FACTORING CLIMATE VARIABILITY AND CHANGE INTO CROP MODELS FOR ENHANCING SORGHUM PERFORMANCE IN THE WEST AFRICAN SEMI-ARID TROPICS AKINSEYE, Folorunso Mathew Major Supervisor: Prof. S. O Agele (FUTA-Nigeria) Co-Supervisor: Dr. P. C. S. Traore (ICRISAT- Mali) German Adviser: Prof. Dr. A. M Whitbread(UG& ICRISAT-India) Department of Meteorology and Climate Science Federal University of Technology, Akure, Ondo State. PhD Final Presentation
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
FACTORING CLIMATE VARIABILITY AND CHANGE INTO CROP MODELS FOR ENHANCING SORGHUM
PERFORMANCE IN THE WEST AFRICAN SEMI-ARID TROPICS
AKINSEYE, Folorunso MathewMajor Supervisor: Prof. S. O Agele (FUTA-Nigeria)
Co-Supervisor: Dr. P. C. S. Traore (ICRISAT- Mali)German Adviser: Prof. Dr. A. M Whitbread(UG& ICRISAT-India)
Department of Meteorology and Climate ScienceFederal University of Technology, Akure, Ondo State.
PhD Final Presentation
Why is climate variability so importantto agriculture ?
Agriculture is the largest employer of labour, a guarantee for food
security in the world and is probably the most weather-dependent of all
human activities.
Climate variability has been, and continues to be, the principal source of
fluctuations in global food production, particularly in the semi- arid
tropics.
Throughout history, climatic extremes has wreaked havoc on
agriculture, water resources etc.
In addition, with other physical, social, political and economic factors,
climate variability contribute to vulnerability of economic losses,
hunger, famine and dislocation
IntroductionWest African semi-arid is home to some 300 million people with at least 70%
engaged in agricultural activity (FAO,2007), it accounts for 35% of the GDP,
(World bank, 2000) and ~ 90% of cropland managed under rainfed conditions
(FAOSTAT,2005).
Rainfall is one of the most important natural resources and rainfall variability
manifests intra-annual, inter-annual and decadal scales.
Crucial problem for rainfed agriculture: Decision about the optimal planting date
for current season
- Planting as early as possible to avoid wastage of valuable growth time
- Planting too early /late may lead to crop failures and high economic losses
Low crop yield (productivity) of major cereal crops attributed to constraining
environmental conditions ,depleted soil fertility (Nitrogen and phosphorus),
diseases ( e.g. Midges),high costs of fertilizers (Winterbottom et al., 2013)
Introduction Cont’d• In the semi-arid tropics, sorghum and millet contribute to more
than 80% of the food needs and has mean yield of 800kg/ha
(Maredia et al., 1998, 2000)
• In 2008, sorghum was cultivated in Mali on an area of 990 995
ha with a production of 1, 027,202 tons and yield average is
1036kg/ha(http://faostat.fao.org/site)
• Crop growth models are used around the world as a research
tool for yield forecast because models
– provide dynamical estimates of climate driven potential yield,
and yield components as well as water balance
– useful for assessing the agricultural risks of climate change in
Introduction Cont’d Decision Support for Agro-technology Transfer (DSSAT) (Jones et al., 2003).
Agricultural Productions Systems sIMulator (APSIM) (McCown et al., 1996;Keating et al., 2003).
Samara Version 2 implemented on the Ecotrop platform of the CentreInternational de Recherche Agronomique pour le De´veloppement (CIRAD)Dingkhun et al., (2003)
• DSSAT model was previously used in simulation studies by Adiku et al., 2007;
MacCarthy et al., (2013) over Ghana) and Traore et al., (2007) in the Sahel zone
• APSIM model was also used in previous studies in West Africa by MacCarthy
et al., 2009 and Apkonikpe et al., (2010).
• comparative evaluation of these models has not been undertaken for
sorghum growth and development in West Africa
Crop simulation models integrate the interaction of genotypic traits,
environmental factors (e.g. soils, weather) and management (G x E x M)
Literature Review
• Lobell et al., (2011), the potential yield loss due to the climate change impact is
about 5% for each degree Celsius of global warming.
• IPCC (2014) predicts an approximate 50% decrease in yields from rain-fed
agriculture by 2020 in some countries.
