global Telecommunications System (GTS) since the
beginning of daily operations in 1979. An archive of First GARP Global Experiment (FGGE)
level II-b data was also on-site at ECMWF. To provide a comprehensive set of input data for
ERA data were acquired from a number of additional sources (Simmons et al., 2007; Uppala
et al., 2008).
The initial and lateral boundary conditions used in this experience with RegCM4.4 is ERA
Interim 15 years reanalyse form ECMWF version TS3.10.1. simulation are obtained from the
new ERA Interim 2.5°×2.5° gridded reanalysis (Simmons et al., 2007; Uppala et al., 2008),
which is the third generation ECMWF reanalysis product. The main advances in this
reanalysis compared to ERA-40 are a higher horizontal resolution (0.75°×0.75° but available
also at 1°×1° and 2.5°×2.5°), four-dimensional variational analysis, a better formulation of
background error constraint, a new humidity analysis, an improved model physics, variational
bias correction of satellite radiance data, and an improved fast radiative transfer model. ERAInterim uses mostly the sets of observations acquired for ERA-40 with a few exceptions:
acquisition of a new altimeter wave-height that provides data of more uniform quality, use of
50
reprocessed Meteosat data for wind and clear-sky radiance, and new ozone profile
information from 1995 onwards.
3.3 Data Analysis
3.3.1 Observation data processing
3.3.1.1 Significance t-test of Differences
The method used to calculate the significance t-test values of differences between two
different decades (1980s, 1990s and 2000s). It was assumed that the variance of the two
samples is the same.
(3.1)
(3.2)
Where
and
is df
are number of observations for each experiment and the degrees of freedom
.
= average, s = standard deviation
(3.3)
(3.4)
This average
) is used to compute the decadal mean of rainfall and NDVI
The Pooled variance (PV) and Standard Error (SE) are given by (3.5) and (3.6)
(3.5)
51
(3.6)
The mean difference and the significance t-test are given by equation (3.7) and (3.8):
(3.7)
(3.8)
The cutoff t value is calculated based on df and the significance level.
3.3.1.2 Standardized Precipitation Index
The Standardized Precipitation Index (SPI) was developed by McKee et al. (1993, 1995) to
provide a spatially and temporally invariant measure of the precipitation deficit (or surplus)
for any accumulation time scale. The SPI is a probability index that considers only
precipitation. It is computed by fitting a parametric cumulative distribution function (CDF) to
a homogenized precipitation time series and applying an equiprobability transformation to the
standard normal variable. This gives the SPI in units of number of standard deviations from
the median. It is negative for drought, and positive for wet conditions. It is given by:
(3.9)
where
is the precipitation of the year i,
is the precipitation averaged over thirty years and
is the standard deviation of the series.
Table 3. 1 Drought categories from SPI (McKee et al., 1993)
SPI
Drought category
Mild drought
Moderate drought
Severe drought
Extreme drought
52
3.3.1.3 Correlation between rainfall and NDVI
The
correlation between rainfall and NDVI is plotted for the whole
region. Rainfall and NDVI monthly data over 29 years (1982 - 2010) the sample size N is
equal to 348. No time lag was observed between the two parameters.
To pick up the seasonal climatology, NDVI and rainfall have been averaged over months
from 1982 to 2010 over six sites throughout the region: two sites over the Guinea region
)
(Mali and Benin)
two sites over Sudan region
precisely, in Burkina
Faso and Niger
3.3.2 Model Setting and Simulation
3.3.2.1 Model Description and Simulation
Figure 3.8 shows the simulation domain with its relief shown by the contour lines. The study
area over West Africa is bordered by the red box.
53
Figure 3. 8: Model simulation domain with the topography (in contour) the zone of interest in
red box.
54
ICTP-RegCM4 is the last version of the regional climate modeling system developed by
ICTP. It is an improved evolution of its previous version RegCM3 (Giorgi et al., 2012). The
model description and the changes between RegCM4 and the previous versions are described
by Giorgi et al. (2012). However, it is a hydrostatic, compressible, sigma-p vertical
coordinate model run on an Arakawa B-grid in which wind and thermo dynamical variables
are horizontally staggered. The land surface model used to couple RegCM4 with is
Biosphere-Atmosphere Transfer Scheme (BATS) which will define the interaction land
surface-atmosphere (Dickinson et al., 1993). Convective precipitation is calculated with the
scheme of Grell et al. (1994) applying the Fritsch and Chapell (1980) closure assumption.
Resolvable precipitation processes are treated with the sub-grid explicit moisture scheme
(SUBEX) of Pal et al. (2000), which is a physically based parameterization including subgrid scale cloud fraction, cloud water accretion, and evaporation of falling raindrops. Table
3.1 shows briefly the summary of the simulation description. The projection is Normal
Mercator (NORMER), the SST type (OI_WK) and Era interim 15 (EIN15) for the boundary
conditions.
Table 3.2: Summary of the model configuration
Stanzas
Parameters
Dimensions
Domain
Number of points in N/S direction
100
Geo-parameter
Map projection
NORMER
Boundary conditions
SST type
OI_WK
Convection scheme
-
Grell over land and ocean
Simulation periods
-
2005-2006-2007
Subex
Cevapland*
0.003
Cevapland: parameter of the SUBEX moisture scheme, raindrop evaporation rate coefficient over land
Cevapoce: parameter of the SUBEX moisture scheme, raindrop evaporation rate coefficient over ocean
The modelling aspect of the thesis is approached through this section. It is a case study of the
impact of land use change on rainfall and atmospheric parameters over two years 2005 and
55
2006. The experiments are based on (1) simulating the rainfall and atmospheric parameters
initially without any changes on land surface cover as a control; then (2) the land cover over
West Africa savannah zone is replaced by the short grasses before performing the simulation
of the rainfall and atmospheric parameters. The model is evaluated with CRU and GPCP.
Comparison was made between atmospheric patterns before changes and after changes in
June-July-August-September (JJAS) of 2005 and 2006 and same with the rainfall.
3.3.2.2 Model Evaluation
The evaluation of the model was based on two parameters which rainfall and temperature.
The rainfall was validated with respect to two observation data namely CRU and GPCP. And
CRU was used to evaluate the temperature. The model has been set based on above
parameters during JJAS of 2005, 2006 and 2007.
3.3.2.3 Land Surface Model
For the soil moisture-
wo regions
(1) oceanic regions and (2) continental regions with and without snow cover (our case). For
the nonsea-ice-covered oceanic regions the surface temperature Tg1 is prescribed from
observational data in the standard model. For other regions, the computation of Tg1 depends
on the current conditions of snow cover soil moisture, type of surface and temperature of the
first layer of the atmosphere.
The vegetation part of the code is only executed for grid squares with vegetation cover
greater than 0.001. A mean wind within the canopy is obtained from the mean wind outside
the canopy times the square root of the drag coefficient which equals the friction velocity u.
The coefficient of transfer of heat and momentum from leaves is calculated (Fig. 3.9).
Foliage water is modified by intercepted rainfall. The temperature of the foliage (leaves) is
56
calculated. Any rain or snow intercepted by leaves in excess of their maximum capacity is
determined as falling to the ground and saved for soil-water or snow-budget calculations.
57
Figure 3. 9: Schematic of individual physical process (Dickinson, 1993)
58
3.3.2.4 Structure of BATS
In the BATS scheme (excluding the snow sub-model), there are 3 soil layers and 1 vegetation
layer, accounting for 7 prognostic variables: canopy temperature ( ), surface soil
temperature (
water (
), subsurface soil temperature (
), total soil water (
), surface soil water (
), and canopy water store (
), root-zone soil
). There are 18 surface-
cover types which are based on Olson et al. (1983), Matthews (1983) and Wilson and
Henderson-Sellers (1985). The soil type data are based on Wilson and Henderson-Sellers
(1985). For each vegetation type, there are about 27 derived parameters which determine the
morphological, physical and physiological properties of vegetation and soil.
Rawls et al. (1993) have provided a detailed review of the theory of soil water movement or
internal soil water fluxes. The rate of soil water movement is important in surface runoff,
groundwater recharge and evapotranspiration. Computation of the internal soil water fluxes
within the soil column of 1-10 m thick that is coupled to the atmosphere is difficult to
implement in climate models. The reasons are due to the coarseness of the grids, usually 2-5
levels, as limited by practical considerations, and the heterogeneous distribution of
topography, soil and vegetation types across the GCM grid-square.
The soil surface evaporation and the internal soil water fluxes in the BATS model are
parameterized based on the multilayer soil model integrations. The capillary movement of
water from the rooting zone into the surface soil layer is given by
(3.9)
And from the total column into the rooting zone is
(3.10)
59
Where
= ratio of soil water content within the total column to its maximum amount. The
gravitational drainage from the surface soil layer to rooting zone is
(3.11)
And from the rooting zone to the total column is
(3.12)
Where
And
3.3.2.5 Experimentation: Change in Land Surface State
The effect of deforestation on rainfall variability in the region was evaluated using the
regional climate model (RCM) RegCM version 4. The model was to simulate deforestation
by attributing short grasses to the transition zone located between the Guinea and Sudanian
region (
-
-
) which was originally characterised by woodland
savannah with tall grasses. Two experimentations were applied. For the first, a simulation (1)
was made without any change in vegetation cover and it was referred as the control (CTL).
Then, for the second, a simulation (2) was a sensitivity test. It was done after the change in
vegetation cover (change mentioned above) and it is referred as the sensitivity of the model to
the changes in vegetation cover (Sens). The period is detailed in the next sub section
(3.3.2.5). Figure 3.10 shows the change made in the vegetation cover. Countries crossed by
60
the change are Guinea,
, Ghana, Benin, Togo, Nigeria and Cameroon. Based on
the initial land cover distribution, the area was mainly occupied by the mixed woodland and
evergreen broadleaf tree. Some characteristics attributed to the short grasses will be found
into the general description of the vegetation and soil types in BATS (Dickinson et al., 1986).
For each of the land grid points, three other variables are defined in subroutine ALBEDOvisible solar albedo of vegetation ( <0.7 m), near-infrared albedo of vegetation ( >0.7 m),
and soil albedo. The land use type in BATS scheme defined the vegetation classes are
mentioned in Table 3.2 based on RegCM user manual version 4.4 (Giorgi et al., 2014).
Because of the lack of some types of vegetation in our region all the vegetation type
represented in the user manual are not shown in the Table.
61
Figure 3. 10: Initial land cover obtained after RegCM4 domain simulation before changes in
land cover (left) and after making changes in land cover (right); more details about the legend
in Table 3.2.
Table 3.3: Land cover/vegetation classes
Short grass
Evergreen needle leaf tree
Deciduous broadleaf tree
Evergreen broadleaf tree
Desert
Irrigated crop
Semi-desert
Bog or marsh
Ocean
Evergreen shrub
Mixed woodland
Forest/Field mosaic
62
The values for the albedo of vegetation were determined from a variety of sources, in
particular but also with reference to Monteith (1959), Barry and Chambers (1966), Federer
(1968), Oguntoyinbo (1970), Stewart (1971), Tucker and Miller (1977), Rockwood and Cox
(1978), Kriebel (1979), Fuller and Rouse (1979) and Kukla and Robinson (1980).
(3.13)
Where
is the albedo for a saturated soil and where the increase of albedo due to dryness
of surface soil is given for <0.7 m as a function of the ratio of surface soil water content
to the upper soil layer depth
,
(3.14)
63
Table 3.4: BATS vegetation/land-cover Dickinson, 1993
/
<
>
64
3.3.2.6 Case study of 2005 and 2006
The simulation covered three years 2005 to 2007. However focus has been on the years 2005
and 2006. The two years were selected due to the contrasts between some patterns which are
tudies have revealed differences
between 2005 and 2006. For example during the AMMA campaign the sub period called
Enhanced Observing Period (EOP) covering 2005-2007 is designed to provide a detailed
documentation of the annual cycle of the surface and atmospheric parameters from
convective scales of a few kilometres up to regional scales. The year 2005 was found to be
The regional SST
field is one of the key factors in West Africa rainfall (Lebel et al., 2010). The ocean plays a
key role in the WAM dynamics, especially during the onset phase. Warmer Gulf of Guinea
SSTs induces stronger surface sea breeze convergence in the Guinea Coast region and
abundant rainfall over the southern part of West Africa. The modification of the TEJ/STJ
leads to more humid atmospheric conditions throughout western Africa region and possibly
to changes in the triggering and behaviour of AEWs. The large-scale response is still
accompanied by changes in land use and vegetation cover. Furthermore, the monsoon has
started early in 2005 when in the following year (2006) it started later. One fundamental goal
of AMMA in terms of process understanding was to document the various atmospheric
structures described above at the appropriate scale in order to better understand their
interactions (Lebel et al., 2010).
65
Figure 3. 11: Diagram showing the differences on SST in Gulf of Guinea and monsoon onset
between 2005 and 2006 over West Africa.
66
Chapter 4
RESULTS AND DISCUSSION
4.1 Mean NDVI and Rainfall over Last Three Decades
4.1.1 Vegetation and Spatial Distribution of Rainfall over West Africa
The NDVI spatial distribution reveals a vegetation gradient increasing from the North to the
South (Fig. 4.1). With rainfall amount of about 600 mm yr-1 and monomodal rainfall pattern
(Konaté and Kampman, 2012), vegetation indices was low (below
in
extreme North of Burkina Faso, Niger, Mali and Mauritania. It corresponds to the Sahel with
vegetation composed of semi arid grasslands, savannah, steppes and thorn shrublands.
However, droughts and dust storms are frequent, and major constraints are overgrazing and
soil erosion and desertification (Boateng, 2013).
-savannah mosaic scattered with agricultural zones stretch up to northern
part of coastal countries
go,
Benin and Nigeria up to the central part of Africa (Congo watershed). The NDVI values are
in between 0.4 and 0.6. These regions are characterized by monomodal rainy season
alternated with pronounced dry season (Konaté and Kampman, 2012). The annual rainfall
varies between 800 and 1200 mm yr-1. The vegetation is dominated by grassland and trees
with low density. Lowland rainforest wetter, drier and mixed types with NDVI values above
0.6 are distributed over coastal regions from Gambia up to Cameroon. The landscape is
mostly flat with remnants of dense humid forests, with annual rainfall higher than 1500 mm
distributed in a bimodal pattern (Konaté and Kampman, 2012). The vegetation distribution is
consistent with rainfall spatial distribution.
67
Figure 4. 1: Map of West Africa showing the rainfall climatology (1971-2000) based on CRU
observation data (contour); NDVI climatology of yearly sum (shaded); the filled triangles
represent the site where rainfall and NDVI have been selected for the intra variability study.
68
4.1.2
NDVI Decadal Variability over West Africa during 1981-2010
4.1.2.1 Decadal Mean of NDVI
The decadal means of rainfall and NDVI are shown in Figure 4.2. Decades 80s, 90s and 00s,
are shown respectively in Figures 4.2(a-c) for the rainfall and Figures 4.2(d-f) for the NDVI.
No significance level of changes could be depicted the level of from one decade to another.
