Transcript Climate, Water and Agriculture: Impacts and adaptation in Africa

Climate, Water and Agriculture:
Impacts and adaptation in Africa
Core funding from GEF plus complementary funding from others (WBI
Finish Trust, NOAA, CEEPA McArthur, WB ARD, IWMI, FAO)
2002 - 2005
WWW.CEEPA.CO.ZA/CLIMATE CHANGE
EEE Program Seminar, ICTP June 10, 2003
Rashid Hassan, CEEPA
Motivation
• Agro-ecosystems in Africa – most vulnerable to climate
change (CC)
– Climate already hot in most parts of Africa
• productivity decline with warming - crop yields
• More pressure:
–
–
–
–
Higher demand for land conversion
More water for irrigation
Increased intensification (pollution, erosion, etc.)
Introduction of GM plants and animals (biodiversity)
– High dependence on agriculture – livelihoods
– Low ability of African farmers to adapt
• Limited access to capital and technological options
• Poor public infrastructure (roads, information, research, extension)
Objectives
• Improve the capacity of research and
policy in participating countries and the
region to:
– Assess the impacts of CC on agro-ecosyst.
– Evaluate alternative adaptation options
• Generate improved information and
knowledge on impacts of CC and possible
options for adaptation
Scope
• Covers 11 countries from north, south, east and
west Africa
– Burkina Faso, Cameroon, Egypt, Ethiopia, Ghana,
Kenya, Niger, Senegal, South Africa, Zambia and
Zimbabwe (work began in 8)
• Involves collaboration with a number of
•
institutions on country level and regional
analyses
IWMI, FAO, YALE working with multidisciplinary
country teams
Approach and Methods
• Three approaches integrated:
– Ricardian approach to economic impact
assessment
– River basin hydrology models for assessing
impacts on runoff and availability of water
– Crop-water response simulation models for
assessing the biological impacts (crop yields)
The Ricardian Approach 1
• Based on the hypothesis that impacts of
•
•
changes in climate attributes (temperature,
rainfall) like other long-term economic
phenomena are capitalized in land values
CC affects crop yields and hence farm revenues
capitalized in land value changes over time
(present value of the stream of future revenues)
Regression of land values on various
determinants of net revenue including climate
variables
VLt = t NRt (1+ )-t
NRt
=
F(RFt, TMt, Zt, Xt, TKt)
The Ricardian Approach 2
• Difficult to have long time series on land values
•
•
•
or net revenue by region and all regressors
Cross-sectional data reflecting spatial diversity in
climate attributes and all other determinants of
land value (NR) alternatively used
Partially controls for adaptation (substitution
between inputs in production)
Other adaptation missed (crop switching,
transitory adjustment costs – capital goods)
Water-crop response simulation models 1
• Describe crops’ behavior (physiological and
development) as a function of:
– Climate (weather) factors
– Soil conditions (fertility, water holding capacity, etc.)
– Other determinants under farmer control and
management (planting dates and input levels)
• Simulate yield response to variability in climate
• Other response variables such as crop water
consumption and irrigation requirements as
climate changes (evapotranspiration, soil
moisture, water excess-deficit at critical stages
of growth, etc.)
Water-crop response simulation models 2
• Compliments the cross-sectional analysis –
Ricardian (both spatially implemented)
• Overestimates impacts – no control for
adaptation (recent modifications to allow)
• Crop specific and location specific
• Experimental – real world replications of
conditions (calibrations to actual locations’
data)
River basin Hydrology
• Models how CC affects agriculture also
indirectly through impacts on basin
hydrology (runoff):
– Rainfall, temp. and evaporation affect seasonal
patterns of river flow and hence availability of
water for agriculture
• Compliments the cross-sectional analysis –
Ricardian method
– Both spatially-based
Data and Plans of analyses 1
• District level data
– Farm survey data at district level (same year 2002)
aggregated by country and region (river basins)
• Crop yields, prices and production costs (NR)
• Other agric. response variables (% land under crops)
– Climate and soil attributes data by district again
aggregated at river basin
– Other data at district level (population, proximity to
markets, literacy and disease rate, etc.)
• Regional data-base to sport the cross-country
assessment of economic impacts
Data and Plans of analyses 2
• Hydrology models analyses’ generate input
variables to the Ricardian regression by district
– Change in runoff, soil moisture, etc.
• Crop water response models calibrate the
biological response of crops to CC – by district
– Yield
– Water requirements
• The Ricardian analyses of the economic impact
•
applied at both country and district levels
First Year:
– Surveys designed and data collection instruments
developed and tested in the various countries
– Training on the three methods and approaches
conducted