Reference
Climate
model Crop model Scenario Area Horizon Crop Baseline
Adejuwon (2006) HadCM2 EPIC 1%/year in CO2 Nigeria
2035/2055/
2085
Cassava, maize,
millet, rice, sorghum 1960/1990
Jones and
Thornton (2003) HadCM2
CERES maize
(DSSAT) Not found WA (details) 2055 Maize
1990
climate
normals
Liu et al., (2008) HadCM3 GEPIC A1FI, B1, A2, B2
SSA, WA
(details) 2030
Global, cassava,
maize, millet, rice,
sorghum, wheat 1990/1999
Lobell et al.,
(2008) 20 GCMs Empirical A1B, A2, B1 WA 2030
Cassava, groundnut,
maize, millet,
rice, sorghum, wheat,
yams 1998/2002
Parry et al.,
(2004) HadCM3
Empirical +
BLS
A1FI, A2A, A2B,
A2C,
B1A, B2A, B2B WA
2020/2050/
2080 Global 1990
Salack (2006) Scenario DSSAT 4
(+1 8C, +1.5 8C,
+3 8C)/ (+5%,
+10%, +20%)
Niger/
Burkina
2020/2050/
2080
Millet( mtdo/ zatib
genotypes), sorghum 1961/1990
Table 1:
Future projections suggest a drier
western Sahel (e.g., Senegal, part of Mali)
A wetter eastern Sahel (e.g., Mali, Niger)
No change or slight increases in annual
rainfall towards more southern locations
(e.g., Ghana, Nigeria) (Hulme et al.,
2001,Adiku et al.,2014).
Literature Review Cont’d
Fig.1b: Median Temperature change (%) for Mid-
century RCP8.5 over West Africa
Fig.1a: Median Precipitation Change (%) for Mid-
century RCP8.5 over West Africa
RESEARCH QUESTIONS
How do process-based crop models perform on diverse photoperiod
sensitive sorghum varieties under current climate system and near
future climate change scenarios in the terms of yield potentials across
semi-arid region?
Which definition of onset of rain is most appropriate to define the start
of growing season (OGS) and fitted into farmer’s planting time, for
major cereal crops(maize, millet and sorghum) across agroecological
zones of Mali?
AIMS AND OBJECTIVESAims:“To address the need for substantial improvement in the characterization of foodsecurity risks and enhance the development of adaptation measures for Sub-Sahara Africa (SSA) in the circumstances of the changing growing environmental(biophysical) conditions”.
Specific objectives are to;
evaluate the onset and length of growing season in order to establish the
most suitable dates for planting major cereal crops in the agro-ecological zones
of Mali;
determine the effect of sowing date on photoperiod sensitive sorghum
genotypes and yield potentials under non-limiting water and nutrient supply;
assess the process-based crop growth models (DSSAT, APSIM and Samara)
improvements through model calibration and validation for phenology and
yield prediction in sorghum;
provide comparison of the sensitivity of the current system to climate change,
and then recommend the most suitable adaptation strategies.
Fig.3: Map of the Mali showing the selected rainfall station and
ecological zones in accordance with the annual mean rainfall
OGS was evaluated
from four (4)
definitions of onset
of rain by;
Def_1 -Sivakumar,
(1988)
Def_2 -Kasei and
Afuakwa, (1991).
Def_3 - Omotosho et
al., (2000)
Def_4 –FAO, (1978)
CGS - Cessation of
growing season
defined after Traore
et al., (2000)
LGS = CGS - OGS
Research Methodology – PART 1
Hypothesis• OGS - onset dates was validated with farmers sowing window
for maize, millet and sorghum
– Accept null: if the mean onset date provided at least 7days
to farmers planting date
• LGS was evaluated with duration to maturity of some major
crops varieties (FAO, 2008)
Crop type Local name Selected name Breeder Variety
maturity
Duration from
planting to
Maturity(days)
Maize Zangueréni Zangueréni IER Early 80 - 90
Dembagnuman Obatanpa CIMMYT/CRI Medium 105-110
Sotubaka Suwan 1-SR CIMMYT/IITA Late 110–120
Millet Sossat Sossat c-88 ICRISAT/IER Early 90
Toroniou Toroniou IER Medium 100 -110
M9D3 M9D3 IER Late 125 -130
Sorghum Jakumbe CSM63E IER Early 100
Jigui Seme CSM388 IER Medium 125
Soumalemba IS15-401 CIRAD/ICRISAT Late 145
Table 2: Characteristics of the most cultivated crop varieties within the West Africa semiarid tropics.