However, the spatial distribution of the rainfall shows that the rainfall is mainly concentrated
over two zones in coastal region of West Africa located from Bissau Guinea to the Southwest
of
and from Southeast of Nigeria to Cameroun due to high relief in these
regions. Ogungbenro et al. (2015) found intense mesoscale convective system (MCS) over
these regions of West Africa. This is explained by the effect of local climatological features
on MCS occurrences and the abundance of rainfall in these localities. It shows that NDVI
high values (above 0.6) are located in area with high rainfall concentration. This highlights
the positive relationship between rainfall and NDVI.
69
Figure 4. 2: Mean annual Rainfall (mm yr-1) shown in [a]; [b] and [c] for respectively decades
80s 90s and 00s and mean annual NDVI shown in [d]; [e]; [f] for respectively decades 80s;
90s and 00s.
70
4.1.2.2 Seasonal Variability of NDVI
Figure 4.3 shows the seasonal distribution of the NDVI averaged over three decades. Figures
4.3(a-d) indicate the seasonal mean of 80s corresponding to December-January-February
(DJF), March-April-May (MAM), June-July-August (JJA) and September-OctoberNovember (SON) respectively. DFJ is characterized by low values of NDVI over Sudan
savannah, the NDVI does not exceed 0.3(Fig. 4.3a). Only the extreme southwest of the region
has some high values of NDVI.
The NDVI values in MAM that is equal to 0.3 have decreased in the southern Sudan, while it
shows increasing values towards the Northern region (Fig. 4.3b) but remains unchanged in
the extreme North.
In JJA an important shrinkage of areas with 0.3 and 0.2 are observed for the benefit of the
growing vegetation over Sudan savannah meanwhile, there is reinforcement of NDVI over
Guinea region with appearance of areas with NDVI above 0.7 (Fig. 4.3c).
In SON the most observed change is the enlargement of the area with 0.7 as NDVI value over
western and southern part of the region.
The decades 90s (Fig. 4.3(e-h)) and 00s (Fig. 4.3(i-l)) show the same spatio-temporal
distribution scheme of NDVI over the region as obtained in decade 80s. However, the main
seasonal change from decade to decade is the increase of NDVI values (above 0.7) and
enlargement of area with high NDVI values during the last decade 00s over Guinea Coast.
71
Figure 4. 3: NDVI decadal mean showing changes in vegetation cover with progressive
southward increase in decade 80s (a, b, c and d); decade 90s (e, f, g and h) and decade 00s (i,
j, k and l).
72
4.2 Spatio-Temporal Distribution of the Rainfall over Three Last Decades
This section describes of the rainfall over thirty year period (1981-2010). The whole period is
split into three decades. First of all, all the decades were compared to the normal. Then an
inter-comparison was made between the different decades. Furthermore, the spatial
distribution of the SPI over some sites is given. Finally, the section was concluded by
observing the seasonal changes between decades.
4.2.1 Changes compare to thirty years climatology
The comparisons between each decade and the thirty years climatologic mean (1981-2010)
are shown in Figure 4.4. Compare to the last three decade means (the normal), decade 80s
indicates the driest years (Fig. 4.4a). A rainfall deficit was observed all over the region with
significant decrease in northern part of Nigeria and Liberia while the western and eastern
Sahel were the most affected regions. Decade 90s was wetter compare to the normal. The
rainfall has increased in the regions where the deficit was observed namely western and
eastern Sahel (Fig. 4.4b). A slight but not significant deficit was observed in the central part
of Ivory Coast. In 00s, the rainfall was observed to have increased over some areas. Thus, the
wet condition was maintained over Liberia and only extended to Ivory Coast in 00s, whereas
it has decreased over Guinea regions. Some deficits are mainly observed over Benin and
Nigeria (Fig. 4.4c).
Among the three decades, decade 80s was the driest. The whole region suffered below
normal rainfall. This result is in agreement with earlier studies in the West African region
(Nicholson et al., 1998; Giannini et al., 2003; Zeng, 2003; Dai et al., 2004; Lebel and Ali,
2009 and Lebel et al., 2010). However, there was a recovery in 90s over many areas of the
region (Nicholson, 2005; Evan et al., 2015 ) this drought was confirmed in 00s over some
73
because some of areas are going back to deficit conditions.
Figure 4. 4: Spatial distribution of rainfall significant changes in decade [a] 80s; [b] decade
90s and [c] decade 00s compare to 30 years average (1981-2010) over West Africa at a level
of 95%.
74
4.2.2 Decade to decade changes
The rainfall evolution from decade to decade was computed using the significance t-test at a
level of 95% between the decadal mean of the considered decades (Fig. 4.5). Figure 4.5a
shows the difference between mean rainfall of 90s and 80s. Positive changes were observed
over large area of the region with strong differences (greater than 300 mm yr-1) in Northern
Nigeria, Liberia and Guinea. This result confirmed the positive anomaly observed in sub
section 4.2.1 in 90s. Over western and eastern part of West Africa, the rainfall
increase in
90s is significant at 95%. Finally, a slight decrease in the rainfall is observed over some parts
of
and Ghana. Also, Figure 4.5b shows the differences between decade 00s and
80s, similar to 90s, decade 00s indicates wetter than 80s over large area of West Africa except
in some part of Ghana and Benin where the rainfall had decline. It is still noted some narrow
zones where the change is positive but not significant in Mali and Mauritania.
Figure 4.5c presents the difference between 00s and 90s. This was the most contrasted in
terms of spatial distribution of positive and negative differences. A decrease in rainfall was
shown in many areas mainly over Nigeria, Benin, Togo, Ghana and the Republic of Guinea
during 00s after a recovery period in the 90s. On the other hand, the rainfall kept increasing
mainly over
, Senegal and Mauritania. The areas which were going back to
drought condition in 00s (see sub section 4.2.1) are still wetter than 80s. So the level of the
drought in 80s is not achieved.
75
Figure 4. 5: Decadal changes in rainfall seasonal spatial distribution over West Africa at a
level of 95%. Blue lines are areas with significant changes.
76
4.2.3 Rainfall distribution
The time latitudinal diagrams of the rainfall decadal mean and the rainfall decadal anomalies
in 80s, 90s and 00s are shown in Figure 4.6.
Decade 80s was marked mainly by weak magnitude of rainfall (Fig. 4.6a) hence the
negative anomaly all over the study period (Fig. 4.6d). The rainfall amount above 240 mm
month-1 is reached over Sudanian region later in September.
In Decade 90s an increase in rainfall in August over Sudanian region was observed (Fig.
4.6b). Furthermore, an enlargement of the area with the rainfall amount above 240 mm
month-1 was also observed. However the anomaly shows a clear contrast between the long
rainy season and the short rainy season (Fig. 4.6e). The long rainy season was wetter
compared to the short rainy season. The third decade at some slight difference was similar to
the second decade (Fig 4.6c and Fig. 4.6f).
The differences between the three decades are mainly about the rainy period and the rainfall
magnitude. From 80s to 90s the rainfall has shifted on time. Therefore, the extreme North
was reached by the rainfall belt in August in 80s however it is reached earlier around July in
90s and 00s. In terms of rainfall magnitude, it was more intense in 90s and 00s contrary to
80s.
77
Figure 4. 6: Time latitudinal diagrams of rainfall seasonal mean (a, b and c) in decade 80s;
decade 90s and decade 00s respectively and seasonal anomaly (d, e and f) for decade 80s;
decade 90s and decade 00s respectively.
78
4.2.4 Upward Trend of the Rainfall over the Region
To monitor the severity of the drought, the Standardized Precipitation Index (SPI) is
computed at different sites over Sahelian, Sudanian and Guinea regions. It shows at these
sites an alternation of dry and wet sequences (Fig. 4.7). The deficit in rainfall occurred
mainly at the beginning of 80s till the middle or the end of decade 80s for some sites over the
Sahel and Sudanian region. Numerous studies referred to early 70s as the beginning of the
drought over the West African region. Sanni et al. (2012) evaluating the drought severity in
the Sudano-Sahel Region (SSR) of Nigeria, they found that most of the drought severity with
the highest magnitude occurred between the 70s and 90s. However, contrary to number of
finding, no severe drought sequences were noted during the period of this study. Thus, this
can be explained by the selected climatology period which is 1981-2010. However, it was
observed that since early 90s and 00s the recurrence of positive value of the SPI over the
region. More wet years are observed in decade 90s at all the considered sites. This
observation is in agreement with Dai et al. (2004) who found that large multi-year
oscillations appear to be more frequent and extreme after the late 80s than previously (before
1980). Thus for them, that may mean
become more unstable and
prone to droughts after the prolonged severe droughts from the early 70s to late 80s (Dai et
al., 2004).
79
Figure 4. 7: SPI at six locations in Burkina Faso and Niger (Sahel); Mali and Benin (Sudan)
and Ivory Coast and Ghana (Guinea Coast) over West Africa showing more wet condition
mainly apart from decade 1990. The climatology is based on 1981-2000 rainfall mean.
80
4.2.5 Changes over Seasons
The differences between seasonal rainfall in 80s, 90s and 00s were shown in Figure 4.8. The
first differences between decade 90s and 80s. The result for DJF shows very narrow areas
with significant positive changes in Liberia and Guinea. However, a deficit was observed
over the southern part of Nigeria (Fig. 4.8a). For MAM there was an enlargement of the area
with significant positive changes over Guinea coast and central Sudan region (Fig. 4.8b). JJA
period shows positive changes and moved northward in the Sahel region (Fig. 4.8c) with an
increase in rainfall trend in 90s. Some decreases have been observed in the central part of
little rainy season (SON),
extreme West and East of the region were the wettest (Fig. 4.8d). These regions with
significant positive changes coincided with the areas where rainfall had significantly
increased between 80s and 90s.
Almost the same observation was obtained when compared 00s and 80s (Fig. 4.8(e-h)), this
implies that decade 00s and 90s was not much different in terms of rainfall spatial
distribution at a seasonal scale. This could explain the scatted distribution of significant
differences between the two decades. However, the difference between 00s and 90s shows an
important mitigated distribution of the rainfall in term of quantity. The rainfall has decreased
over some areas during the last decade namely from Ghana to Nigeria and Guinea to Sierra
Leone in MAM and JJA (Fig. 4.8(i-l)).
There was consistent increase in rainfall over some areas during the three decades, for
instance
nd Liberia where an
increase in rainfall was observed during the dry season (DJF). The result was opposite over
southern part of Ghana where rainfall amount shows a decline in JJA. However, in most cases
it is fluctuating between wet and dry conditions. The differences showed that 90s was wet
81
than 80s over a large area of West Africa except
and Nigeria where 90s was dry
in JJA and also in a southern part of Ghana for SON.
Figure 4. 8: Decadal changes in rainfall seasonal spatial distribution between decade 80s-90s
(a, b, c and d), decade 90s-00 (e, f, g and h) and decade 80s-00s (i, j, k and l) over West
Africa at a level of 95%; Blue lines are areas with significant changes.
82
However, the observed increase in rainfall between 80s and 90s over large area could be
explained the fact that from 90s till now, more wet years are observed (Fig. 4.8). High rainfall
amount and a greater number of wet periods were observed contrary to the 80s within which
there were many dry sequences of rainfall over many areas throughout West Africa
(Nicholson, 1993 and Omotosho, 2008). The extreme and severe droughts occurred in 1983
and 1984 in many areas of West Africa. This result supported previous work on extreme and
severe drought in the region based on Palmer indices (Dai, 2011 and Hua et al., 2013).
Compare to 80s, there was an increase in rainfall in 00s over the western part of West Africa
Studies have revealed a
rainfall recovery over Sahel regions starting in early 90s and still going on in some areas
(Nicholson, 2005; Herrmann et al., 2005; Ali et al., 2008 and Lebel and Ali, 2009).
The whole Guinea zone have also registered a slight increase in the rainfall amount from
, the same result was found in the northern part of Nigeria by
Herrmann et al. (2005). However, in the last decade there was appearance of some areas with
negative changes but very small within Nigeria and Benin, this implies non-coherence of the
spatial distribution and Changes in surface energy budgets resulting from land cover change
4.3 Spatio-Temporal Distribution of the Vegetation over Three Last Decades
This section describes the vegetation dynamics through the Normalized Difference
Vegetation Index (NDVI). The decadal anomaly was firstly computed, followed by the inter
comparison between decades. Finally, a spatio temporal variability and the seasonal change
over decade were analysed.
83
4.3.1 NDVI Decadal Anomaly Variability over West Africa
Compare to the whole period, decade 80s was widely negative in term of vegetation growing
all over the region. This was highlighted throughout Guinea and Sudanian region. However
some narrow areas with positive greenness were observed over central part of Ghana and
Cot
a).
In 90s the re-greening has started in central West African region and over West Sudanian
is period
the change in vegetation indices is still negative over Guinea region (Fig. 4.9b).
In the last decade 00s the rate of re-greening is the highest over Guinea, Sudanian regions and
over western Sahel. So contrary to 80s, in addition to some areas over Sudanian region, the
positive change has extended to the coastal region, meaning that the vegetation has greened
progressively southward when the decreases are becoming more significant over Sahel region
(Fig. 4.9c). In the Gulf of Guinea, the re-greening is the highest all over the West African
region.
Above this region it is still observed
significant. Furthermore, an enlargement of the area with negative differences is observed
kina. These
refer to a decrease of the vegetation cover over these areas.
84
Figure 4. 9: Annual significance t-test computed between decades showing significant
positive and negative changes in vegetation cover over West Africa in a, b and c for decades
80s; 90s and 00s respectively compare to 30 years average (1981-2010) at a level of 95%.
85
4.3.2 NDVI Decadal Variability over West Africa
The difference between 90s and 80s shows positive increase in vegetation cover over
Sudanian
4.10a). After the severe drought
occurred in 80s, the vegetation has been significantly greened over the Sudanian region
ian region during
this period could be due to the rainfall recovery which has started in decade 90s and may
have induced some vegetation cover recovery.
It was not the case in Guinea region where there were no changes or even negative changes
were observed during this period. When over Guinea region significant increase in vegetation
compare to 80s.
central West African region and to
Sudanian region of West Africa (Fig.
4.10b). This positive difference is highlighted over coastal region by significance difference.
So contrary to 80s, the positive change has extended to the coastal region, meaning that the
vegetation has greened progressively southward when the negative changes are becoming
more significant over Sahel region. The NDVI provides information about the green leaf area
index (Myneni and Williams, 1994). Anyamba and Tucker (2005) using the same set of data
but over 23 years found a gradual recovery from drought conditions.
The comparison between 00s and 90s shows the persistence of significant positive changes in
the coastal region
still observed some positive changes up
however, not significant (Fig. 4.10c). Furthermore, an enlargement of area
with negative differences is observed over Sah
over Niger, Mali and Burkina. These referrers to a decrease of the vegetation cover over these
86
areas. Therefore, the opposite phenomenon is observed over Guinea coast with the greening
process going on until 90s.
Figure 4. 10: Annual significance t-test computed between decades showing the positive and
negative changes in vegetation cover over West Africa between decade 80s-90s [a]; decade
80-00s [b] and decade 00s-90s [c] at a level of 95%.
87
4.3.3 Decadal Change on NDVI
The time latitudinal diagrams of the NDVI decadal mean and decadal anomalies in 80s, 90s
and 00s are shown in Figure 4.11.
