Title: Toward an integrated modelling of crop-climate interactions: including tropical croplands in a land surface model
1 Représentation des zones cultivées tropicales
dans le modèle de surface continentale de
lIPSL, ORCHIDEE apport à létude des
interactions climat/agriculture
Alexis BERG
Thèse réalisée au (1) LSCE IPSL, Paris, France
(Directrice de thèse Nathalie de Noblet) (2)
LOCEAN IPSL, Gif-sur-Yvette, France
(co-directeur Benjamin Sultan)
2The crop / climate system
Studying the relationships between climate and
agriculture raises a twofold issue
3The crop / climate system
Studying the relationships between climate and
agriculture raises a twofold issue
- IMPACT of climate on crop productivity
- Climate mean productivity
- Weather interannual variability
Impact of climate variability, impact of climate
change ?
4The crop / climate system
Studying the relationships between climate and
agriculture raises a twofold issue
- FEEDBACK of croplands expansion to climate
- biogeochemical effects (e.g., CO2)
- biogeophysical effects albedo, evaporation,
roughness
Foley et al. (2003)
5The crop / climate system
Studying the relationships between climate and
agriculture raises a twofold issue
- FEEDBACK of croplands expansion to climate
- biogeochemical effects (e.g., CO2)
- biogeophysical effects albedo, evaporation,
roughness
Cropland area 12
Ramankutty et al., 2008
Impact of land-use change on climate ?
6Crop / climate interactions in the TROPICS
- Tropical regions (Africa) vulnerable to climate
change - dependence on climate
- poverty
- demographic pressure
7Crop / climate interactions in the TROPICS
- Tropical regions (Africa) vulnerable to climate
change - dependence on climate
- poverty
- demographic pressure
Present
Future
Cereals per capita
Estimates by 2050, food production demand x 2
in Asia and Latin America x 5 in Africa
Kg/capita
Africa
2000
1960
1980
Developed countries
Importance of climate change impacts on crop
yields in the TROPICS (AFRICA)
Africa
World
Asia developing
Latin / Caribbean
Population increase gt food supply increase
8Crop / climate interactions in the TROPICS
Past land-use change (1992-1860)
Future land-use change (2100-1992)
Davin et al. 2007
A2 socio-economic scenario
9Crop / climate interactions in the TROPICS
Past land-use change (1992-1860)
Future land-use change (2100-1992)
Davin et al. 2007
A2 socio-economic scenario
- Projected future land-use change Tropics
Importance of land-use change for future climate
in the TROPICS
101) Modeling the impact of climate on crop yields
Climate
Plot-scale, process-based crop model
Different crops different crop models
111) Modeling the impact of climate on crop yields
- Linking crop models and climate models crop
yields projections
Crop model i
Crop model i1
Crop model i2
Climate model
121) Modeling the impact of climate on crop yields
- Linking crop models and climate models crop
yields projections
Crop model i
Crop model i1
Crop model i2
Climate model
- scale mismatch
- local assessments / spatially heterogeneous
impacts - / different climate and crops models
131) Modeling the impact of climate on crop yields
- Linking crop models and climate models crop
yields projections
Crop model i
Crop model i1
Crop model i2
Climate model
- scale mismatch
- local assessments / spatially heterogeneous
impacts - / different climate and crops models
IPCC 2007 - Cereal yield change with global
warming
mid-high latitudes
low latitudes
69 studies at multiple simulation sites
Yield change ()
Inconsistent / qualitative assessment Temperate
1-3 K threshold Tropics 1-2 K threshold
Maize
Maize
T change
141) Modeling the impact of climate on crop yields
- Linking crop models and climate models crop
yields projections
Crop model i
Crop model i1
Crop model i2
Climate model
- scale mismatch
- local assessments / spatially heterogeneous
impacts - / different climate and crops models
IPCC 2007 - Cereal yield change with global
warming
Need for more consistent, large-scale, spatially
distributed and quantitative climate change
impact studies
69 studies at multiple simulation sites
Inconsistent / qualitative assessment Temperate
1-3 K threshold Tropics 1-2 K threshold
152) Modeling the feedback of land-use change on
climate
ORCHIDEE (IPSL) 8 tree types C3 and C4 grass
C3 and C4 crops
162) Modeling the feedback of land-use change on
climate
ORCHIDEE (IPSL) 8 tree types C3 and C4 grass
C3 and C4 crops
Land-use change
172) Modeling the feedback of land-use change on
climate
ORCHIDEE (IPSL) 8 tree types C3 and C4 grass
C3 and C4 crops
croplands are approximated by grasslands But
croplands ? grasslands
Need to account for croplands more realistically
in land surface models (LSMs)
18General idea include a representation of
croplands in LSMs, derived from crop
models large-scale, spatially explicit response
of crop yields to climate ? more realistic
feedback to climate ?