Research Methodology –PART 2
Fig. 4: Study Area
The field experiment was
conducted under non-limiting
water and nutrients supply
CLIMATIC CONDITION AT FIELD SITE
Fig.4c: Climatic pattern of the experimental site
Fig.4b: An Automatic weather station less than 500m away from sorghum field trial
Code Genotypes
Name
Race/type Geographical
Origin
Target use quality of
Stover
grain quality Plant
type
G1 CSM63E Guinea Mali Biomass Poor Good int
G2 621 B Caudatum Senegal Dual purpose High Good short
Fig. 10a: Model-simulated total leaf numbers (TLN) against the observed TLN
values for all cultivars used over the three sowing dates (Jun14, July 09, Aug.05).
0
2
4
6
8
10
0 2 4 6 8 10
Sim
ula
ted
M
ax L
AI(
m2/m
2)
Observed Max LAI(m2/m2)
APSIM
DSSAT
SAMARA
Fig. 10b: Model-simulated maximum leaf area Index (MaxLAI) against the
observed MaxLAI values for all cultivars used over the three sowing dates (Jun14,
July 09, Aug.05).
APSIM: RMSE =2.2, NRMSE =
10.6 %, R2= 0.88;
DSSAT: RMSE =2.0, NRMSE =
9.6%, R2= 0.86;
Samara: RMSE =1.3,NRMSE =
6.4 %, R2= 0.96
APSIM:
RMSE=2.4,NRMSE = 85
%, R2= 0.1;
DSSAT:
RMSE=2.6,NRMSE = 92
%, R2= 0.5;
Samara: RMSE=
0.9,NRMSE = 33 %, R2=
0.4
Model-calibrated and observed for TLN and Max LAI
Fig.11: Comparison of model-validation for duration to flowering and maturity with field
observed
Model performance against independent trials for phenology under different growing season, locations and planting densities
0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
CSM63E CSM335 Fadda IS15401
Gra
in y
ield
(k
g/h
a)
Cultivars
Observed
DSSAT
APSIM
SAMARA
Fig12a
0
5000
10000
15000
20000
25000
CSM63E CSM335 Fadda IS15401
To
tal b
iom
ass(k
g/h
a)
Cultivars
Observed
DSSAT
APSIM
SAMARA
Fig12b
Table 9: Grain yield(kg/ha)
APSIM DSSAT SAMARA
RMSE(kg/ha) 833 753.0 810.0
NRMSE(%) 40.0 36 38
R2 0.6 0.6 0.4
Total biomass(kg/ha)
APSIM DSSAT SAMARA
RMSE(kg/ha) 3798 3144 3653
NRMSE(%) 40 33 39
R2 0.8 0.8 0.5
Models performance against an independent dataset for grain yield and total biomass under different growing season, locations and planting densities
OBJECTIVE 4- CLIMATE CHANGE SCENARIOS AND IMPACTS ON SORGHUM PRODUCTION
• Climate scenarios from CMIP5 GCMs using a 30-year baseline dailyweather of MODERN-ERA RETROSPECTIVE ANALYSIS FOR RESEARCH ANDAPPLICATIONS (MERRA) dataset(1980-2009)
• For future projections (2040-2069), five GCMs namely CCSM4, GFDL-
ESM2M, Had GEM2-ES, MIROC5, and MPI-ESM-MR (Rosenzweig et al.
2013) were used for the RCP 8.5 scenario that assumes an elevated
CO2 concentration of 571 ppm compared with the current 390 ppm.
Projected decline change towards
western Sahel significant increase
change towards eastern and southern
Sahel
All the GCMs seasonal rainfall
projected changes differs across the
station, CCSM4 and MIROC5 projected
above baseline except Nioro du Rip
Fig. 13: Projected change (%) in the growing season (May to October) rainfall between Baseline (1980-2009)
and GCM’s future projection (2040- 2069) .