First,
in 80s it is shown that
April period corresponding to the dry season (Fig. 4.11a). Then there was a narrowing of this
area tow
, they are
concentrated in Guinea zone between April and November
Sudanian region during August-September. However, the anomaly does not show significant
difference in 80s (Fig. 4.11d). In 90s the same spatio-temporal distribution of the NDVI was
observed (Fig. 4.11b) like in 80s, except the fact that in the South
-
sa
decrease in NDVI from mid July up to September. The anomaly shows a contrast between the
first eight months of the year (January to August) and the last four months (August to
December). The positive anomalies are observed during the first eight months contrary to the
last four months where negative anomalies are observed (Fig. 4.11e). In 00s the NDVI value
has increased in the South the 0.5 NDVI is observed early apart from January and the 0.6
starts before April (Fig. 4.11c). Furthermore, there was an appearance of 0.7 NDVI value
between October and November in the Sudano-Guinea region. However, the decrease has
extended from the Guinea up to Sudano-Guinea region between July and September. The
anomaly shows the same contrast as in 90s (Fig. 4.11f).
88
Figure 4. 11: Time latitudinal diagrams of seasonal NDVI shown in a, b and c for decade 80s,
90s and 00s respectively and NDVI seasonal anomaly
shown in d,
e and f for decades 80s, 90s and 00s.
89
4.3.4 NDVI seasonal variability over West Africa
The difference between decade 80s and 90s shows narrow significant positive changes
scattered mainly
and Ghana in DJF (Fig. 4.12a). In the following season
MAM positive changes (Fig. 4.12b), have slightly moved northwards in the sub Sudanian
region with some significance in Guinea Republic. However, it is noted some negative
changes in the South of
voire, Ghana and Guinea in JJA (Fig. 4.12c). Then in SON
(Fig.
4.12e).
The difference between decades 90s and 00s is positive and significant in the Guinea region
from Senegal to Nigeria (Fig. 4.12f). In MAM there is not upwards movement of the positive
change however, the negative change has highly increased over Sahel region (Fig. 4.11g).
The following seasonal difference in JJA is slightly positive in some narrow places over
Sahel band. When in the Guinea region the change has become significantly negative (Fig.
4.12h). The difference between DJF NDVI of decade 00s and 80s shows significant positive
change over Guinea region (Fig. 4.12i). In MAM the change is still positive over Guinea
region negative change is observed above 12
4.12j). For JJA, the positive change is
scattered over Sahel region when over Guinea region the negative change is widely
significant (Fig. 4.12k). In SON the positive change is mainly significant and located between
4.12l).
90
Figure 4. 12: Positive and negative changes in vegetation cover over West Africa between
seasons in decade 80s-90s (a, b, c and d); seasons in decade 80-00s (e, f, g and h) and seasons
in decade 00s-90s (i, j, k and l) at a level of 95%.
91
From these analysis, it is observed that the difference between decade 80s and 90s show
northwards seasonal increase in vegetation covert. But this difference seems to be less
significant. Compare to the previous test, the positive difference between the second and third
decades is almost sedentary for the two first seasons DJF and MAM then widely negative in
JJA. The difference between 90s and 00s is better highlighted than the previous difference
thus more significant. The last test is almost similar to the second meaning that 90s compare
to 80s is quite similar to 00s compare to 80s this induces that there was not important spatial
changes between 90s and 00s. The vegetation covert seems to be constant.
With respect to the NDVI spatial distribution, it is noted that the test of means difference
between the three decades shows a strong seasonal variability from a season to season and a
decade to another (Fig. 4.12). The difference between decade 80s and 90s is not showing in
general an expended significant changes. However, the positive changes occur in Guinea
. The most important changes
occur in SON over sub Sudanian and Sahel zone
negative changes throughout Guinea region. The difference between decade 00s and 90s is
significantly positive all over Guinea region in DJF and MAM contrary to the Sahel region
where the changes are negatives mainly in Niger, North Burkina Faso and Mali. JJA is
relatively negative over Sahel and Guinea region. The last test which is computed between
00s and 80s shows
DJF and MAM. In JJA, it is identified some negative changes over the Sahel and Guinea
regions. Except the southern part of
, Liberia and some regions in Nigeria, the
whole
and 00s.
92
4.3.5 Frequency distribution of the NDVI
The level of vegetation greenness was defined based on the method applied by Lim and
Kafatos (2002) who divided vegetation
Over Niger site, the high frequency is located in between 0.24 and 0.25 (Fig. 4.13) which
ur case the Sahel region of West Africa where the mean
rainfall is about 469.76 mm yr-1.
Figure 4. 13: Annual frequencies and distribution of NDVI at Niger (left) and Burkina Faso
(right) sites over Sahel region.
The vegetation of West Africa is characterised by a combination of factors related to both
climate and soil, essentially exhibiting the same longitudinal zonation as rainfall. The
structure of the vegetation also changes progressively from North to South. The further
South, the taller the vegetation, the greater the proportion of woody species (trees, shrubs,
bushes), and the higher amount of ground cover.
occasional woody species of small trees or shrubs. Grasses are perennials, generally not taller
than 80 cm, and animals; woody species are often thorny, like the typical Acacia. This is the
Sahel proper. Here the surface is to a large extent bare soil and the vegetation tends to be
93
clustered in sites of favourable conditions of soil or runoff, creating a mosaic pattern
(Nicholson, 1993).
The higher frequency was obtained between 0.32 and 0.34 at Burkina Faso site this range
belongs to medium green in Lim and Kafatos (2002) classification (Fig. 4.14).
Figure 4. 14: Annual frequencies and distribution of NDVI at Mali (left) and Benin (right)
sites over Sudanian region.
The Sudanian Savanna is characterized by the coexistence of trees and grasses. Dominant tree
species are often belonging to the Combretaceae and Caesalpinioideae some Acacia species
are also important. The dominant grass species are usually Andropogoneae, especially the
genera Andropogon and Hyparrhenia, on shallow soils also Loudetia and Aristiada. Much of
the Sudanian Savanna region is used in the form of parklands, where useful trees, such as
shea, baobab, locust-bean tree and others are spared from cutting, while sorghum, maize,
millet or other crops are cultivated beneath.
94
Figure 4. 15: Annual frequencies and distribution of NDVI at
(right) sites over Guinea region.
(left) and Ghana
The Guinea Forests of West Africa hotspot encompasses all of the lowland forests of political
West Africa, stretching from Guinea and Sierra Leone eastward to the Sanaga River in
Cameroon. This includes the countries of Liberia,
, Ghana, Togo, Benin, and
Nigeria, which maintain remnant fragments of the forests. The hotspot also includes four
islands in the Gulf of Guinea: Bioko and Annobon, which are both part of Equatorial Guinea,
and Sao Tomé and Principe, which together form an independent nation. Bioko is a
continental-shelf island, whereas the remaining three are oceanic.
The Guinea forests consist of a range of distinct vegetation zones varying from moist forests
along the coast, freshwater swamp forests (for example, around the Niger Delta), semideciduous forests inland with prolonged dry seasons. Of all West African countries, only
Liberia lies entirely within the moist forest zone, although a substantial portion of Sierra
Leone also falls within the boundaries.
4.4 Relationship between Rainfall and NDVI over West Africa
This section focused on the relationship between rainfall and vegetation. It is first of all based
on statistical analysis of the relationship between the two parameters at different sites over
West Africa.
95
4.4.1 Intra Annual Variability of Rainfall and NDVI
The statistical analysis of the rainfall and NDVI data was performed over three major
climatic zones in West Africa (see Figure 4.1). The behaviour of the vegetation cover was
analysed from 1982 to 2010. Parameters like rainfall and NDVI mean and standard deviation
are given in Table 4.1. The correlation between rainfall and NDVI is also shown. The
interannual variability over Guinea region is the highest in term of both rainfall and NDVI
compare to Sahel and Sudan where most correlation are the Sahel region 0.56 (Niger) and
0.58 (Burkina) followed by the Guinea region 0.45
correlation is weak for the Sudanian region 0.24 and 0.21 at respectively Mali site and Benin
site. The low correlation could be due to the time lag between rainfall and vegetation growing
which is not always systematic as it is assumed in this work.
Table 4.1: Descriptive Statistics for the rainfall and NDVI time series
Rainfall
Sahel
Sudan
Sites
Niger site
Burkina site
Mali site
Benin site
Guinea
Ghana site
Mean
469.76
595.48
1126.6
923.54
1244.4
1198.3
NDVI
STD.
117.04
123.49
136.47
119.88
181.37
141.25
Mean
0.24
0.33
0.50
0.49
0.63
0.63
STD.
0.012
0.018
0.016
0.018
0.036
0.040
Corr
0.56**
0.58**
0.24
0.21
0.45*
0.40*
**. Correlation is significant at the 0.01 level (2-tailed);
*. Correlation is significant at the 0.05 level (2-tailed).
The rainfall seasonality could be responsible for the strong variability of rainfall and NDVI
over Guinea region where the region is characterised by four seasons two rainy seasons and
two dry seasons. Also the land cover type is an important factor to Wang et al. (2003) found
an average of 0.85 for grassland and 0.79 for forest.
96
4.4.2 Relationship between Rainfall and NDVI over Guinea Region of
The correlations between NDVI and rainfall vary strongly from one month to another (Table
4.2 and 4.3). The rainfall of some months is significantly correlated with its NDVI. It is the
case of April, May, August and September at Lamto and January, February, April and
December at Daloa. But it is not always the case in some cases it is observed 2 or 3 months
gap between rainfall and NDVI. That is the case of July rainfall and NDVI of August, August
rainfall and NDVI of October and November at Lamto. However, the correlation is
significantly negative in November (-0.60).
97
Table 4.2: Monthly Cross-correlation between AHVRR NDVI and Rainfall at Lamto station
(1981-2000); Correlation is significant at 0.05 level and (**) correlation is significant at 0.01
level. The indices p and n are respectively rainfall and NDVI.
JANp
FEBp
MARp
APRp
MAYp
JUNp
JULp
AUGp
SEPp
OCTp
NOVp
DECp
JANn
FEBn
MARn
APRn
MAYn
JUNn
JULn
AUGn
SEPn
OCTn
NOVn
0.29
0.13
-0.22
-0.04
0.03
-0.04
-0.10
-0.31
-0.23
-0.22
-0.19
0.39
-0.17
-0.01
0.17
-0.26
-0.27
0.05
0.07
-0.45*
-0.23
0.16
0.08
0.41*
-0.18
0.09
0.24
0.08
-0.09
-0.18
0.56*
0.16
-0.25
-0.01
-0.17
0.03
-0.28
0.34
0.41*
-0.26
0.22
-0.32
-0.31
-0.18
0.44*
-0.03
0.13
0.24
0.15
0.18
-0.24
0.22
0.55*
-0.07
0.04
-0.04
0.59**
0.27
0.46*
-0.60**
0.43*
-0.25
-0.15
-0.07
-0.25
0.01
DECn
-0.27
-0.23
-0.30
0.13
-0.14
-0.06
0.01
-0.17
0.02
-0.06
-0.24
0.12
98
The situation at Lamto and Daloa is almost similar but with some slight differences.
Therefore at Lamto the February to February correlation is low 0.39 contrary to Daloa where
the two parameters are significantly correlated 0.54, and then in May the correlation is
significant between rainfall and NDVI at Lamto (0.41) when it is too weak at Daloa (-0.01).
For both localities June and July are not showing any correlation between the two parameters
on time. The rainfall of March and April is negatively correlated with NDVI of April and
positively correlated with NDVI of October at Daloa.
This could be explained by the rainfall regime and the vegetation types over the two sites.
Lamto is located in the Sudano Guinea savannah zone where the rainfall has much influence
on vegetation contrary to Daloa site where there is still some forest and the vegetation is less
influenced by the rainfall. However, the effect of rainfall can be seen during the dry season
like in December, January and February (Table 4.3).
99
Table 4.3: Monthly Cross-correlation between AHVRR NDVI and Rainfall at Daloa (19812000); Correlation is significant at 0.05 level and (**) correlation is significant at 0.01 level.
The indices p and n are respectively rainfall and NDVI.
100
4.4.3 Rainfall Intra-seasonal Variability
The intra annual variability of the rainfall at the six selected sites over the three main climatic
zones of West African region is differently observed. The selected sites over Sahel region are
and are located in Niger (Fig. 4.16a) and Burkina Faso (Fig. 4.16b). The
rainfall regime is mono modal with the peak in August. Over the two Sahel sites the
difference between the three decades is mainly observed at Niger site with 90s as the rainiest
(about 200 mm in August) and 80s the driest (about 100 mm in August). This finding is also
seen by Hagos and Cook (2007). At Burkina Faso site the visual analysis does not show
considerable changes.
Over Sudanian region the rainfall regime is still mono modal with the peak in August 350
mm and 250 mm respectively in Mali and Benin. But the rainfall amount is higher compare
to the rainfall in Sahel. The important change noted was the shift of the rainy period during
the last decade at Mali site (Fig. 4.16c). This is due to a late onset and late secession of the
rainfall and less precipitation in the core of the rainy season finding in agreement with Louvet
et al. (2015). In Benin the rainy period did not change only some slight change has been
observed in rainfall amount in August (Fig. 4.16d).
At last the two selected sites over Guinea regions
(Fig. 4.16e) and Ghana (Fig.
4.16f) show a bimodal regime of the rainfall. The two peaks are obtained respectively in
May-June and October due to the presence of two rainy seasons. Detailed description of this
regime could be found in Konate and Kampmann (2012). The change occurs in the second
rainy season with and an increase in the rainfall over the last decade. However many studies
and research programmes (African Monsoon Multidisciplinary Analysis) built up over West
Africa have shown the deep implications of the monsoon in the rainfall amount and spatial
101
distribution over the West African region (Pospichal et al., 2010; Gosset et al., 2010 and
Lebel et al., 2010).
Figure 4. 16: Decadal rainfall at the monthly timescale plotted for the six selected sites in the
Sahel region ([a] Niger and [b] Burkina Faso), the Sudanian region ([c] Mali and [f] Benin)
and the Guinea region ([d]
and [e] Ghana).
102
onset and the early cessation. Thus, the rainfall amount increases whereas the length of the
rainy seasons decreases. The distribution of the rainfall over the year has changed. The
rainfall drops in August due to the monsoon jump (Hagos and Cook, 2007). Hence, during
this period, the rainfall is mainly due to locale conditions and some remote parameters. This
exposition is supported by Odekunle and Eludoyin (2008). Yaw and Ian (1994) found
positive correlations between the high SST in the Gulf of Guinea and high rainfall; same
colder oceans are associated with lower rainfall in June-July and September-October in
Ghana. Many studies and research programmes (African Monsoon Multidisciplinary
Analysis) built up over West Africa have shown the deep implications of the monsoon in the
rainfall amount and spatial distribution over the West African region (Pospichal et al., 2010;
Gosset et al., 2010 and Lebel et al., 2010).
4.4.4 NDVI Intra-seasonal variability
The intra annual variability of NDVI is analysed over decades 80s, 90s and 00s (Fig. 4.17) at
the six selected sites. Like the rainfall, the NDVI has a monomodal evolution over the year.