19General idea include a representation of
croplands in LSMs, derived from crop
models large-scale, spatially explicit response
of crop yields to climate ? more realistic
feedback to climate ?
For various reasons, most agro-LSMs focus first
on a few crops and/or regions
MODEL Group Crops Scale
LPJ-ml PIK-Postdam 11 crop types World
Agro-Ibis Univ.Wisconsin Corn, soybean, wheat Cont.US
ORCHIDEE-STICS IPSL Wheat, maize, soybean Europe
GLAM-MOSES Hadley Center Groundnuts Tropics
SiBcrop Univ.Colorado Corn, soybean, wheat Cont.US
20General idea include a representation of
croplands in LSMs, derived from crop
models large-scale, spatially explicit response
of crop yields to climate ? more realistic
feedback to climate ?
For various reasons, most agro-LSMs focus first
on a few crops and/or regions
MODEL Group Crops Scale
LPJ-ml PIK-Postdam 11 crop types World
Agro-Ibis Univ.Wisconsin Corn, soybean, wheat Cont.US
ORCHIDEE-STICS IPSL Wheat, maize, soybean Europe
GLAM-MOSES Hadley Center Groundnuts Tropics
SiBcrop Univ.Colorado Corn, soybean, wheat Cont.US
Here develop a representation of tropical, C4
crops in ORCHIDEE.
ORCHIDEE-mil IPSL Tropical C4 cereals Africa, India
21General idea include a representation of
croplands in LSMs, derived from crop
models large-scale, spatially explicit response
of crop yields to climate ? more realistic
feedback to climate ?
ORCHIDEE-mil
SARRAH
millet/sorghum in West Africa
22My objectives
Here, stronger focus on impacts
Impacts
Feedbacks
- To what extent is ORCHIDEE-mil able to represent
tropical crops and to simulate large-scale
impacts of climate on crops in the TROPICS? - - what is the scope of climate change impacts on
crop yields by the end of this century in the
TROPICS?
- How does a more realistic representation of
croplands feed back on land-atmosphere
interactions and climate ?
23- Outlines
- Model development
- Model validation on-site simulations and
large-scale applications - Feedback on land/atmosphere interactions and
climate - Projected impacts of climate change on regional
crop yields over Africa and India
24- Outlines
- Model development
- Model validation on-site simulations and
large-scale applications - Feedback on land/atmosphere interactions and
climate - Projected impacts of climate change on regional
crop yields over Africa and India
25Model development
1) Comparison ORCHIDEE / SARRAH over Bambey,
Senegal. SARRAH proxy of observations
Bambey climate 1997
Bambey research station (on the rainfall map of
West Africa).
ORCHIDEE
SARRAH Obs
26Model development
1) Comparison ORCHIDEE / SARRAH over Bambey,
Senegal. SARRAH proxy of observations
Bambey climate 1997
Bambey research station (on the rainfall map of
West Africa).
ORCHIDEE
SARRAH Obs
Potential climatic yield
- Intensive farming
- modern cultivar
- no nutrient stress, no pests
- high density
27Model development
1) Comparison ORCHIDEE / SARRAH over Bambey,
Senegal. SARRAH proxy of observations
Bambey climate 1997
Bambey research station (on the rainfall map of
West Africa).
ORCHIDEE
SARRAH Obs
LAI
2) Changes to ORCHIDEE - phenology -
carbon allocation scheme - yield elaboration -
LAI computation Assimilation, hydrology are not
modified.
Vegetative phase
Flowering
Grain filling
Dessication
July
Aug
Sept
oct
SARRAHs crop cycle on Bambey in 1997
28Model development
1) Comparison ORCHIDEE / SARRAH over Bambey,
Senegal. SARRAH proxy of observations
Bambey climate 1997
Bambey research station (on the rainfall map of
West Africa).