RCP8.5 analyses – Climate change impact on moisture regime between Baseline and GCM’s future projectionSeasonal Rainfall
-40
-30
-20
-10
0
10
20
30
40
Ch
an
ge
in
sea
so
na
l ra
infa
ll(%
)
Nioro du Rip, Senegal Current average = 720 mm(a)
-40
-30
-20
-10
0
10
20
30
40
Ch
an
ge
in
se
as
on
al ra
infa
ll(%
) Sadore, Niger Current Average = 517 mm(b)
-20.0
-15.0
-10.0
-5.0
0.0
5.0
10.0
15.0
20.0
Ch
an
ge
in
sea
so
na
l ra
infa
ll(%
)
Navrongo, Ghana, Current Average = 903mm(c)
Onset of growing season (OGS)
Fig. 14: Comparison between Baseline(1980-2009) and GCMs mean
projection (2040-2069) for estimated Onset of growing season
Low significant change ( -5 to +7days) -
uncertainty would lies in the distribution of
rainfall during the growing period
Sadore and Navrongo projected early OGS
– corroborate the projection of more wetter
future climate.
08-Jun
13-Jun
18-Jun
23-Jun
28-Jun
03-Jul
08-Jul
On
se
t o
f g
row
ing
sea
so
n
Nioro du Rip, Senegal
Baseline (1980-2009)
Median_RCP 8.5 (2040-2069)
(a)
02-Jun
04-Jun
06-Jun
08-Jun
On
se
t o
f g
row
ing
sea
so
n
Sadore, Niger
Baseline (1980-2009)
Median_RCP 8.5 (2040-2069)
(b)
04-Jun
05-Jun
06-Jun
07-Jun
08-Jun
On
se
t o
f g
row
ing
sea
so
n
Navrongo, Ghana
Baseline (1980-2009)Median_RCP 8.5 (2040-2069)
(f)
Length of growing season (LGS)
Fig. 15: Comparison between Baseline(1980-2009) and GCMs mean projection
(2040-2069) for estimated length of growing season (LGS)
Sadore: LGS shows significant increase in
(4) and decrease in (1); inter-annual
variability is high
Nioro: LGS decrease (3), no change (2),
variability remains high
Navrongo: No change variability remains
moderate
CCSM4 projected increase across the
stations except Nioro du Rip
80
100
120
140
160
180
Len
gth
of
gro
win
g s
ea
so
n(d
ays
)
Nioro du Rip
Baseline (1980-2009)
Median_RCP 8.5 (2040-2069)
(a)
80
90
100
110
120
130
140
Len
gth
of
gro
win
g s
ea
so
n(d
ays
)
Sadore, Niger
Baseline (1980-2009)
Median_RCP 8.5 (2040-2069)
(b)
80
100
120
140
160
180
Len
gth
of
gro
win
g s
easo
n(d
ays)
Navrongo, Ghana
Baseline (1980-2009)
Median_RCP 8.5 (2040-2069)
(c)
Fig. 16: Comparison of average monthly variability of minimum temperature between the
Baseline (1980-2009) and GCMs Scenario (2040-2069) for the selected stations
Both Tmax and Tmin uniformly increase
throughout growing season between
baseline and the GCMs projection
Tmin projected faster in magnitude than
Tmax
Suggests increase in GDD for the
crops,
Exacerbated moisture stress in rainfed
agriculture leads to grain weight loss
Climate change impact on temperatures regime between Baseline and GCM’s future projection
10
15
20
25
30
35
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Ave
rag
e T
min
(0C
)
Nioro du Rip, Senegal
BASELINE
CCSM4
GFDL-ESM2M
HadGEM2-ES
MIROC5
MPI-ESM-MR
(a)
Growing season
10
15
20
25
30
35
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Ave
rag
e T
min
(0C
)
Sadore,Niger
BASELINE
CCSM4
GFDL-ESM2M
HadGEM2-ES
MIROC5
MPI-ESM-MR
(b)
Growing season
20
25
30
35
40
45
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Ave
rag
e T
ma
x (
0C
)
Navrongo, Ghana
BASELINE
CCSM4
GFDL-ESM2M
HadGEM2-ES
MIROC5
MPI-ESM-MR
Growing season
(c)
Fig. 17: Projected change for minimum temperature between baseline (1980-2009
and GCMs scenario (2040-2069)
All GCMs project increased
temperature at varying
magnitudes across six stations
Highest value was projected
by HadGEM2-ES followed by
MPI-ESM-MR while the least
warming is projected by CCSM4
except at Nioro du Rip
Minimum temperatures projection change
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
Ch
an
ge
in
Ave
rag
e T
min
(0C
)
Nioro du Rip, Senegal, Current average = 23.7 0C, ∆=0.12 0C(a)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
Ch
an
ge
in
Ave
rag