The NDVI is weak over Sahel (<0.4) at both sites Niger (Fig. 4.17a) and Burkina Faso (Fig.
4.17b). Compare to the three decades, the NDVI has decreased during the first semester in
00s. However, the peak of 80s NDVI is the lowest. The vegetation has increased in JulyAugust-September during the two last decades (90s and 00s).
In the Sudanian region, no significant change can be observed from January to July. On the
other hand, it can be observed an increase of the NDVI during decades 90s and 00s from
August to December. As shown in the previous section the higher amount of rainfall occurs in
August however decade 00s was not the rainiest but the vegetation response was the highest
(Fig. 4.17c and 4.17d).
103
In Guinea region the NDVI seasonal variability follows the rainfall regime with two peaks.
No important visual change is observed between decade 80s and decade 90s. However, it is
observed a strong intra annual variability over the last decade. The vegetation starts growing
early in the last decade and end the growing process late however it decreased in August (Fig.
4.17e and 4.17f).
104
Figure 4. 17: NDVI decadal mean averaged over months at six different points over Sahel
region [a] Niger and [b] Burkina Faso, Sudanian regions [c] Mali and [f] Nigeria and Guinea
105
4.4.5 Rainfall and NDVI Monthly Climatology
Figure 4.18 shows the NDVI and rainfall intra annual variability. These are monthly mean of
rainfall and NDVI over 30. Two types of annual distribution of the two parameters are
shown. The NDVI shape is following the rainfall seasonal distribution. In the Sahel region
the rainfall is mono modal. The peak is achieved in August (around 150 mm) when the NDVI
peak occurred one to two months later (Fig. 4.18a and 4.18b).
106
Figure 4. 18: NDVI and rainfall monthly mean averaged respectively over 1981-2012 and
1981-2006 at six different points over Sahel region [a] Niger and [b] Burkina Faso, Sudanian
region [c] Mali and [f] Benin and Guinea region [d]
and [e] Ghana.
107
The site selected over Sudanian region at the southern part of Mali (Fig. 4.18c) and northern
part of Benin (Fig. 4.18d) the rainfall can achieve 300 mm in August and the NDVI 0.7 in
October. In April the NDVI start growing and will continuous up to September. The
vegetation over these regions is more sensitive to rainfall. Mali and Benin sites are close to
The Guinea region which is characterized by two rainy seasons, two peaks of NDVI peaks is
observed (Fig. 4.18e and 4.18f). Contrary to the rainfall, the second peak of NDVI is the
highest as shown in
(Fig. 4.18e). The amplitude is higher in the case of mono
modal distribution of NDVI namely in Sahel regions. The amplitude of seasonal evolution
shows that the saturation occurs early in region with high density of vegetation (Eklundh and
Olsson, 2003) as Guinea region of West Africa contrary to Sahel where vegetation takes some
time to accumulate water.
The period of year identified as having a consistent upward trend in time series NDVI
corresponding to the beginning of measurable photosynthesis in the vegetation canopy are
mid June over Sahel region, April for Sudanian region and early in March over Guinea
region.
The response of the vegetation to the rainfall seasonality is mainly seen in MAM and SON
after a period of higher rainfall. Lamb (1980) revealed that approximately 83% of the rainfall
occurs within this period of the year. In turn, they alter the equilibrium state of the vegetated
surfaces (Monteny, 1986). As shown by Fensholt and Rasmussen (2011) and Herrmann et al.
(2005), an annual positive trend in vegetation greenness and rainfall were noted throughout
the study over the sub Sudanian region during the last decades. This fact has been assigned to
a desertification reverse by some authors. The rate of the increase of vegetation greenness can
reach 50% in some areas in parts of Mali, Mauritania and Chad. Some areas where the
108
changes are not significant are observed thus, moving from the south of Benin to the
4.4.6 Relationship between Rainfall and NDVI over West Africa
For the relationship studies, focus was made on correlation between rainfall and NDVI. The
correlation between rainfall and NDVI is widely positive (>0.4) over large area of West
4.19). In decade 80s the correlation can reach 0.8 over
Republic of Guinea, Senegal and in the Southern part of Mali, this positive difference is
mainly significant at 95% level over the region (Fig. 4.19a). However it is weak over coastal
region from Sierra to Nigeria mainly included in between -0.4 and 0.4. The same situation is
observed in decade 90s nevertheless with some slight differences (Fig. 4.19b). So contrary to
80s, the area with high values of correlation with significant level has decreased. Furthermore
slight enlargement of the zone with weak correlation values northward apart from the coastal
region up to sub-Sudanian regions have been noted during this period. In 00s the area with
high positive correlation has shrunk with an accentuation of northward expansion of the area
with weak values over Sudanian region (Fig. 4.19c). In sum the area with high correlation
kept dropping all over the region but precisely over Republic of Guinea, Mali and Senegal
during the three decades. The report is that in general the vegetation growing is controlled by
the rainfall availability over the region. However in the coastal region which is mainly
covered by forest, vegetation is weakly influenced by rainfall. As shown in the previous
sections theses region have significantly greened over the two last decades. This finding is
similar to Richard and Poccard (1998) result who found weak sensitivity of the NDVI to
rainfall over coastal areas, mountain regions and flooded areas they concluded that the
interannual variability of the rainfall does not have a significant effect on the photosynthetic
activity.
109
Figure 4. 19: Spatial correlation between NDVI and rainfall over [a] decade 1980s, [b]
decade 1990s and [c] decade 2000s significance areas at 95% confident level.
110
Our findings are in agreement with those of Yuan et al. (2015) who found that grass
vegetation is most sensitive to the changes in precipitation at about 250 mm. According to
them, the correlation between rainfall and vegetation decreases with the increase in
precipitation when precipitation exceeds 250 mm. It may be due to the limits of grass
vegetation RUE, indicating that rainfall might not be the main constraint factor for grass, and
the increase of NDVI might be related to other factors, such as temperature, radiation and so
on.
111
Figure 4. 20: Scatter plot showing correlation and linear equation between rainfall and NDVI
over 1981-2010 at six different points over Sahel region [a] Niger and [b] Burkina Faso,
Sudanian region [c] Mali and [f] Benin and Guinea region [d]
and [e] Ghana.
112
The correlation between rainfall and NDVI was high in large area of the region mainly over
savannah areas it
Nigeria. The high values are mainly observed in region where the annual rainfall is around
1000 mm. So the vegetation growing depends directly on rainfall. This shows some linear
relationship between rainfall and NDVI over these regions. This finding is shared by
Nicholson et al. (1990) who found a linear relationship between rainfall and NDVI in the
Sahel below a rainfall threshold of about 1000 mm per year. At certain amount of rainfall the
vegetation becomes less sensitive to rainfall. Using the REMO to simulate the rainfall, Paeth
et al. (2005) found that rainfall is associated with large-scale circulation and less sensitive to
the annual cycle of vegetation cover.
At a decadal scale, it is noted that the response of the NDVI to rainfall is strong during dry
decade the case of 80s, however, this decrease for the wet periods like 90s and 00s. This
observation has been also done by Wang et al. (2003) at yearly scale. After a year to year
relationship analysis between rainfall and NDVI they conclude that NDVI responded more
rapidly to precipitation during dry years, and during a year immediately after four
consecutive dry years. By contrast, NDVI responded more slowly to precipitation during wet
year, and during year immediately after a wet year.
113
4.5 Changes in Atmospheric Parameters
Compared to CRU data, RegCM4 shows some wet condition over some areas of Nigeria and
over the southern part of Sierra Leon,
GPCP rainfall
data; the model output shows that the western part of the coastal area is wet over land and sea
(Fig. 4.21c and 4.21d).
114
Figure 4. 21: Rainfall monthly mean biases computed over June-July-August-September
(JJAS) based on CRU a) and b) and GPCP c) and d).
115
Figure 4. 22: Temperature monthly mean biases computed over June-July-August-September
(JJAS) based on CRU a) 2005 and b) 2006.
116
Nikulin et al. (2012) found TRMM drier than GPCP over tropical Africa, this point of view is
shared by Sylla et al. (2013) who found GPCP more consistence with gauge based
observations. But, founding some similarity between GPCP and TRMM Diallo et al. (2014)
used both the two observation data to detect the performance of HadGEM3-RA in monsoon
onset studies ov
temperature is studied based on CRU temperature data. Compare to CRU the model gives hot
biases over Sahel region mainly over Senegal, Mauritania, Mali and Niger and some cold
biases in Guinea region Sierra Leon, Southern part of
and Nigeria (Fig. 4.22a
and 4.22b). Our findings in term of 2 m air temperature biases are similar to those determined
by Diallo et al. (2014) using CRU to validate HadGEM3.
Table 4.4: Brief description of observational datasets and model used to set the simulation of
RegCM4.4
CRU
GPCP
RegCM3
Rainfall and Temperature
Rainfall
Rainfall and Temperature
50 km
50 km
50 km
Land only
Land and ocean
Land and ocean
1901-2006 (rainfall)
1991-2010
1989-2008
Harris et al., 2013
Sylla et al., 2012
117
4.5.1 Upper, middle and lower levels tropospheric winds
The atmospheric Jets have an important role in rainfall spatial distribution over this region.
The three main types were shown in the sub sections 4.3.1, 4.3.2 and 4.2.3. Wind spatial
distribution is shown at 200 hPa, 700 hPa and 850 hPa before changes and after making
changes on vegetation cover. These different levels represent respectively the location of the
Tropical Easterly Jet (TEJ), African Easterly Jet (AEJ) and the monsoon fluxes averaged over
JJAS which corresponds to the core of the rainy season over the sahelian region in 2005 and
2006.
4.5.2 Tropical Easterly Jet
The experimentation has given in general a TEJ speed ranging from 2 to 8 m s-1 over West
Africa for both experiments and the core of the wind is located over the Indian Ocean. So the
comparison between TEJ before (Fig. 4.23a and 4.23c) and after changes in vegetation cover
is not showing strong changes. However, only some slight changes on the wind speed occurs
in region above Nigeria and Cameroon (Fig. 4.23b and 4.23d). The TEJ is said to be one of
the most intense features over equatorial Africa and is one mechanism for the formation of
AEWs. Its average location ranges between 5 -10 N in August and 5 -10 S in January. The
TEJ is also produced from the thermal contrast. Central Africa is in the west exit region of the
Asian branch of the TEJ; this exit region will enhance upper-level divergence and lower-level
convergence, promoting convective activity. Variability in the TEJ is associated to
perturbations in the Tibetan high and so this feature could have a remote impact upon central
African rainfall (Farnsworth et al., 2011).
118
Figure 4. 23: JJAS mean Tropical Easterly Jet at 200 hPa before changes [a] and [c]; and after
119
The TEJ (200 hPa wind) according to Janicot et al. (2008) is the high-level anticyclonic
structure, which is the sign of the Indian and African monsoons induces an easterly wind field
on its southern flank. Configurations with stronger monsoon winds tend to have a stronger
core of the TEJ (Klein et al., 2015). During the AMMA field campaign in 2006 Janicot et al.
(2008) found the same result by studying month per month from May to September. They
went beyond the West African region and located the core speed greater than 20 m s 1 centred
over the Indian Ocean. Recently Klein et al. (2015) using the WRF model and the
Atmospheric convective Model Version 2 (ACM2) spread ranges from 20 m s 1 and for ERAInterim, the maximum winds in the core exceed 20 m s
1
at 200 hPa. And some parallelisms
are found with the monsoon activity in July and August.
4.5.3 African Easterly Jet
The highest speed of AEJ (4 m s-1) is located in the Western part of the region along Senegal,
rence between AEJ before and
after changes
al area boarding Liberia
and Republic of Guinea. The averaged ITD was
4.24a and Fig. 4.24
in 2006.
The intensity of the African Easterly Jet is said to be a result of the communication of the
surface temperature gradient into the lower troposphere (Cook, 1999). According to many
authors (Payne and McGarry, 1977; Chen and Ogura, 1982 and Rowell and Milford, 1993),
its presence has been associated with the occurrence of African wave disturbances and,
arguably, with the modulation or even the instigation of intense, small-scale precipitation
events. However, the jet is thought to be hydrodynamically unstable and African wave
disturbances may be an expression of this instability (Thorncroft and Hoskins, 1994).
120
Figure 4. 24: JJAS mean Africa Easterly Jet at 700 hPa before changes [a] and [c]; and after
121
However, there are opposing views in the literature. Schubert et al. (1991), for example,
suggest that the reversed potential vorticity gradients that mark the region of instability
between the ITD and the African Easterly Jet are due only to the presence of a well-defined
ITD over West Africa. Thorncroft and Rowel (1998) finds that both the jet and the ITD
contribute to the reversal of the potential vorticity gradient and, therefore, the unstable
environment. The result in section 4.3.2 about AEJ is different from Janicot et al. (2006) over
the region. This could be due to the considered level (600 hPa for Janicot et al. (2006) and
700 hPa in our case) attributed to the AEJ. The effects occur mainly above the region where
changes have been made on vegetation cover. Studies have revealed the sensitivity of the AEJ
to the land surface (Sylla et al., 2010). The convection has been linked to African Easterly
Waves (AEWs) which development and maintain is based on the presence of AEJ (Berry and
Thorncroft, 2005; Li et al., 2015). According to Diedhiou et al. (2002) over land the waves in
the ITD are mainly located in the neighbourhood of the jet and result mainly from barotropic
instability of the jet. AEJ is prior to AEWs development and enhanced convection cross West
Africa and Atlantic at later gaps (Alaka and Maloney, 2012). There is a marked increase in
AEW activity in June. For Thorncroft and Hodges (2001) the increased activity over the land
in June is perhaps consistent with the increased solar heating at the surface at this time and
the development of a deep well-mixed boundary layer. They also noted that another notable
This equatorward intense activity of the AEWs was even more pronounced in September.
4.5.4 Monsoon fluxes
The zero-isoline of the meridional wind component shown in red is delimitating the monsoon
fluxes or easterlies and the northern Sahel winds or harmattan this is the location of the ITD
over the continent during the monsoon. The key factor of West African Monsoon rainfall is
122
the position of the ITD and its annual cycle. In addition, the seasonal variation of temperature
is associated with the seasonal variation of the ITD. Thus, temperature is at
maximum/minimum value during the period when ITD is at its lowest/highest mean position
around 5° N/20 22° N (Abatan et al., 2014). Based on experimentation, the ITD was located
(Fig. 4.25a and c)
(Fig. 4.25b and d) over
West Africa. This position of the ITD has been observed through many dataset ERA15,
regional climate model REMO from the Max-Planck Institute for Meteorology in Hamburg,
CRU (Paeth et al., 2005), ERA40 (Mekonnen et al., 2006), NCEP. The moisture is brought
toward the continent during the rainy period by the wind at 925 hPa.
123
Figure 4. 25: JJAS mean Monsoon fluxes at 850 hPa before changes [a] and [c]; and after
124
4.5.5 Zonal Wind, Convection and Wind Velocities
4.26a).
These areas are the zones where the changes in vegetation cover seem to affect. The areas
easterlies. The wind movement is dominated by the westerlies over the region during JJAS.
