Post on 11-Nov-2014
description
Landscape level hydrological modeling
&
Farm-scale modelingFred Kizito, Katrien Descheemaeker, Sabine Douxchamps 3 / 7 / 2012
Landscape level hydrological modeling
Fred Kizito, Katrien Descheemaeker, Sabine Douxchamps 3 / 7 / 2012
Andes • Ganges • Limpopo • Mekong • Nile • Volta
Study objectivesModeling hydrological dynamics to quantify water fluxes for achieving optimal crop‐livestock productivity
o Assess sub‐basin scale water balance thresholds at target sites
o Develop water allocations framework in target sites
o Recommend best‐fit integrated rainwater management strategies that maximize productivity
Andes • Ganges • Limpopo • Mekong • Nile • Volta
Study sites
Landscape hydrological modeling:
o Conduct sub‐basin water balance thresholds
o Develop a water allocations framework in target sites
o Assess water productivity in specific crop‐livestock systems
Andes • Ganges • Limpopo • Mekong • Nile • Volta
• Baseline characterization has been conducted in target sites at the household level
• Tools: and
• SWAT hydrological modeling is physically based– Weather, soil properties, topography, vegetation,
and land management practices data sets
• DEM:– Used at 90 m resolution– Watershed delineation; Stream network
Methods
Andes • Ganges • Limpopo • Mekong • Nile • Volta
6
Crop water use trends in Golinga
Data Source: Ministry of Food and Agriculture, GhanaProduction estimates and Regional Crop Acreage data for 1992 to 2010 ‐ Complemented and verified with V2 Household survey data
Andes • Ganges • Limpopo • Mekong • Nile • Volta
7
Water, crops and livestock distribution for Golinga
Source: Processed from FAO Geo‐portal data‐Not checked against V2 HH data
Source: Ramankutty et al, 2000Processed from Global Croplands database;Complemented with Ghana MoFA Data and V2 Household data
Andes • Ganges • Limpopo • Mekong • Nile • Volta
8
Water Balance Components for Golinga
0
200
400
600
800
1000
120019
8019
8119
8219
8319
8419
8519
8619
8719
8819
8919
9019
9119
9219
9319
9419
9519
9619
9719
9819
9920
0020
0120
0220
0320
0420
0520
0620
0720
0820
0920
1020
1120
1220
1320
14
Rainfall (m
m) a
nd Discharge (m
m)
Rainfall (mm) Surface Water Discharge (mm) Groundwater Discharge (mm)
Percolation (mm) Evapotranspiration (mm)
Warm‐up Calibration ValidationSim
ulated
Andes • Ganges • Limpopo • Mekong • Nile • Volta
Milestones:• Cropping density and livestock distribution ascertained for all study
sites; Water balance thresholds calculated for all study sites• Currently developing crop‐livestock water productivity maps for all
target sites• Landscape outputs from water allocations and water balance will
complement farm‐level flows analysis
Conclusion• Hydrological analysis indicated that reservoirs play a critical role in
maintaining storage and reducing surface runoff losses at sub‐basin scale
Conclusion
Farm-scale modeling
Fred Kizito, Katrien Descheemaeker, Sabine Douxchamps 3 / 7 / 2012
Andes • Ganges • Limpopo • Mekong • Nile • Volta
ObjectivesIdentify and evaluate promising interventions for improved farm productivity• Extrapolating field results in space and time• Aggregate field level outputs to farm level• Scenario analysis: exploring options• Risk analysis• Tradeoff analysis (tradeoffs in resource allocation)• Identifying issues for further (field) research• Discussion and decision support tool: informing the
innovation platform
Andes • Ganges • Limpopo • Mekong • Nile • Volta
NPKNPK
NPK
Options
Giller et al. 2010
Andes • Ganges • Limpopo • Mekong • Nile • Volta
NUANCES-FARMSIM: farm-scale modeling approach
Tittonell et al. (2007) Fld Crops Res. 100, 348-368; Rufino et al. (2007) Livestock Sci. 112, 273-287; Chikowo et al. (2008) Ag. Syst. 97, 151-166; Tittonell et al. (2009) Ag. Syst. 101, 1-19; van Wijk et al. (2009) Ag. Syst. 102, 89-101; Tittonell et al. (2010) E. J Agron.32, 10-21.