e T
min
(0C
)
Sadore, Niger Current average = 25.6 0C ∆= 0.14 0C(b)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
Ch
an
ge i
n a
vera
ge T
min
(0C
)
Navrongo, Ghana, Current average = 22.9 0C ∆=0.11 0C(c)
Fig. 18: Projected change for maximum temperature between baseline
(1980-2009 and GCMs scenario (2040-2069)
Maximum temperatures projection change
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
Ch
an
ge
in
Ave
rag
e T
ma
x(0
C)
Nioro du Rip, Senegal , Current Average = 34.40C , ∆=0.140C(a)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
Ch
an
ge
in
Ave
rag
e T
ma
x(0
C)
Sadore, Niger Current average = 36.9 0C ∆= 0.19 0C(b)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
Ch
an
ge
in
ave
rag
e T
ma
x(0
C)
Navrongo,Ghana, Current average = 33 0C ∆ =0.13 0C (c)
0
1000
2000
3000
4000
5000
6000
Ba
se
lin
e y
ield
(k
g/h
a)
CSM63E
APSIM
DSSAT
Samara
(a)
0
1000
2000
3000
4000
5000
6000
Ba
se
lin
e y
ield
(k
g/h
a)
CSM335
APSIM
DSSAT
Samara
0
1000
2000
3000
4000
5000
6000
Ba
se
lin
e y
ield
(k
g/h
a)
Fadda
APSIM
DSSAT
Samara
(c)
0
1000
2000
3000
4000
5000
6000
Ba
se
lin
e y
ield
(k
g/h
a)
IS15401
APSIM
DSSAT
Samara
(d)
Models sensitivity under baseline climate
Fig. 19: Simulated yield of CSM63E, CSM335, Fadda and IS15401 under the baseline climate (1980–
Results• CSM63E- DSSAT simulated lower grain yield compared to
APSIM and Samara, low inter-annual variability except at Mopti
and Kano by DSSAT
• CSM335 -DSSAT and Samara shows higher inter-year
variability across the sites compared to APSIM model, highest
grain yield simulated at Koutiala and lowest grain yield at
Sadore.
• Fadda –exhibited high grain yield potential, inter-annual
variability remains high across the models and sites
• IS15401 – model simulated low grain yield across the sites
-30
-20
-10
0
10
20
30
Rela
tiv
e C
han
ge i
n
gra
in y
ield
(%
)
CSM63E - without Adaptation
Mopti
Sadore
Nioro du Rip
Kano
Koutiala
Navrongo
(a)
Impact of projected GCMs scenario on sorghum cultivars without adaptation
-30
-20
-10
0
10
20
30
Rela
tiv
e
Ch
an
ge i
n g
rain
yie
ld (
%)
CSM335 - without Adaptation
Mopti
Sadore
Nioro du Rip
Kano
Koutiala
Navrongo
(b)
-30
-20
-10
0
10
20
30
Rela
tiv
e C
han
ge i
n g
rain
yie
ld (
%)
Fadda - without Adaptation
Mopti
Sadore
Nioro du Rip
Kano
Koutiala
Navrongo
(c)
-30
-20
-10
0
10
20
30
Rela
tiv
e C
han
ge i
n g
rain
yie
ld (
%)
IS15401 - without Adaptation
Mopti
Sadore
Nioro du Rip
Kano
Koutiala
Navrongo
(d)
Fig. 20: Comparison of the relative change (%) in yield projection for the cultivars between the baseline and
future projected climate scenario (2040-2069) without Adaptation across selected sites
-30
-20
-10
0
10
20
30
Rela
tiv
e c
han
ge i
n g
rain
yie
ld(%
)
CSM63E With Adaptation
Mopti
Sadore
Nioro du Rip
Kano
Koutiala
Navrongo
(a)
-30
-20
-10
0
10
20
30
Rela
tiv
e c
han
ge i
n g
rain
yie
ld (
%)
FADDA – With Adaptation
Mopti
Sadore
Nioro du Rip
Kano
Koutiala
Navrongo
(c)
-30
-20
-10
0
10
20
30
Rela
tive
Cha
ng
e in
g
rain
yie
ld(%
)
IS15401 – With Adaptation
Mopti
Sadore
Nioro du Rip
Kano
Koutiala
Navrongo
(d)
-40
-30
-20
-10
0
10
20
30
40
Rela
tiv
e C
han
ge i
n g
rain
yie
ld (
%)
CSM335- With Adaptation
Mopti
Sadore
Nioro du Rip
Kano
Koutiala
Navrongo
(b)
Impacts of adaptation measure on genotypic difference under climate change
Fig. 21 : Comparison of the relative change (%) in yield projection for the cultivars between the baseline
and future projected climate scenario (2040-2069) with Adaptation across selected sites
Discussions• Medium and late maturity cultivars found to be photoperiodically sensitive
and strong response to variation in sowing dates
• Calibration shows the models capability to predict crop duration for theagronomically relevant range of sowing dates.