Strong zonal wind is observed when changes are made in vegetation cover. Thus in 2005 after
changes 1 m s-1 is widely represented (Fig. 4.26b) when in 2006 it is observed the appearance
and persistence of 1.5 m s-1 and 2 m s-1 over the region. Referring to the findings of Diro et
al. (2011) who found consistent wind anomaly at 850 mb during deficit rainfall years, our
results is prelude to a probable drop of rainfall due to changes in vegetation. Consistent low
level wind deals with deficit rainfall.
125
Figure 4. 26: Time latitudinal variability of JJAS zonal wind mean at 850 hPa averaged along
vegetation cover (c and d) in 2005 and 2006.
126
Figures 4.27 and 4.28 show the vertical cross-section of JJAS mean wind velocity averaged
-
respectively for 2005 and 2006. Negative (positive) values of the velocity
correspond to upward (downward) motion (Fontaine et al., 2002). According to the wind
vertical speed different phases are noted. The negative wind speed shows convection areas.
Base on the negative velocity location and height, generally three convection levels are
observed
between 700 hPa and 200 hPa. The third
positive wind speed corresponding to subsidence zone in the Sahara
desert is observed. The first convection zone correspond the ITD location in May-June before
its abrupt jump (Sultan and Janicot,
located in the abrupt jump of ITD
corresponding to the monsoon onset over Sahel region.
127
Figure 4. 27: Vertical crossof 2005 showing convergence and divergence zones a) before the change and b) after the
change.
Figure 4. 28: Vertical crossof 2006 showing convergence and divergence zones a) before the change and b) after the
change.
128
When it comes to the difference between winds velocities before change has been made on
vegetation cover (Fig. 4.27a and 4.28a) and after making change on vegetation cover (Fig.
4.27b and Fig. 4.28b), it is not too much perceptible. However, the weakest convection values
are always high before the change in vegetation cover.
The ensemble plots in Figure 4.29 show the time latitude diagram of wind velocity at 925
hPa. The progression of the velocity on time over the latitude
positive and
negative vertical wind speed as the previous. The negative wind represents the ascension
zones over the region. But the convection time progression shows continuous ascension
before changes in vegetation cover when after changes the convection is discontinuous in
time in 2005 (Fig. 4.29a and Fig. 4.29b). And in 2006 the convection is the stronger before
the change (Fig. 4.29c and Fig. 4.29d). The change in convection occurs at area where the
vegetation has been changed. Strong motion ascension occurs at the end of June till the end
of August before the change and after the change the strong ascension occurs late in middle
of July and stay only a month. These mean that the change has affected the convection on
time. The abrupt jump could be explained by the strongest of the convection.
129
Figure 4. 29: Time latitudinal variability of JJAS wind velocity mean at 925 hPa averaged
anges in
vegetation cover [c] and [d] in 2005 and 2006.
130
4.5.6 Relative humidity and Convection
Figures 4.30 and 4.31 show the vertical cross-section of relative humidity and wind velocity
-
Fig. 4.30a) and 2006 (Fig. 4.30b) for the
control. Negative (positive) values of the velocity correspond to upward (downward) motion
(Fontaine et al.,
in general. Furthermore the moisture is well distributed by the TEJ above 200 hPa. Other
hPa) and poor moisture content. However, specifically the year 2005 was the stronger in term
of moisture convection it is note
N convection speed about 0.08 m s-1 with 60%
of moisture when in 2006 it was 0.06 and 50% of moisture. The change in vegetation cover
It is observed a
decrease of the moti
wind speed which corresponds to subsidence zone in the Sahara desert is observed. The first
convection zone correspond the ITD location in May-June before its abrupt jump (Sultan and
Janicot,
D
Sahel region.
131
2005 [a] and in JJAS of 2006 [b].
132
When it comes to the difference between winds velocities without change has been made on
vegetation cover (Fig. 4.30a and Fig. 4.31a) and after making change on vegetation cover
(Fig. 4.30b and Fig. 4.31b), it is not too much perceptible. However, the weakest convection
is always high before the change in vegetation cover. In the next the velocity will be observed
in time to see how the change in vegetation cover is affecting it. With the initial land cover,
strong motion ascension occurs at the end of June till the end of August. However, after the
change the ascension occurs late in middle of July and stay only a month. These mean that
the change has affected the convection on time. The abrupt jump could be explained by the
strongest of the convection.
133
velocity (contour) after changes in vegetation cover averaged along
2005 [a] and in JJAS of 2006 [b].
134
4.5.7 Change in Surface Temperature, Evapotranspiration Flux and Albedo
Initially some differences are noted between 2005 and 2006 in term of Evapotranspiration,
surface temperature and albedo averaged over the experiment band previous to the change in
vegetation cover. It is noted that the surface temperature averaged over JJAS was equal to
er
s of evapotranspiration the high
value 3.84 mm day-1 is observed in 2005 when in 2006 it is computed 3.60 mm day-1. No
no change in surface albedo between the years. However after making change in vegetation
cover some differences are released depending on the year, control 2005 and Sens 2005 and
(Table
4.5). The evapotranspiration has been reduced at about 0.52 mm day-1 when change is made
in vegetation cover in 2005 and at about 0.42 mm day-1 in 2006. The change has increased
the albedo at about +0.03 in both 2005 and 2006.
Table 4.5: Changes in Evapotranspiration, Temperature and Albedo due to vegetation cover
change in JJAS of 2005 and 2006 over changed band.
Initial difference
CTL 2006-2005
Evapotranspiration
-0.24 mm/day
Impact of land
Sens-CTL 2005
Sens-CTL 2006
-0.52 mm/day
-0.45 mm/day
surface change
Temperature
0.6
Albedo
0
+0.03
+0.03
135
Therefore, the drop in evapotranspiration is about 0.43 mm day-1 and 0.37 mm day-1 namely
13% and 12.50% respectively in 2005 and 2006 over the area where changes have been made
in vegetation cover. This is equivalent to a loss of about 53 mm and 45 mm during the whole
season JJAS.
4.5.8 Rainfall Seasonal Variability
The distribution of the intraseasonal variability of the rainfa
Guinea
region is studied in this section. It is based on rainfall daily and monthly means covering
JJAS. Figure 4.32 and Figure 4.33 show daily and monthly time latitude variability of the
rainfall over West Africa in 2005 and 2006 respectively. Northward progression of the ITD is
the end of June in 2005.
Contrary to 2005, in 2006 the rapid jump of the ITD occurs later in July almost at the end of
July (Mounkaila et al., 2014) then, in September starts a southward displacement of the ITD
movement. An important feature for the monsoon and ITD progression is the Saharan heat
low (Janicot et al., 2008).
The different phases of the monsoon within the period JJAS are well captured by the
in Sahel. In 2005 the abrupt shift of the monsoon occurs early at the end of June contrary to
2006 where it occurs in July. The change in vegetation has an impact on monsoon onset it
seems to delay the monsoon onset over Sahel for instance in 2005 where it is observed an
early monsoon onset in Sahel before making changes in surface state. After the changes the
136
time gap between the end of monsoon in Guinea region and its onset in Sahel is considerable
more than one months in 2005 and around one month in 2006. In this changes are not clearly
seen in the daily data, there are clearly observed in monthly plots. At the end of September
there is a southwards trend of the rainfall displacement this is the beginning of the second
season of the rainfall over Guinea region. The change in surface state over Guinea region is
going to affect the rainfall seasonal variability over Sahel region by delaying the onset of
rainfall over there. This finding is in agreement with some previous studies of Charney
(1975). The African Sahel is a transition zone between the arid Sahara Desert and the more
humid Gulf of Guinea registering about 80% of annual rainfall between June and September
during the south-west monsoon season (Monerie et al., 2013).
137
138
The change in land surface cover will modify the surface albedo by increasing it. This finding
is sustained by numerous previous studies realised over Sahel (Charney, 1975). Fuller and
Ottke (2002) at the end of their study on land cover, rainfall and land surface albedo in West
Africa they conclude that albedo and rainfall are related only modestly at short time scales
(monthly and annual). This could explain the observed changes in the monsoon onset over
Sahel.
139
140
The spatial distribution of the rainfall biases computed between the two surface states (before
and after changes) is shown in Figure 4.34. In 2005 dry biases are spread out all over the
Fig. 4.34a).
The phenomenon is slightly different in 2006 with an accentuation of dry biases in region
N and wet biases over Guinea region (Fig. 4.34b). So in general, strong dry biases
are observed over sub Sudanian
gCM4
used for this study supports previous findings which linked the drought in Sahel to the
changes in land surface state over Guinea region.
Figure 4. 34: JJAS rainfall biases between the experiment without changes in vegetation
cover and the experiment with changes in vegetation cover in 2005 [a] and 2006 [b].
141
So in general, strong dry biases are observed over sub Sudanian band. The changes in
el zones
previous findings which linked the drought in Sahel to the changes in land surface state over
Guinea region (Charney, 1975). As the sub Sudanian area is the most affected in term of
rainfall biases distribution, in Figure 4.35 based on zonal averaged rainfall time series over
Sudano-Guinea
-
-
es
on time. Thus, in 2005 (Fig. 4.35a) after making changes in vegetation cover the rainfall have
dropped about 0.42 mm day-1 and 0.23 mm day-1 in 2006 respectively 5% and 3% of the
rainfall over the region (Fig. 4.35b). These results are similar with the grass scenario applied
by Salih et al. (2013) over Republic of Sudan.
142
Figure 4. 35: Rainfall averaged over sub Sudanian band before changes (CTL) and after
changes (Sens) in vegetation cover in 2005 [a] and in 2006 [b].
143
During the whole monsoon period, the rainfall has dropped over the region probably caused
by the changes in vegetation cover which in turn induced some changes in surface albedo
thus in the energy budget. It is observed that before the monsoon surge there was no
difference between the rainfall before and after changes in vegetation cover. However, the
difference starts with the monsoon onset. So as soon as the rainfall starts to go beyond 8 mm
per day the impact of the changes on vegetation cover start to appear in August with the
decrease in rainfall amount after making changes in vegetation cover. Using a short range of
time the two experimentations clearly depict the impact of vegetation cover change on West
Africa Monsoon thus in rainfall. This finding is consistent with previous studies linking
drought in this area to the changes in vegetation cover over Sudano-Guinea regions of West
Africa. Charney (1975) found about 40% drop in rainfall over Sahel region due to the change
in vegetation cover.
144
Figure 4. 36: Rainfall averaged over sub Sudanian band before changes and after changes in
vegetation cover in 2005 and 2006.
145
The 0.52 mm day-1 and 0.42 mm day-1 decrease in evapotranspiration observed for
respectively 2005 and 2006 seems to be low compared to those found by some authors over
the region. For instance, these values are not as high as those of Abiodun et al. (2008) who
found a uniform decrease in Evapotranspiration about 2 mm day-1 which contributes to
decrease in rainfall about 20%. This difference should have been caused by the assumption
made by them. However, our experimentation area was covering only the transition zone
between Guinea and Sudanian regions which is widely less than their one covering the entire
guinea region. In the other hand, two zones are identified with positive differences these are
Guinea
Land
evapotranspiration (ET) is a key component of the coupling between the land surface and the
atmosphere. This contributes to the decrease in the rainfall observed over Sudano-Guinea and
Sudanian
Also the change has
increased the albedo at about +0.03 in both 2005 and 2006.
146
Chapter 5
CONCLUSION AND RECOMMANDATIONS
5.1 Conclusion
Decadal variability of the rainfall and the vegetation over West Africa is revisited from 1981
to 2012 using CRU, station observation rainfall data and NDVI from NOAA. From decade
80s to 90s, significant return to wet condition was observed over West Africa, this was
sustained during decade 00s except over Central Benin and all the western side of Nigeria
where there were observed decreases in annual rainfall magnitudes. From decades 80s to 90s,
a re-greening of the Central Sahel and Sudano-Sahel regions was also observed. From decade
90s to 00s, this re-greening belt was observed extended to the South and the Coastal areas,
mainly over the Guinea Coast, Sudano-Guinea and Western Sahel regions. Over the Sahel,
observed changes in the rainfall pattern are mainly changes in magnitude during the core of
rainy season (July, August and September) and length of the rainy season starting sooner
during the two last decades. Over the Sudanian region, observed changes in the rainfall
pattern are also in the magnitude during the peak of the rainy season and a shift of the rainy
period (JJAS - JASO), starting and ending later during the two last decades. Over the Guinea
Coast, the changes were observed mainly during the little rainy season which becomes more
intense in magnitude and longer in duration during the last two decades. The NDVI intraannual variability shows generally the same evolution pattern compared to the rainfall, but
during the last two decades, significant NDVI values are found one to two months after the
end of the rainy season over the entire region.
147
Correlations between rainfall and NDVI were significant over the Sahel, Sudan and northern
part of Guinea Coast, but they become weaker in magnitude Guinea Coast from decade 80s to
00s meaning that in wetter conditions, there is no linear relationship between NDVI and
rainfall over this region.
It is quite clear from the result of this study that there is recovery of rainfall over some part of
West African region after the long drought period. The increasing tendency observed in
vegetation greenness is moving from the South to North. Although for the decade 90s the regreening process was mainly below latitude 10 N however, in decade 00s it has significantly
reached
After the severe drought of the 80s, the vegetation has been
significantly re- greened over the Sudanian
increase in vegetation over Sudanian region during this period could be due to the rainfall
recovery which has started in decade 90s and may have induced some vegetation cover
recovery. This is however not the case in Guinea region where there were no changes were
observed during this period. The comparison between 00s and 90s shows the persistence of
significant positive changes in the coastal region
The RegCM4 model has been able to simulate the early onset of monsoon in 2005 and the
late onset in 2006. It also captured the impact of vegetation cover change on rainfall spatial
and temporal distribution over the West African region. The model results has also shown that
the change in vegetation cover during the peak of the rainy season in the Sahel (JJAS) does
not have a clear effect on synoptic dynamic patterns like TEJ, AEJ and monsoon fluxes. Also
the mean position of ITD is not affected. However, what has changed is the time lag between
the start of rains in the Guinea and Sudan savannah, which was observed to be about a month
without vegetation cover and a month and a half with vegetation cover. The reliability of this
study resides in the fact that its findings are coherent with some previous findings on this
topic.
148
The results of the study have shown the impact of deforestation over the West African
savannah zone on rainfall spatial and temporal variability and provided maps of rainfall and
vegetation index variability that can guide decision making for policy makers to prevent
further deforestation and soil degradation.
5.2 Recommendations
Despite the fact that some important results have been found on vegetation evolution over
these last decades over West Africa and the impact of vegetation cover change on rainfall
spatio-temporal variability, some gaps still remain. So as recommendation to improve further
understanding, (1) other parameters such as the Leaf Area Index and vegetation types should
be taken into account in order to characterize the vegetation types. Also, above a given
rainfall threshold the NDVI is no longer sensitive to rainfall in wet regions. (2) Modeling
aspect should be done over a long period of time in order to achieve more accuracy. (3)
Future studies could involve the use of different models to better understand the model
strengths and limitations. (4) The policy makers should really care about the destruction of
the forest cover over the region because of its impact on rainfall spatio temporal distribution.