ORCHIDEE
SARRAH Obs
LAI
2) Changes to ORCHIDEE - phenology -
carbon allocation scheme - yield elaboration -
LAI computation Assimilation, hydrology are not
modified.
Vegetative phase
Flowering
Grain filling
Dessication
July
Aug
Sept
oct
SARRAHs crop cycle on Bambey in 1997
GDD sum
29Model development
1) Comparison ORCHIDEE / SARRAH over Bambey,
Senegal. SARRAH proxy of observations
Bambey climate 1997
Bambey research station (on the rainfall map of
West Africa).
ORCHIDEE
SARRAH Obs
LAI
2) Changes to ORCHIDEE - phenology -
carbon allocation scheme - yield elaboration -
LAI computation Assimilation, hydrology are not
modified.
Vegetative phase
Flowering
Grain filling
Dessication
July
Aug
Sept
oct
SARRAHs crop cycle on Bambey in 1997
GDD sum
ORCHIDEE-mil
30Model development and calibration
Bambey, 1997
J
J
O
F
M
A
M
J
A
S
N
D
Rainfall ORCHIDEE SARRAH(obs) ORCHIDEE-mil
31- Outlines
- Model development
- Model validation on-site simulations and
large-scale applications - Feedback on land/atmosphere interactions and
climate - Projected impacts of climate change on regional
crop yields over Africa and India
32Model validation how ?
- Local scale on-site measurements ?
- Large scale
- against satellite data ?
-
- against large-scale yield data (FAO) ?
scarcity of sites and data
observations are not specific of croplands
uncertainties of yield observations
FAO Food and Agriculture Organization of the
United Nations
33Model validation how ?
- Local scale on-site measurements ?
- Large scale
- against satellite data ?
-
- against large-scale yield data (FAO) ?
scarcity of sites and data
observations are not specific of croplands
uncertainties of yield observations
FAO Food and Agriculture Organization of the
United Nations
34Model validation on-site simulations
Wankama, 2005 -2007 (AMMA site) On-farm
yield Extensive farming
Bambey, other year Potential climatic
yield Intensive farming
Mean rainfall
35Model validation on-site simulations
Wankama, 2005 -2007 (AMMA site) On-farm
yield Extensive farming
Bambey, other year Potential climatic
yield Intensive farming
Mean rainfall
Bambey, 1996
ORCHIDEE SARRAH(obs) ORCHIDEE-mil Rainfall
36Model validation on-site simulations
Wankama, 2005 -2007 (AMMA site) On-farm
yield Extensive farming
Bambey, other year Potential climatic
yield Intensive farming
Pictures by N.Boulain, IRD
Bambey, 1996
Low inputs
ORCHIDEE SARRAH(obs) ORCHIDEE-mil Rainfall
Extensive
70 bare soil
37Model validation on-site simulations
Wankama, 2005 -2007 (AMMA site) On-farm
yield Extensive farming
Bambey, other year Potential climatic
yield Intensive farming
Wankama
2005
Bambey, 1996
Aboveground biomass
2006
2007
ORCHIDEE SARRAH(obs) ORCHIDEE-mil Rainfall
38Model validation how ?
- Local scale on-site measurements ?
- Large scale
- against satellite data ?
-
- against large-scale yield data (FAO) ?
scarcity of sites and data
observations are not specific of croplands
uncertainties yield observations
FAO Food and Agriculture Organization of the
United Nations
39(No Transcript)
40(Berg et al. 2010, Climatic Change)
Model validation regional simulations
Mean yields
Kg/ha
Correct spatial distribution
Mean simulated yields (1965-2000)
Large mean overestimation
41Model validation regional simulations
(Berg et al. 2010, Climatic Change)
Mean yields
- Mean overestimation yield gap between
climatic potential yields and actual, on-farm
yields
Data show - on-farm crop yields are limited by
soil fertility, - potential yields are
consistent with ORCHIDEE-mil results
ICRISAT data, Niger (Vlek and Mokwunye 1988).
42Model validation regional simulations
(Berg et al. 2010, Climatic Change)
Mean yields
- Mean overestimation yield gap between
climatic potential yields and actual, on-farm
yields
Data show - on-farm crop yields are limited by
soil fertility, - potential yields are
consistent with ORCHIDEE-mil results
Without nutrient stress potential
On-farm
ICRISAT data, Niger (Vlek and Mokwunye 1988).