– A near perfect fit was observed for the phenological growth stagesbetween the crop model-simulated and field-observed values
– the uncertainty lied in the prediction of total grain yield and biomass
• Total biomass and grain yield varied strongly among the models, thevariation from models output could be linked to model internal mechanismor quality of the field data.
• On the sensitivity of current systems to climate change:
– Decline changes in yield output between baseline and 5GCMs for all themodels across sites
– Models showed effect of the latitude and photoperiod on the cultivars(e.g. Fadda)
– High demand for water (CSM335 and IS15401) which resulted in low yield
– the increase in rainfall amounts projected by some GCMs (e.g. CCSM4)does not match with the projected increase in mean simulated grainyields
– Tmin projected faster than Tmax that suggests increase in GDD
Conclusions The determination of onset date of growing season from single
method across AEZ of Mali may lead to false onset or too late date
estimation.
Based on the estimated LGS across AEZ and evaluation with
duration to maturity of major crops varieties, the results suggest
early-maturing varieties for Sahelian zone,
early and medium maturing varieties for Sudano-sahelian zone,
All level of maturity for Sudanian and Guinean zones provided the flowering
time would occur 15-20days prior to CGS (e.g. sorghum and millet) or varieties
that can withstand the terminal drought(CGS) during grain filling
The novel and apparent merit of this study is that
Crop modelling is found as a valuable tool to understand
genotype × environment × management (G × E × M) interactions
on crop growth and yield potential
Nearly all the widely used crop models tested showed their
capability in assessing climate impacts/risk for range of
photoperiod sensitive sorghum cultivars
Conclusions cont’d The study confirmed warming across the dryland West
Africa (high confidence) – seemingly faster in cooler areas
(e.g. Nioro du Rip, Senegal).
Rainfall may likely increase eastwards, decrease westwards
and slight increase/no change southward: this suggests
climate adaptation will be local
Impacts of projected changes by GCM’s vary significantly
across different study sites compared and cultivars.
Projected yields changes from three crop models at different
contrasted sites, it suggests an insight on the need for climate-
smart varieties as long-time plan adaptation strategy to
ensure increase productivity under warming projected climate.
CONTRIBUTION OF THE RESEARCH TO KNOWLEDGE
Strengthened the prediction skill to define the onset of growing
season, as well as the length of growing season in semi-arid region in
order to minimize climatic risk especially for staple crops(maize, millet
and sorghum)
Crop models improvement through calibration of photoperiod sensitive
sorghum for the growth parameters and yield development was
established
Application of multi-model climate change scenarios projection (GCMs)
into dynamic crop models for enhancing sorghum productivity in West
Africa semi-arid tropics and the development of the adaptation
strategies.
Recommendations Further evaluations of onset date via participatory approach
with farmers, agrometeorologists and agriculture extensionofficers, for ‘on-line’ dissemination to farmers;
As modelling can help reduce number of field experiments andcan save resources, it is therefore recommended that a reliableyield projection should be cultivar specific through modelcalibration and validation with data sets from carefully-conducted experiments;
Crop breeders should work closely with both climate and cropmodellers in the region to improve on climate-smart traits insorghum varieties that would be more resilient to elevatedmean temperature during the growing period;
Many, many more models exist and much, much moreuncertainty subsists. Regional capacity to operate modelsand interpret projections is lacking and must be aggressivelydeveloped – e.g. through science-policy platforms
ACKNOWLEDGEMENT• This research study was funded by Federal Ministry
of Education and Research (BMBF) through the West African Service Centre on Climate Change and
Adapted Land Use (WASCAL), Graduate Research Program (GRP). Financial support is gratefully
acknowledged.
• Grateful to the University management and Department for my study leave.