5.3 Limitations of the Study
Our research has some limitations based on the fact that the vegetation over Guinea Coast is
heterogeneous and the vegetation over Sudan is more homogenous. The link between rainfall
and vegetation is still complex to be clearly defined. So it could be better to add more
parameters such the Leaf Area Index, the Photosynthetically Active Radiation Fraction
(FPAR) and the soil types. The vegetation type is not taken into account. The unavailability
of climate data from stations is real issue over the region. The simulation period covered only
three years, this should be extended to more years. Also the model resolution was 50 km x 50
km this could be increased by some downscaling methods in order to fit well with the
149
vegetation indices. The simulation facilities were not accessible on time that had an impact
on the length of the simulation.
150
REFERENCES
Abatan, A. A., Abiodun, B.J. and Omotosho, B.J. (2014). On the characteristics of sea breezes
over Nigerian coastal region. Theor Appl Climatol, 116:93 102.
Abiodun, B.J., Pal, J.S., Afiesimama, E.A., Gutowski, W.J. and Adedoyin, A. (2008).
Simulation of West African monsoon using RegCM3 Part II: impacts of deforestation
and desertification. Theor. Appl. Climatol., 93: 245 261.
Adefolalu, D.O. (1986). Rainfall trends in Nigeria. Theor. and Appl. Climat., 37:205-219.
Adedoyin, J.A. (1997). Instability of squall-inducing waves, global sea-surface temperature
anomalies and climate change in tropical North Africa. Meteorol. Atmos. Phys., 62: 7989.
Adefolalu, D.O. (1985). Further aspects of the Sahelian drought as evident from rainfall
regime in Nigeria. Archive of Meteorology, Geophysics and Bioclimatology, 36: 277-295.
Adler, R.F., Gu, G., and Huffman, G.J. (2010). Estimating climatological bias errors for the
GPCP, J. Hydrometeorology, submitted.
Alaka, G.J.J.R. and Maloney, E.D. (2012). The Influence of the MJO on Upstream Precursors
to African Easterly Waves. J. Climate, 25: 3219-3235.
Ali, A., Lebel, T. and Abou, A. (2008).
sahel Sécheresse. Sciences et changement planétaires/Sécheresse, 19 (4):227-35.
Albignat, J.P. and Reed, R.J. (1980). The Origin of African Wave Disturbances during Phase
III of GATE. Mon. Wea. Rev., 108, 1827 1839.
Anyamba, A. and Tucker, C.J. (2005). Analysis of Sahelian vegetation dynamics using
NOAA-AVHRR NDVI data from 1981 2003. Journal of Arid Environments, 63:596 614.
Bagayoko, F. (2006). Impact of land use intensity on evaporation and surface runoff:
Processes and parameters for eastern Burkina Faso, West Africa. Cuvillier Verlag, 2006,
Vol. 40, Ecology and Development Series, West Africa, Bonn, Germany.
Balas, N., Nicholson, S.E. and Klotter, D. (2007). The relationship of rainfall variability in
West Central Africa to sea-surface temperature fluctuations. Int. J. Climatol., 27: 1335
1349. DOI: 10.1002/joc.1456.
(2012). Effects of coastal topography on climate: high-resolution simulation with a
regional climate model. Clim. Res., Vol. 52: 159 174.
151
Barry, R.G. and Chambers, R.E. (1966). A preliminary map of summer albedo over England
and Wales. Q. J. Roy. Meteor. Soc., 92: 543-548.
Barry, R.G. and Chorley, R.J. (2003). Atmosphere, Weather and Climate. Routledge, (eds) 8.
Bates, B.C., Kundzewicz, Z.W., Wu, S. and Palutikof, J.P. (2008) Climate Change and
Water. (Eds) Technical Paper of the Intergovernmental Panel on Climate Change, IPCC
Secretariat, Geneva, 210p.
Bertrand, A. (1983). La déforestation en zone de forêt en Côte d'Ivoire. Rev. Bois et Forêt des
Tropiques 202:3-17.
Berry, G.J., and Thorncroft, C.D. (2005). Case study of an intense African easterly wave.
Mon. Wea. Rev., 133: 752 766.
Bliefernicht, J., Hamann, I., Kunstmann, H., Legutke, S., Laux, P., Wagner, S., Rhmmler, T.,
Omotosho, J., Gochis, D., Sarr, A., Konaré, A., Lamptey, B., Annor, T., Waongo, M.,
Hingerl, L. and further members of the WASCAL team (2012). DKRZ Project 726 Regional Climate Simulations and Land-Atmosphere Simulations for West Africa,
Poster.
Boateng, P.K. (2013). Agricultural Production, Land-use/cover Change and the
Desertification Debate in the West African Savannah: An Adapted Political Ecology
Approach. Journal of Arts and Humanities, Vol. 2, No. 8: 21-35.
Boko, M., Niang, I., Nyong, A., Vogel, C., Githeko, A., Medany, M., Osman-Elasha, B.,
Tabo, R. and Yanda, P. (2007). Africa Climate Change 2007: Impacts, Adaptation and
Vulnerability, Contribution of Working Group II to the Fourth Assessment Report of the
Intergovernmental Panel on Climate Change, Parry, M.L., Canziani, O.F., Palutikof, J. P,
van der Linden, P.J. and Hanson, C.E. (eds) Cambridge University Press. Cambridge
UK. pp433 467.
Bonan, G.B. (2008). Ecological Climatology, Cambridge Univ. Press, Cambridge, (eds). 2.
Browning, K.A. and Roberts, N.M. (1994). Use of satellite leading to frontal imagery to
diagnose events thunderstorms: part I of a case study, Met Apps 1: 303-310.
Browning, K.A. and Roberts, N.M. (1995). Use of satellite leading to frontal imagery to
diagnose events thunderstorms: part II of a case study, Met Apps 2: 3-9.
Burpee, R.W. (1972). The origin and structure of easterly waves in the lower troposphere of
North Africa. J. Atmos. Sci., 29: 716-726.
Camilla, T. (2009). Climate change in Africa (African Arguments). Zed Books Ltd. 15-31, 5069: 145-152.
152
Carter, T.R., Parry, M.L., Harasawa, H. and Nishioka, S. (1994). Intergovernmental panel on
climate change technical guidelines for assessing climate change impacts and adaptation.
University College London and Center for Global Environmental Research, London.
Charney, J.G. (1975). Dynamics of deserts and drought in the Sahel. Q. J. Roy. Meteorol. Soc.
101:193 202.
Che, M., Chen, B., Innes, J.L., Wang, G., Dou, X., Zhou, T., Zhang,, H., Yan, J., Xu, G. and
Zhao, H. (2014). Spatial and temporal variations in the end date of the vegetation growing
season throughout the Qinghai Tibetan Plateau from 1982 to 2011. Agricultural and
Forest Meteorology, 189 190, 81 90.
Chen, Y.-L. and Ogura, Y. (1982). Modulation of convective activity by large-scale flow
patterns observed in GATE, J. Atmos. Sci., 39: 1260 1279.
Chen, X., Yang, D., Chen, J. and Cao, X. (2015). An improved automated land cover
updating approach by integrating with downscaled NDVI time series data. Remote Sensing
Letters, 6:1, 29-38.
Cook, K.H. (1999). Generation of the African Easterly Jet and Its Role in Determining West
African Precipitation. J. Climate, 12: 1164-1184.
Cracknell, A.P. (2001). The exciting and totally unanticipated success of the AVHRR in
applications for which it was never intended. Advances in Space Research, 28: 233 240.
Dai, A., Lamb, P.J., Trenberth, K.E., Hulme, M., Jones, P.D. and Xie, P. (2004). Comment
the recent Sahel drought is real. Int. J. Climatol., 24: 1323 1331.
Dai, A. (2011). Characteristics and trends in various forms of the Palmer Drought Severity
Index during 1900 2008. J. of Geophysical Research, Vol. 116, D12115, doi:
10.1029/2010JD015541.
Dardel, C., Kergoat, L., Mougin, E., Grippa, M. and Tucker, C.J. (2014). Re-greening Sahel:
30 years of remote sensing data and field observations (Mali, Niger). Remote Sensing of
Environement, 140: 350-364.
Diallo, I., Bain, C.L., Gaye, A.T., Moufouma-Okia, W., Niang, C., Dieng, D.B. M. and
Graham, R. (2014). Simulation of West Arican monsoon onset using the HadGEM3-RA
regional climate model. Clim. Dyn. (accepted).
Diawara, A., Yoroba, F., Kouadio, K.Y., Kouassi, K.B., Assamoi, E.M., Diedhiou, A., and
Assamoi, P. (2014). Climate Variability in the Sudano-Guinea Transition Area and Its
Impact on Vegetation: The Case of th
Advances in
Meteorology, Article ID 831414, 11p.
153
Dickinson, R.E., Henderson-Sellers, A., Kennedy, P.J. and Wilson, M.F. (1986). Biosphere
Atmosphere Transfer Scheme (BATS) for the NCAR Community Climate Model. NCAR
Technical Note, NCAR, TN275+STR, 69p.
Dickinson, R.E., Errico, R.M., Giorgi, F. and Bates, G.T. (1989). A regional climate model
for the Western United States. Clim Change,
.
Dickinson, R.E., Henderson-Sellers, A., and Kennedy, P.J. (1993). Biosphere-Atmosphere
Transfer Scheme (BATS) version 1E as coupled to the NCAR Community Climate
Model. NCAR Tech. rep. TN-387+STR, 72p.
Dickinson, R.E. (2003). Land Atmosphere interactions, Georgia Institute of Technology,
Atlanta, GA, USA, Elsevier Science Ltd.
Diedhiou, A., Janicot, S., Viltard, A. and P. de Felice, (2002). Energetics of easterly wave
disturbances over West Africa and the tropical Atlantic: a Climatology from 1979-95
NCEP/NCAR reanalyses. Clim. Dyn., 18: 487-500.
Diedhiou, A., Janicot, S., Viltard, A., P. de Felice and Laurent, H. (1999). Easterly wave
regimes and associated convection over West Africa and tropical Atlantic: results from
the NCEP/NCAR and ECMWF reanalyses. Clim. Dyn., 15: 795-822.
Dietz, A.J., Ruben, R. and Verhagen, A. (2004). The Impact of Climate Change on Drylands:
With a focus on West Africa. (eds) Environment & policy, ISSN 1383-5130, vol. 39,
465p.
Diro, G.T., Grimes, D.I.F. and Black, E. (2011). Large scale features affecting Ethiopian
rainfall, in Williams, C.J.R. and Kniveton, D.R. (eds) African Climate and Climate
Change. Adv. Global Change Res., vol. 43: 13 50.
Eischeid, J.K., Diaz, H.F., Bradley, R.S., and Jones, P.D. (1991). A comprehensive
precipitation dataset for global land areas. DOE/ER-6901T-H1, U.S. Department of
Energy, Washington, DC, [Available from National Technical Information Service, U.S.
Dept. of Commerce, Springfield, VA 22161] 81p.
Eklundh, L. and Olsson, L. (2003). Vegetation index trends for the African Sahel 1982 1999,
Geophysical Research Letters, Vol. 30, NO. 8, 1430, doi: 10.1029/ 2002GL016772.
Eltahir, E.A.B. and Gong, C. (1996). Dynamics of wet and dry years in West Africa, J.
Climate, 9(5): 1030 1042.
Evan, A. T., Flamant, C., Lavaysse, C., Kocha, C. and Saci, A. (2015). Water Vapor Forced
Greenhouse Warming over the Sahara Desert and the Recent Recovery from the Sahelian
Drought, J. Climate, 28: 108-123.
154
FAO (2000). In: The Challenges of Sustainable Forestry Development in Africa: Twenty-First
FAO Regional Conference for Africa, Yaounde, Cameroon, February 21 25. Food and
Agriculture Organization of the United Nations, Rome, Italy.
Farnswort, A., White, E., Williams, C.J.R., Black, E. and Kniveton, R.D. (2011).
Understanding the Large Scale Driving Mechanisms of Rainfall Variability over Central
Africa. Advances in Global Change Research, 43: 101-122.
Fensholt, R. and Rasmussen, K. (2011). Analysis of trends in the Sahelian 'rain use
efficiency' using GIMMS NDVI, RFE and GPCP rainfall data. Remote Sensing of
Environment, 115: 438 451.
Fontaine, B., Philippon, N., Trzaska, S., and Roucou, P. (2002). Spring to summer changes in
the West African monsoon through NCEP/NCAR reanalyses (1968 1998). Journal of
Geophysical research, Vol. 107, NO. 0,10.1029/2001JD000834.
Fritsch, J.M. and Chappell, C.F. (1980). Numerical Prediction of Convectively Driven
Mesoscale Pressure Systems, Part I: Convective Parameterization. J. Atmos. Sci., 37,
1722 1733.
Fuller, S.P. and Rouse, W.R. (1979). Spectral reflectance changes accompanying a post-fire
recovery sequence in a subarctic spruce lichen woodland. Remote Sensing of
Environment, 8: 11-23.
Fuller, D.O. and Ottke, C. (2002). Land cover, rainfall and land-surface albedo in West
Africa. Climatic Change, 54: 181 204.
Giannini, A., Saravanan, R. And Chang, P. (2003). Oceanic forcing of Sahel rainfall on
interannual to interdecadal time scales. Science, 302: 1027-1030.
Giorgi, F. (1990). Simulation of regional climate using a limited area model nested in a
general circulation model. J. Climate,
.
Giorgi, F., Bates, G.T. and Nieman, S. (1993). The multi-year surface climatology of a
regional atmospheric model over the Western United States. J. Climate,
.
Giorgi, F. and Mearns, L.O. (1999). Introduction to special section: regional climate
modeling revisited. J. Geophys Res., 104:
.
Giorgi, F., Pal, J.S., Bi, X. Q., Sloan, L.C., Elguindi, N. and Solmon, F. (2006). Introduction
to the TAC special issue: the RegCNET network. Theor Appl Climatol,
.
Giorgi, F., Coppola, E., Solmon, F., Mariotti, L., Sylla, M.B, Bi, X., Elguindi, N., Diro, G.T.,
Nair, V., Giuliani, G., Turuncoglu, U.U., Cozzini, S., Güttler, I.,
, T.A., Tawfik, A.B.,
Shalaby, A., Zakey, A.S., Steiner, A.L., Stordal, F., Sloan, L.C. and Brankovic, C. (2012).
155
RegCM4: model description and preliminary tests over multiple CORDEX domains, 52:
7 29.
Giorgi, F., Nellie, E., Stefano, C. and Graziano, G. (2014). Regional Climatic Model RegCM
Gosset, M., Zahiri, E.P. and Moumouni, S. (2010). Rain drop size distribution variability and
impact on X-band polarimetric radar retrieval: Results from the AMMA campaign in
Benin. Q.J.R. Meteorol. Soc., 136(s1): 243 256.
Gray, T.I. and McCrary, D.G. (1981). Meteorological satellite data a tool to describe the
-NI-04042, Johnson Space
Center, Houston, Texas, 7p.
Grell, G., Dudhia, J. and Stauffer, D.R. (1994). A description of the fifth generation Penn
State/NCAR Mesoscale Model (MM5) National Center for Atmospheric Research Tech
Note NCAR/TN-398 + STR, NCAR, Boulder, CO, 121p.