43Model validation regional simulations
(Berg et al. 2010, Climatic Change)
Mean yields
- Mean overestimation yield gap between
climatic potential yields and actual, on-farm
yields
Data show - on-farm crop yields are limited by
soil fertility, - potential yields are
consistent with ORCHIDEE-mil results
Wankama density as a proxy for intensification ?
44(No Transcript)
45Model validation regional simulations
Moderate score (R0.48) what causes ?
46Model validation regional simulations
Moderate score (R0.48) what causes ?
Land-use map
Simulated yields aggregated at country-level
- Climate data
- NCC (NCEP corrected by CRU)
- 1965-2000
- 1x1
ORCHIDEE-mil
Vs
Yield data detrended national yields from FAO
47Model validation regional simulations
Moderate score (R0.48) what causes ?
Land-use map
Simulated yields aggregated at country-level
- Climate data
- NCC (NCEP corrected by CRU)
- 1965-2000
- 1x1
ORCHIDEE-mil
?
Vs
Yield data detrended national yields from FAO
- Uncertainties in model accuracy
48Model validation regional simulations
Moderate score (R0.48) what causes ?
?
Land-use map
Simulated yields aggregated at country-level
- Climate data
- NCC (NCEP corrected by CRU)
- 1965-2000
- 1x1
?
ORCHIDEE-mil
Vs
?
Yield data detrended national yields from FAO
- Uncertainties in model accuracy
- Uncertainties in input data
49Model validation regional simulations
Moderate score (R0.48) what causes ?
?
Land-use map
Simulated yields aggregated at country-level
- Climate data
- NCC (NCEP corrected by CRU)
- 1965-2000
- 1x1
ORCHIDEE-mil
?
Vs
?
?
Yield data detrended national yields from FAO
- Uncertainties in model accuracy
- Uncertainties in input data
- Uncertainties in observations how much
climate do they really contain ?
50Uncertainties in observations how much climate
do they really contain ?...
51Model validation regional simulations
Moderate score (R0.48) what causes ?
?
Land-use map
Simulated yields aggregated at country-level
- Climate data
- NCC (NCEP corrected by CRU)
- 1965-2000
- 1x1
ORCHIDEE-mil
?
Vs
?
?
Yield data detrended national yields from FAO
Model Region Time period Crops R
ORCHIDEE-STICS European countries 1972-2003 Wheat Maize 0.1 - 0.3 0.4 - 0.7
GLAM India 1966-1989 Groundnut 0.76
MCWA Chinese provinces 1985-2002 Maize 0.4 - 0.8
ORCHIDEE-mils skill is similar to other
large-scale crop models.
52Model validation regional simulations
Moderate score (R0.48) what causes ?
?
Land-use map
Simulated yields aggregated at country-level
- Climate data
- NCC (NCEP corrected by CRU)
- 1965-2000
- 1x1
ORCHIDEE-mil
?
Vs
?
?
Yield data detrended national yields from FAO
Correction of daily rainfall intensity and
frequency by observations
R0.55
(Berg et al. 2010, GRL)
53Model validation regional simulations
Moderate score (R0.48) what causes ?
?
Land-use map
Simulated yields aggregated at country-level
- Climate data
- NCC (NCEP corrected by CRU)
- 1965-2000
- 1x1
ORCHIDEE-mil
?
Vs
?
?