Andes • Ganges • Limpopo • Mekong • Nile • Volta
APSIM (Agricultural Production Systems sIMulator)
Andes • Ganges • Limpopo • Mekong • Nile • Volta
In-house feeding
Grazing
Feed gap
Constraint analysisExample of feedbase in villages around Golinga reservoir
Andes • Ganges • Limpopo • Mekong • Nile • Volta
Baseline situation• 1.5 ha farm• household of 8 people• crops: millet, sorghum and cowpea intercropped• no crop residue stored for cattle• 3 breeding cows, sells at 4-5 years, herd of 8-10
Scenario Analysis
Adapted from McDonald (2010)
Andes • Ganges • Limpopo • Mekong • Nile • Volta
Baseline
Animals sold (10y) 5‐6
Animals on hand 12‐13
Forage deficit 7000
Wet season labour +50
Cattle revenue 34000
Gross Margin* 515000
Cash balance ‐3000
* - including home consumption
Scenario Analysis
Adapted from McDonald (2010)
Andes • Ganges • Limpopo • Mekong • Nile • Volta
Baseline Manure (4 t/ha)
Animals sold (10y) 5‐6 6‐7
Animals on hand 12‐13 13
Forage deficit 7000 6000
Wet season labour +50 +20
Cattle revenue 34000 37000
Gross Margin 515000 637000
Cash balance ‐3000 109000
Scenario Analysis
Adapted from McDonald (2010)
Andes • Ganges • Limpopo • Mekong • Nile • Volta
Baseline Manure (4 t/ha)
Crop residue harvesting
Animals sold (10y) 5‐6 6‐7 7‐8
Animals on hand 12‐13 13 13
Forage deficit 7000 6000 3000
Wet season labour +50 +20 +10
Cattle revenue 34000 37000 41000
Gross Margin 515000 637000 671000
Cash balance ‐3000 109000 140000
Scenario Analysis
Adapted from McDonald (2010)
Andes • Ganges • Limpopo • Mekong • Nile • Volta
Baseline Manure (4 t/ha)
Crop residue harvesting
Sell cow, buy 10 sheep &
fatten
Calves sold (10y) 5‐6 6‐7 7‐8 6‐7
Cattle on hand 12‐13 13 13 9‐10
Forage deficit 7000 6000 3000 4400
Wet season labour +50 +20 +10 +50
Livestock revenue 34000 37000 41000 96000
Gross Margin 515000 637000 671000 739000
Cash balance ‐3000 109000 140000 205000
Scenario Analysis
Adapted from McDonald (2010)
Andes • Ganges • Limpopo • Mekong • Nile • Volta
Discussion support tool Learning tool
Scenario Analysis
Adapted from McDonald (2010)
Andes • Ganges • Limpopo • Mekong • Nile • Volta
Simulation experiment
Andes • Ganges • Limpopo • Mekong • Nile • Volta
Lessons:- Fertilizer increases average yield, but also production risk- Information on risk is useful for insurance providers (partner in the IPs?)- Water and nutrient use efficiency are interlinked
Simulation experiment
Andes • Ganges • Limpopo • Mekong • Nile • Volta
Understanding resource allocation decisions
Resources are finite; directing them to one objective will penalize other objectives
• Labor: weeding vs. marketing produce• Cash: fertilizers vs. hiring labor for weeding • Crop residues: soil organic matter vs. livestock feeding
Tradeoff analysis
Andes • Ganges • Limpopo • Mekong • Nile • Volta
concentrates
ferti
lizer
Tradeoff analysis
Andes • Ganges • Limpopo • Mekong • Nile • Volta
concentrates
ferti
lizer
Tradeoff analysis
Andes • Ganges • Limpopo • Mekong • Nile • Volta
concentrates
ferti
lizer
Tradeoff analysis
Andes • Ganges • Limpopo • Mekong • Nile • Volta
concentrates
ferti
lizer
Tradeoff analysis
Andes • Ganges • Limpopo • Mekong • Nile • Volta
concentrates
ferti
lizer
Tradeoff analysis
Andes • Ganges • Limpopo • Mekong • Nile • Volta
concentrates
ferti
lizer
Lessons:- Tradeoff analysis helps us in systems understanding- Linked with understanding of socio-institutional settings (e.g. market) and farmers’
objectives, this can be used to design well-adapted interventions
Tradeoff analysis
Andes • Ganges • Limpopo • Mekong • Nile • Volta
Farm systems models are useful tools for research to - Understand complex farm dynamics, including farmer
decision making- Identify topics for further (field) research
for development through - Assisting in the development of adapted interventions- Generation of information for discussion support (in IPs)
! Need for high quality input data
Conclusions
Andes • Ganges • Limpopo • Mekong • Nile • Volta
Merci pour votre attention!Thanks for your attention!