Grist, J.P. and Nicholson, S.E. (2001). A study of the dynamic factors influencing the
variability of rainfall in the West African Sahel, J. Climate, 14: 1337 1359.
Hagos, S.M. and Cook, K.H. (2007). Dynamics of the West African Monsoon Jump, J.
Climate, 20: 5264-5284.
Hamatan, M., Mahe, G., Servat, E., Paturel, J.E. and Amani, A. (2004). Synthèse et
, Note de recherche,
évaluation des prévisions saisonnières e
Sécheresse, 15 (3) : 279-86.
Harris, I., Jones, P.D., Osborn, T.J. and Lister, D.H. (2014). Updated high-resolution grids of
monthly climatic observations the CRU TS3.10 Dataset. Int. J. Climatol., 34: 623 642.
Henke, D., Smyth, P., Haffke, C. and Magnusdottir, G. (2012). Automated analysis of the
temporal behavior of the double Intertropical Convergence Zone over the east Pacific,
Remote Sensing of Environment, 123: 418 433.
Hernandez, M., Miller, S.N., Goodrich, D.C., Goff, B.F., Kepner, W.G., Edmonds, CM. and
Jones, K.B. (2000). Modeling runoff response to land cover and rainfall spatial variability
in semi-arid watersheds, Environmental Monitoring and Assessment, 64: 285 298.
Herrmann, S.M., Anyamba, A. and Tucker, C.J. (2005). Recent trends in vegetation dynamics
in the African Sahel and their relationship to climate, Global Environmental Change,
15:394 404.
Holben, B.N. (1986). Characteristics of maximum-value composite images from temporal
AVHRR data. Int. J. Remote Sensing, 7: 1417 1434.
156
Hua, T., Wang, X. M., Zhang, C. X. and Lang, L. (2013). Temporal and spatial variations in
the Palmer Drought Severity Index over the past four centuries in arid, semiarid, and semi
humid East Asia, Chinese science Bulletin, Vol.58 No.33: 4143-4152.
Hulme, M. (1994). Global changes in precipitation in the instrumental period. In Desbois, M.
and Desalmand, F. (eds) Global Precipitation and Climate Change, Springer-Verlag, 387
405.
Hulme, M. and Kelly, P.M. (1997).
links between desertification and climate
Chapter 15, In Owen, L. and Unwin, T.B.H. (eds.) Environmental management:
readings and case studies, Blackwell, Oxford, UK, 448p.
Imo, J.E. and Ekpenyong, N. (2011). Extreme Climatic Variability in North-western Nigeria:
An Analysis of Rainfall Trends and Patterns, Journal of Geography and Geology, 3: 5162.
IPCC (2007). Climate change 2007. The Physical Science Basis, Contribution of working
Group I to the fourth assessment Report of the Intergovernmental Panel on Climate
Change, edited by S. Solomon, J.M., Dyurgerov, and R.S.W. Van de Wal (2007)
Cambridge Univ. Press, Cambridge, U. K., 996 pp.
IPCC (2013). Climate change physical basis, working group I contribution to the fifth
assessment report of the intergovernmental panel on climate change. Cambridge
University Press, Cambridge, United Kingdom and New York, NY, USA, 1535pp.
Jamali, S., Jönsson, P., Eklundh, L., Ardö, J. and Seaquist, J. (2015). Detecting changes in
vegetation trends using time series segmentation. Remote Sensing of Environment, 156:
182 195.
James, R.H. and Gregory, J.H. (2012). An Introduction to Dynamic Meteorology. Academic
Press, vol.88 of International Geophysics (eds) 5, 552pp.
Janicot, S., Thorncroft, C.D., Ali, A., Asencio, N., Berry, G., Bock, O., Bourles, B., Caniaux,
G., Chauvin, F., Deme, A., Kergoat, L., Lafore, J.-P., Lavaysse, C., Lebel, T.,
Marticorena, B., Mounier, F., Nedelec, P., Redelsperger, J.-L., Ravegnani, F., Reeves, C.
E., Roca, R., P. de Rosnay, Schlager, H., Sultan, B., Tomasini, M., Ulanovsky, A. and
ACMAD forecasters team (2008). Large-scale overview of the summer monsoon over
West Africa during the AMMA field experiment in 2006, Ann. Geophys., 26, 2569 2595.
Jones, P.D., New, M., Parker, D.E., Martin, S. and Rigor, I.G. (1999). Surface air temperature
and its variations over the last 150 years, Reviews of Geophysics, 37: 173-199.
Kim, D.H., Sexton, J.O., Noojipady, P., Huang, C., Anand, A., Channan, S., Feng, M. and
Townshend, J.R. (2014). Global, Landsat-based forest-cover change from 1990 to 2000.
Remote sensing of Environment, 155: 178-193.
157
Kinne, S., Schulz, M., Textor, C., Guibert, S., Balkanski, Y., Bauer, S.E., Berntsen, T.,
Berglen, T.F., Boucher, O., Chi, M., Collins, W., Dentener, F., Diehl T., Easter R.,
Feichter, J., Fillmore, D., Ghan, S., Ginoux, P., Gong, S., Grini, A., Hendricks, J., Herzog,
M., Horowitz, L., Isaksen, I., Iversen, T., Kirkevag, A., Kloster, S., Koch, D.,
Kristjansson, J. E., Krol, M., Lauer, A., Lamarque, J.F., Lesins, G., Li, X., Lohmann, U.,
Montanaro, V., Myhre, G., Penner, J.E., Pitari, G., Reddy, S., Seland, O., Stier, P.,
Takemura, T., and Tie, X. (2006). An AeroCom initial assessment optical properties in
aerosol component modules of global models. Atmos. Chem. Phys., 6: 1815 1834.
Kinne, S. (2009). Climatologies of Cloudrelated Aerosols Part 1: Particle Number and Size
From the Strüngmann Forum Report. In, Jost Heintzenberg and Robert J. Charlson (eds)
Clouds in the Perturbed Climate System: Their Relationship to Energy Balance,
Atmospheric Dynamics, and Precipitation, MIT Press ISBN 978-0-262-01287-4.
Klein, C., Heinzeller, D., Bliefernicht, J. and Kunstmann, H. (2015). Variability of West
African monsoon patterns generated by a WRF multi-physics ensemble, Clim. Dyn., 23pp,
DOI 10.1007/s00382-015-2505-5.
Konaté, S. and Kampmann, D. (2012). Biodiversity Atlas of West Africa, (eds) vol. 3: Cote
, Abidjan and Frankfurt/Main, pp: 44-47.
Kpongor, D. (2007). Spatially explicit modeling of sorghum (Sorghum bicolor L) production
on complex terrain of a semi-arid region of Ghana using APSIM. PhD thesis, Ecology
and Development Series 51. Center for Development Research (ZEF), University of
Bonn. Bonn, Germany, 132pp.
Kriebel, K.T. (1979). Albedo of vegetated surfaces: Its variability with differing irradiances.
Remote Sensing of the Environment, 8: 283-290.
Kukla, G.J. and Robinson, D. (1980). Annual cycle of surface albedo. Mon, Wea Rev., 108:
56-67.
Lamb, P.J. (1980). Sahelian Drought, New Zealand, Journal of Geography, 68:12 16.
Laux, P., Kunstmann, H. and Bárdossy, A. (2007).
with atmospheric circulation patterns, Proceedings of Symposium HS2004 at IUGG2007,
Perugia, IAHS Publ., 313pp.
Le Barbe, L., Lebel, T. and Tapsoba, D. (2002). Rainfall Variability in West Africa during the
Years 1950 90, J. Climate, 15:187-202.
Lebel, T., Diedhiou, A. and Laurent, H. (2003). Seasonal cycle and interannual variability of
the Sahelian rainfall at hydrological scales, Journal of Geophysical Research, 108: 8389
8392.
158
Lebel, T. and Ali, A. (2009). Recent trends in the Central and Western Sahel rainfall regime
(1990-2007). Journal of Hydrology, 375 (1 2): 52 64.
Lebel, T., Parker, D.J., Flamant, C., Bourles, B., Marticorena, B., Mougin, E., Peugeot, C.,
Diedhiou, A., Haywood, J. M., Ngamini, J.B., Polcher, J., Redelsperger and Thorncroft C.
D. (2010). The AMMA field campaigns: Multiscale and multidisciplinary observations in
the West African region. Q. J. R. Meteorol. Soc., 136(s1): 8 33.
Li, R., Jin, J., Wang, S. Y. and Gillies, R.R. (2015). Significant impacts of radiation physics
in the Weather Research and Forecasting model on the precipitation and dynamics of the
West African Monsoon, Clim. Dyn., 44:1583 1594.
Lim, C. and Kafatos, M. (2002). Frequency analysis of natural vegetation distribution using
NDVI/AVHRR data from 1981 to 2000 for North America: Correlations with SOI. Int. J.
Remote Sensing, 23:17, 3347-3383.
Louvet, S., Paturel, J.E., Mahé, G., Rouché, N. and Koité, M. (2015). Comparison of the
spatiotemporal variability of rainfall from four different interpolation methods and impact
on the result of GR2M hydrological modeling case of Bani River in Mali, West Africa.
Theor. Appl. Climatol, DOI 10.1007/s00704-014-1357-y.
Mangiarotti, S., Mazzega, P., Hiernaux, P. and Mougin, E. (2012). Predictability of
vegetation cycles over the semi-arid region of Gourma (Mali) from forecasts of AVHRRNDVI signals. Remote Sensing of Environment, Elsevier, 123: 246-257.
Marx, A., Kunstmann, H., Schuttemeyer, D. and Moene, A.F. (2008). Uncertainty analysis
for satellite derived sensible heat fluxes and scintillometer measurements over Savannah
environment and comparison to mesoscale meteorological simulation results. Agricultural
and forest meteorology, 148: 656 667.
Mathon, V. and Laurent, H. (2001). Life cycle of Sahelian mesoscale convective cloud
systems. Q. J. R. Meteorol. Soc., 127: 377 406.
Matthews, E. (1983). Global vegetation and land use: New high-resolution data bases for
climate studies, J. Clim. Appl. Meteorol., 22: 474-487.
McGuffie, K. (1994). Remote Sensing Input to Climatological Datasets. In, Robin, A.
Vaughan, and Arthur, P. Cracknell (eds) Remote Sensing and Global Climate Change,
NATO ASI Series. Vol. I, 24pp.
McKee, T. B., Doesken, N. J. and Kleist, J. (1993). The relationship of drought frequency and
duration to time scales. Preprints, Eighth Conf. on Applied Climatology, Anaheim, CA,
Amer. Meteor. Soc., 179 184.
McKee, T. B., Doesken, N. J. and Kleist, J. (1995). Drought monitoring with multiple time
scales. Proc. Ninth Conf. on Applied Climatology, Dallas, TX, Amer. Meteor. Soc., 233
236.
159
Mekonnen, A., Thorncroft, C.D. and Hall, N.M.J. (2006). Analysis of convection and its
association with African easterly waves, J. Climate, 19: 5405-5421.
Mito, C.O., Boiyo, R.K. and Laneve, G. (2012). A simple algorithm to estimate sensible heat
flux from remotely sensed MODIS data. Int. J. Remote Sensing, 33:19, 6109-6121.
Monerie, P.-A., Roucou, P. and Fontaine, B. (2013). Mid-century effects of Climate Change
on African monsoon dynamics using the A1B emission scenario. Int. J. Climatol, 33:
881 896.
Monteny, B.R. (1986). Importance of the tropical rain forest as an atmospheric moisture
source, Proc. ISLSCP Conference, Rome, Italy 2-6 December 1985, ESA SP-248:449-454.
Mosley, P.M. and McKerchar, A.I. (1993). Streamflow. In Maidment D.R. (eds) Handbook of
Hydrology (p. 8.1), McGraw-Hill, Inc.
Mounkaila, M.S., Abiodun, B.J. and Omotosho, J.B. (2014). Assessing the capability of
CORDEX models in simulating onset of rainfall in West Africa. Theor. Appl. Climatol.,
18pp.
Myneni, R.B. and Williams, D.L. (1994). On the relationships between FAPAR and NDVI.
Remote Sensing of Environment, 49: 200 211.
Myneni, R.B., Hall, F.G., Sellers, P.J. and Marshak, A.L. (1995). The interpretation of
spectral vegetation indexes. IEEE Transactions on Geoscience and Remote Sensing, 33:
481 486.
Monteith, J.L. (1959). The reflection of short-wave radiation by vegetation. Q. J. R. Meteor.
Soc., 85: 386-392.
New, M., Hulme, M. And Jones, P. (1999). Representing Twentieth-Century Space Time
Climate Variability, Part I: Development of a 1961 90 Mean Monthly Terrestrial
Climatology. J. Climate, 12: 829-856.
New, M., Hulme, M. and Jones, P. (2000). Representing Twentieth-Century Space Time
Climate Variability. Part II: Development of 1901 96 Monthly Grids of Terrestrial Surface
Climate. J. Climate, 13: 2217-2238.
Nicholson, S.E. (1993). Sahel, West Africa, Encyclopedia of Environmental biology. Vol. 3,
Academic Press, Inc.
Nicholson, S.E., Devenport, M.L. and Malo, R. (1990). A comparison of vegetation response
to rainfall in Sahel and East Africa using Normalised Difference vegetation index from
NOAA AVHRR. Climatic Change, 17: 209-241.
160
Nicholson, S.E. (1993). An overview of African rainfall fluctuations of the last decade. J.
climate, 6:1463-1466.
Nicholson, S.E., Tucker, C.J. and Ba, M.B. (1998). Desertification, drought, and surface
vegetation: an example from the West African Sahel. Bulletin of the American
Meteorological Society, 79: 815 829.
Nicholson, S.E. (2005). On the question of the recovery of the rains in the West Africa Sahel.
Journal of Arid Environments, 63: 615-641.
Nicholson, S.E. (2009).
over West Africa. Clim. Dyn., 32:1155 1171.
Nikulin, G., Jones, C., Giorgi, F., Asrar, G., Büchner, M., Cerezo-Mota, R., Bøssing, O.
Christensen, Déqué, M., Fernandez, J., Hänsler, A., Meijgaard, E., Samuelsson, P., Sylla,
M.B., and Sushama, L.I. (2012). Precipitation climatology in an ensemble of CORDEX
Africa regional climate simulations. J. Climate, 25:6057 6078.
Norman, A.P. (1956). The general circulation of the atmosphere: A numerical experiment. Q.
J. R. Meteorological Soc., 82: 123-164.
Odekunle, T.O. and Eludoyin, A.O. (2008). Sea surface temperature patterns in the Gulf of
Guinea: their implications for the spatio-temporal variability of precipitation in West
Africa. Int. J. Climatol., 28: 1507 1517.
Odekunle, T.O. (2010). An Assessment of the Influence of the Inter-Tropical Discontinuity
on Inter-Annual Rainfall Characteristics in Nigeria. Geographical Research, 48(3):314
326.
Ogungbenro, S.B., Ajayi, V.O. and Adefolalu, D.O. (2015). Mean state and kinematic
properties of mesoscale convective systems over West Africa. Theor Appl Climatol, 9p.