Yield data detrended national yields from FAO
Kg/ha
Other region India
R0.68
Observed and simulated yield anomalies
Mean simulated yields (1961-1999)
Same mean bias but better score
54- Outlines
- Model development
- Model validation on-site simulations and
large-scale applications - Feedback on land/atmosphere interactions and
climate - Projected impacts of climate change on regional
crop yields over Africa and India
55(No Transcript)
56Modifications of land-atmosphere fluxes
comparison between ORCHIDEE-mil/ORCHIDEE
- ORCHIDEE-mil millet / standard
ORCHIDEE grasslands
Albedo
Surface Net Radiation
Latent Heat flux
Sensible Heat Flux
Ratio between ORCHIDEE-mil and ORCHIDEE for
36-year annual average
ORCHIDEE-mil millet
ORCHIDEE grasslands
57Modifications of land-atmosphere fluxes
comparison between ORCHIDEE-mil/ORCHIDEE
- ORCHIDEE-mil millet / standard
ORCHIDEE grasslands
Surface Net Radiation
Albedo
Latent Heat flux
Sensible Heat Flux
Ratio between ORCHIDEE-mil and ORCHIDEE for
36-year annual average
ORCHIDEE-mil millet
ORCHIDEE grasslands
58Modifications of land-atmosphere fluxes
comparison between ORCHIDEE-mil/ORCHIDEE
Albedo
Surface Net radiation
Latent heat flux, LAI
Sensible heat flux
Mean annual cycle, over the 36 years of
simulation and a sub-region for ORCHIDEE-mil and
ORCHIDEE
59Modifications of land-atmosphere fluxes
comparison between ORCHIDEE-mil/ORCHIDEE
Albedo
Surface Net radiation
Latent heat flux, LAI
Sensible heat flux
Mean annual cycle, over the 36 years of
simulation and a sub-region for ORCHIDEE-mil and
ORCHIDEE
Differences lt shorter plant cycle
(harvest). What effect on the simulated climate
and monsoon system ?
60Modifications of land-atmosphere fluxes impact
on climate
- ORCHIDEE-mil not coupled to the atmosphere model
asynchronously coupled simulation
61Modifications of land-atmosphere fluxes impact
on climate
- ORCHIDEE-mil not coupled to the atmosphere model
asynchronously coupled simulation
NCC (climate data)
LMDZ-OR (climate model)
ORCHIDEE
Average seasonal cycle of LAI
ORCHIDEE-mil
- Two 30-year global simulations
- prescribed SSTs
- 3.75x 2.5.
100
or
Real vegetation (Satellite LAI)
100
62Modifications of land-atmosphere fluxes impact
on climate
Imposed perturbation
LAI
J F M A M J J A O N D
LAI
ORCHIDEE
ORCHIDEE-mil
63Modifications of land-atmosphere fluxes impact
on climate
Imposed perturbation
Simulated climate
mm/d
C
LAI
J F M A M J J A O N D
J F M A M J J A O N D
J F M A M J J A O N D
Rainfall
Surface air temp.
LAI
ORCHIDEE
ORCHIDEE-mil
- small effect on temperatures 1-2 K at the end
of the rainy season - no significant effect on mean precipitation
64Modifications of land-atmosphere fluxes impact
on climate
- Grassland approximation OK ?
- Preliminary assessment / limitations
- the scale of the surface perturbation / global
simulation - representation of L-A coupling and the West
African monsoon in LMDZ ? - fixed vegetation
No generic conclusion should be drawn here.
Other regions when the contrast between
grasslands and croplands is more pronounced yield
more interesting results (e.g., Amazonia, Costa
et al. 2007)
65- Outlines
- Model development
- Model validation on-site simulations and
large-scale applications - Feedback on land/atmosphere interactions and
climate - Projected impacts of climate change on regional
crop yields over Africa and India
66Projections of climate change impacts on
potential C4 crop productivity over Africa and
India
(Berg et al. 2010, subm. to Agr.For.Met.)
With / without CO2 effect
- Climate data
- IPCC climate projections
- 2 scenarios (A1B, A2)
- 7 and 5 models
- 1961-2100
- 2.5x2.5
ORCHIDEE-mil
Potential yield change
No land-use map crops everywhere
67- Regional biases in climate models
Obs
(Cook and Vizy 2006)
Observed and simulated 19492000 JJAS
precipitation rates (mm/day) in 2 IPCC models
68- Regional biases in climate models
analysis by Köppen bioclimatic zones
- based on T and P - based on threshold values
and seasonality - broadly correspond to biomes
Obs
(Cook and Vizy 2006)
Köppen classification based on CRU data over
1951-2000
Observed and simulated 19492000 JJAS
precipitation rates (mm/day) in 2 IPCC models
69- Regional biases in climate models
analysis by Köppen bioclimatic zones
- based on T and P - based on threshold values
and seasonality - broadly correspond to biomes
Obs
Definition of simpler zones
(Cook and Vizy 2006)
Köppen classification based on CRU data over
1951-2000
Observed and simulated 19492000 JJAS
precipitation rates (mm/day) in 2 IPCC models
70- Regional biases in climate models
analysis by Köppen bioclimatic zones
MIROC_Hi
IPSL
Temperate Arid Desert Eq. dry season Eq. humid
Obs
Obs
(Cook and Vizy 2006)
Observed and simulated 19492000 JJAS
precipitation rates (mm/day) in 2 IPCC models
71(No Transcript)
72(No Transcript)
73Yield change
A1B scenario
By 2035 (2020-2050)
- significant vs interannual variability
Arid
Eq. humid
Eq. dry season
Temperate
By 2085 (2070-2100)
Arid
Eq. humid
Eq. dry season
Temperate
74Yield change
A1B scenario
By 2035 (2020-2050)
- significant vs interannual variability
Arid
Eq. humid
Eq. dry season
Temperate
By 2085 (2070-2100)
Arid
Eq. humid
Eq. dry season
Temperate
75Yield change
A1B scenario
By 2035 (2020-2050)
- significant vs interannual variability
Arid
Eq. humid
Eq. dry season
Temperate
By 2085 (2070-2100)
Arid
Eq. humid
Eq. dry season
Temperate
76Rainfall change, Temperature change and Yield
change
Arid
Eq. humid
A1B
Eq. dry season
Temperate
A2
77Rainfall change, Temperature change and Yield
change
Arid
Eq. humid
A1B
Eq. dry season
Temperate
A2
Rainfall effect only in dry zones
78Rainfall change, Temperature change and Yield
change
Arid
Eq. humid
Arid
Eq. humid
A1B
Eq. dry season
Eq. dry season
Temperate
Temperate
A2
Rainfall effect only in dry zones
79Rainfall change, Temperature change and Yield
change
Arid
Eq. humid
Arid
Eq. humid
A1B
Eq. dry season
Eq. dry season
Temperate
Temperate
A2
Rainfall effect only in dry zones
T effect in other zones acceleration of
phenology
80Rainfall change, Temperature change and Yield
change
Arid
Eq. humid
Arid
Eq. humid
A1B
Eq. dry season
Eq. dry season
Temperate
Temperate
A2
Rainfall effect only in dry zones
T effect in other zones acceleration of
phenology
Consistency of yields change on different zones lt
consistency of climate projections for T/P
81CO2 effect
- Little CO2 effect
- no direct fertilization effect
- improved water use efficiency in dry areas (7)
Consistent with expected impact on C4 crops
82Projections of climate change impacts on
potential C4 crop productivity over Africa and
India
Limitations
- Not all climate impacts are considered
- - climate extremes (heat waves, floods)
- - interactions between climate, CO2 and
pests/diseases
83Projections of climate change impacts on
potential C4 crop productivity over Africa and
India
Limitations
- Not all climate impacts are considered
- - climate extremes (heat waves, floods)
- - interactions between climate, CO2 and
pests/diseases
- Scope of potential climate-related impacts only
- possibility to adapt agricultural practices to
climate change water management, cultivars,
cropping systems - possibility to improve yields today fill the
yield gap -
84Conclusion
- More realistic representation of tropical
croplands in ORCHIDEE - Yield overestimation yield gap
- The large-scale relationship between climate and
yields is correctly captured 23 of variance
explained over West Africa and 46 over India - Uncertainties in large-scale modeling -
accuracy of large-scale yield data - - accuracy of climate and
land use data - Over West Africa a more realistic
representation of croplands - modifies land-atmosphere interactions
- shows little subsequent impact on the simulated
monsoon. -
- Impacts of climate change on yields over Africa
/ India - - Robust, moderately adverse effect of
temperature - - uncertainties in precipitations total
impact -29 / 11 - Impacts may be more than compensated by
adaptation yield gap filling ! - Climate change additional stress
85Outlooks
- Model improvement
- -Yield overestimation use of a
spatially-varying scaling factor to account for
levels of intensification (as in LPJ-mL) - - Calculate cultivars as a function of climate
- - Represent other tropical crops (rice,
soybean) - representation of croplands in ORCHIDEE on the
global scale ? -
- Model use
- - Projections with CMIP5 climate change
simulations - - Impacts on shorter time scales (e.g.,
seasonal) - - Impacts of land use change on regional surface
carbon and water budgets - Feedback on climate coupling to the atmosphere
(interactive vegetation) - Consistent framework to investigate
- - interactions between climate, crops and
irrigation - - what land-use scenario provides the best
climate return in terms of crop yields ?
86Thank you for your attention and to
everyone who helped and contributed!