Oguntoyinbo, J.S. (1970). Reflection coefficient of natural vegetation, crops and urban
surfaces in Nigeria, Q. J. Roy. Meteor. Soc., 96: 430-441.
Olson, J.S., Watts, J.A. and Allison, L.J. (1983). Carbon in live vegetation of major world
ecosystems. U.S. Dep. Energy, DOE/NBB-0037, No. TROO4, 164p.
Omotosho, J.B. (1990). Onset of thunderstorms and precipitation over Northern Nigeria. Int.
J. Climatol., 10:849-860.
Omotosho, J.B. (1992). long-range prediction of the onset and end of the rainy season in the
west african sahel. Int. J. Climatol., vol. 12: 369-382.
161
Omotosho, J.B., Balogun, A.A. and Ogunjobi, K. (2000). Predicting monthly and seasonal
rainfall, onset and Cessation of the rainy season in West Africa using only surface data.
Int. J. Climatol., 20: 865 880.
Omotosho, J.B. (2008). Pre-rainy season moisture build-up and storm precipitation delivery
in the West African Sahel. Int. J. Climatol., 28:937 946.
Ouédraogo, E. (2004). Soil Quality Improvement for Crop Production in semi-arid West
Africa. Wageningen, the Neederlands: Wageningen University, Tropical Resource
Management 51p.
Oueslati, B. and Bellon, G. (2015). The double ITCZ bias in CMIP5 models: interaction
between SST, large scale circulation and precipitation. Clim. Dyn., 44:585 607.
Paeth, H., Born, K., Podzun, R. and Jacob, D. (2005). Regional dynamical downscaling over
West Africa: model evaluation and comparison of wet and dry years. Meteorologische
Zeitschrift, Vol. 14, No. 3: 349-367.
Pal, J.S., Small, E. and Eltahir, E. (2000). Simulation of regional-scale water and energy
budgets: representation of subgrid cloud and precipitation processes within RegCM. J
Geophys Res., 105: 29579 29594.
Pal, J.S., Giorgi, F., Bi, X., Elguindi, N., Solomon, F., Gao, X., Francisco, R., Zakey, A.,
Winter, J., Ashfaq, M., Syed, F., Bell, J.L., Diffanbaugh, N.S., Kamacharya, J., Konare,
A., Martinez, D., da Rocha, R.P., Sloan, L. C. and Steiner, A. (2007). The ICTP RegCM3
and RegCNET: regional climate modeling for the developing world. Bulletin of
American Meteorological Soc., 88: 1395 1409.
Payne, S.W. and McGarry, M.M. (1977). The relationship of satellite inferred convective
activity to easterly waves over west Africa and the adjacent ocean during phase III of
GATE. Mon. Wea. Rev., 105: 413 420.
Pilgrim, D.H. and Cordery, I. (1993). Chapter 9: Flood Runoff. In: Maidment D.R. (eds.)
Handbook of Hydrology, McGraw-Hill, New York, USA.
, F.
Pospichal, B., Karam, D.B., Crewell, S., Flamant, C., Hunerbein, A., Bockd,
(2010). Diurnal cycle of the intertropical discontinuity over West Africa analysed by
remote sensing and mesoscale modelling. Q. J. R. Meteorol. Soc., 136(s1):92 106.
Quevauviller, P. (2011). Water Sustainability and Climate Change in the EU and Global
Context
Policy and Research Responses. Issues in Environmental Science and
Technology, 31, Sustainable Water, 24p.
Rasmussen, K., Fog, B. and Madsen, J.E. (2001). Desertification in reverse? Observations
from northern Burkina Faso, Global Environmental Change, 11:271 282.
162
Rawls, W.J., Brakensiek, D.L. and Shirmohammadi, A. (1993). Infiltration and soil water
movement. In: D.R. Maidment (eds) Handbook of Hydrology McGraw-Hill, New York,
pp5.1-5.51.
Reiche, J., Verbesselt, J., Hoekman, D. and Herold, M. (2015). Fusing Landsat and SAR time
series to detect deforestation in the tropics. Remote Sensing of Environment, 156: 276
293.
Richard, Y. and Poccard, I. (1998). A statistical study of NDVI sensitivity to seasonal and
interannual rainfall variations in Southern Africa. Int. J. of Remote Sensing, 19:15, 29072920.
Rockwood, A.A. and Cox, S.K. (1978). Satellite inferred surface albedo over north-western
Africa. J. Atmos. Sci., 35: 513-522.
Rodwell, M.J. and Hoskins, B.J. (1996). Monsoons and the dynamics of deserts. Q. J. R.
Meteorol. Soc., 122: 1385 1404.
Rowell, D.P. and Milford, J.R (1993). On the generation of African squall lines. J. Climate, 6:
1181 1193.
Salih, A.M., Kornich, H. And Tjernstrom (2013). Climate impact of deforestation over South
Sudan in regional climate model, Int. J. Climatol., 33: 2362-2375.
Sandwidi, J.P. (2007). Groundwater potential to supply population demand within the
Kompienga dam basin in Burkina Faso. PhD thesis. Center for Development Research,
University of Bonn. Ecology and Development Series, 2007. 55p. Bonn, Germany.
Available from: http://www.zef.de/fileadmin/template/Glowa/Downloads/Sandwidi_do
c_thesis_2007.pdf.
Sanni, M., Odekunle, T.O. and Adesina, F.A. (2012). Spatio-Temporal Variation of Drought
Severity in the Sudano-Sahelian Region of Nigeria: Implications for Policies on Water
Management. In: W. Leal Filho (eds), Climate Change and the Sustainable Use of Water
Resources, Climate Change Management, DOI: 10.1007/978-3-642-22266-5_26.
Schulz, M., Textor, C., Kinne, S., Balkanski, Y., Bauer, S., Berntsen, T., Berglen, T.,
Boucher, O., Dentener, F., Guibert, S., Isaksen, I.S.A., Iversen, T., Koch, D., Kirkevåg, A.,
Liu, X., Montanaro, V., Myhre, G., Penner, J. E., Pitari, G., Reddy, S., Seland Ø., Stier, P.,
and Takemura, T. (2006). Radiative forcing by aerosols as derived from the AeroCom
present-day and pre-industrial simulations. Atmos. Chem. Phys., 6: 5225-5246.
Simmons, A., Dee, D., Uppala, S. and Kobayashi, S. (2007). Erainterim: new ecmwf
reanalysis products from 1989 onwards. In: ECMWF Newsl, 110. ECMWF, p
.
Sivakumar, M.V.K. (1992). Climate change and implications for agriculture in Niger.
Climatic change, 20 (4): 297-312.
163
Sivakumar, M.V.K. (2007). Interactions between climate and desertification. Agricultural and
Forest Meteorology, 142:143 155.
Schindler, J. (2009). A multi-agent system for simulating land-use and land-cover change in
the Atankwidi catchment of Upper East Ghana, PhD thesis, Center for Development
Research, University of Bonn. Ecology and Development Series, 68p. Bonn, Germany.
Available from: http://hss.ulb.uni-bonn.de:90/2009/1954/1954.htm.
Schubert, W.H., Ciesielski, P.E., Stevens, D.E. and Kuo, H.C. (1991). Potential vorticity
modeling of the ITCZ and the Hadley circulation. J. Atmos. Sci., 48: 1493 1509.
Stewart, J.B. (1970). The albedo of a pine forest. Q. J. R. Meteor. Soc., 97: 561-564.
Stow, D.A., Shih, H.-C. and Coulter, L.L. (2014). Discrete classification approach to land
cover and land use change identification based on Landsat image time sequences. Remote
Sensing Letters, 5:10, 922-931.
Sultant, B. and Janicot, S. (2000). Abrupt shift of the ITCZ over West Africa and intraseasonal variability. Geophys. Res. Lett., 27: 3353 3356.
Sultan, B., Janicot, S. and Diedhiou, A. (2003). The West African Monsoon Dynamics. Part
I: Documentation of Intraseasonal Variability, J. Climate, 16: 3389-3406.
Sultan, B. and Janicot, S. (2003). The West African Monsoon dynamics, Part II: the
monsoon. J. Climate, 16(21):3407 3427.
Sylla, M.
, A., Ruti, P.M. and Giorgi F. (2010). Simulation of the intraseasonal
and the interannual variability of rainfall over West Africa with RegCM3 during the
monsoon period. Int. J. Climatol, 30: 1865 1883.
Sylla, M.B., Giorgi, F. and Stordal, F. (2012). Large-scale origins of rainfall and temperature
bias in high-resolution simulations over southern Africa. Clim Res., Vol. 52: 193 211.
Sylla, M.B., Diallo, I. and Pal, J.S. (2013). West African Monsoon in State-of-the-Science
Regional Climate Models, Climate Variability-Regional and Thematic Patterns, Dr.
Aondover Tarhule (eds), ISBN: 978-953-51-1187-0, InTech, DOI: 10.5772/55140.
Textor, C., Schulz, M., Guibert, Kinne, S., Balkanski, Y., Bauer, S., Berntsen, T., Berglen, T.,
Boucher, O., Chin, M., Dentener, F., Diehl, T., Easter, R., Feichter, H., Fillmore, D.,
Ghan, S., Ginoux, P., Gong, S., Grini, A., Hendricks, J., Horowitz L., Huang P., Isaksen,
T. Iversen, S. Kloster, D. Koch, A. Kirkevag, J. E. Kristjansson, M. Krol, A. Lauer I.,
Lamarque, J. F., Liu, X., Montanaro, V., Myhre, G., Penner, J., Pitari, G., Reddy, S.,
Seland, Ø., Stier, P., Takemura, T., and Tie, X. (2006). Analysis and quantification of the
diversities of aerosol life cycles within AeroCom. Atmos. Chem. Phys., 6: 1777 1813.
164
Thorncroft, C.D. and Hoskins, B.J. (1994). An idealized study of African easterly waves. I: A
linear view. Q. J. R. Meteorol. Soc., 120: 953-982.
Thorncroft, C.D. and Rowell, D. (1998). Interannual variability of African easterly waves in a
GCM. Int. J. Climatol., 18: 1305-1323.
Thorncroft, C.D. and Blackburn, M. (1999). Maintenance of the African easterly jet. Q. J. R.
Meteorol. Soc., 125: 763-786.
Thorncroft, C.D. and Hodges, K. (2001). African Easterly Wave Variability and Its
Relationship to Atlantic Tropical Cyclone Activity, J. Climate, 14:1166-1178.
Tompkins, A.M., Diongue-Niang, A., Parker, D.J. and Thorncroft, C.D. (2005). The African
easterly jet in the ECMWF Integrated Forecast System: 4D-Var analysis. Q. J. R.
Meteorol. Soc., 131: 2861 2885.
Tucker, C.J. and Miller, L.D. (1977). Soil spectra contributions to grass canopy spectral
reflectance. Photogrammatic Engineering and Remote Sensing, 43: 721-726.
Tucker, C.J. (1979). Red and photographic infrared linear combinations for monitoring
vegetation, Remote Sensing of Environment, 8: 127 150.
Tucker, C.J., Pinzon, J.E., Brown, M.E., Slayback, D.A., Pak, E.W., Mahoney, R., Vermote,
E.F. and Saleous, N.E. (2005). An extended AVHRR 8-km NDVI dataset compatible with
MODIS and SPOT vegetation NDVI data. Int. J. of Remote Sensing, No. 20, 26: 4485
4498.
Uppala, S., Dee, D., Kobayashi, S., Berrisford, P. and Simmons A. (2008). Towards a climate
adapt assimilation system: status update of ERA-Interim. ECMWF Newsletter, 115: 12-18.
Vizy, E.K. and Cook, K.H. (2001). Mechanisms by Which Gulf of Guinea and Eastern North
Atlantic Sea Surface Temperature Anomalies Can Influence African Rainfall. J. Climate,
14: 795-821.
Wang, J., Rich, P.M. and Price, K.P. (2003). Temporal responses of NDVI to precipitation
and temperature in the central Great Plains, USA. Int. J. Remote sensing, Vol. 24, 11:
2345 2364.
White, F. (1983). The vegetation of Africa, a descriptive memoir to accompany the
UNESCO/AETFAT/UNSO vegetation map of Africa. UNESCO, Natural Resour. Res., 20:
1-356.
Williams, C.J.R. and Kniveton, D.R. (2011). African Climate and Climate Change, (eds) 101
Advances in Global Change Research 43, DOI 10.1007/978-90-481-3842-5_5, Springer
Science+Business Media B.V.
165
Wilson, M.F. and Henderson-Sellers, A. (1985). A global archive of land cover and soils data
for use in general circulation climate models. J. Climate, 5: 119-143.
Wu, D., Wu, H., Zhao, X., Zhou T., Tang, B., Zhao, W. and Jia, K. (2014). Evaluation of
Spatiotemporal Variations of Global Fractional Vegetation Cover Based on GIMMS NDVI
Data from 1982 to 2011. Remote Sens., 6: 4217-4239; doi:10.3390/rs6054217.
Yaw, O.-A. and Ian, C. (1994). Atlantic Sea Surface Temperatures and Rainfall Variability in
Ghana, J. Climate, 7: 551-558.
Yilma, T. (2006). Modeling farm irrigation decisions under rainfall risk in the Whit Volta
Basin of Ghana. A tool for policy analysis at the farmhousehold level. Bonn, Göttingen,
Germany: University of Göttingen. 170p.
Yuan, X., Li, L. and Chen, X. (2015). Increased grass NDVI under contrasting trends of
precipitation change over North China during 1982 2011, Remote Sens. Letters, 6:1, 6977.
Yuen, B. and Kumssa, A. (2011). Climate Change and Sustainable Urban 3 Development in
Africa and Asia (eds), DOI 10.1007/978-90-481-9867-2_1, Springer Science+Business
Media B.V.
Zhu, Z., Bi, J., Pan, Y., Ganguly, S., Anav, A., Xu, L., Samanta, A., Piao, S., Nemani, R.R.
and Myneni, R.B. (2013). Global Data Sets of Vegetation Leaf Area Index (LAI) 3g and
Fraction of Photosynthetically Active Radiation (FPAR)3g Derived from Global Inventory
Modeling and Mapping Studies (GIMMS) Normalized Difference Vegetation Index
(NDVI3g) for the Period 1981 to 2011. Remote Sens, 5, 927-948.
Zanis, P., Ntogras, C., Zakey, A., Pytharoulis, I. and Karacostas, T. (2012). Regional climate
feedback of anthropogenic aerosols over Europe using RegCM3. Clim. Res., Vol. 52:
267 278.
Zeng, N. (2003). Drought in the Sahel, Science 3.
Zheng, X. and Eltahir, E.A.B. (1998). The Role of Vegetation in the Dynamics of West
African Monsoons. J. Climate, 11: 2078-2096.
166

land use change, vegetation dynamics and rainfall spatio

Transcription

Documents pareils

- Wiley Online Library

as a PDF

Canada - Kahnawake

Role of the Hoggar massif in the West African monsoon - HAL-Insu

ECOLE DOCTORALE DES SCIENCES DE L

How much does grazing-induced heterogeneity impact

IUFRO PILOT SITE DIMAKO CAMEROON

bugs and beavers - Ville d`Estérel