This document discusses using climate predictions from global climate models (GCMs) for agricultural impact studies. It presents several crop modeling approaches that can be used with downscaled climate data to quantify the impacts of climate change on crops and design effective adaptation strategies. These include the EcoCrop, MaxEnt, DSSAT and GLAM models. The models vary in their complexity and data requirements, but can provide information on changes in crop suitability, yields and impacts under future climate scenarios. The document advocates using ensemble modeling approaches to decrease uncertainties and inform stakeholders about adaptation needs.
2. The World is changing….
Population growth
Industrial revolution
Non-environmentally friendly
technologies/practices
LEAD TO GREENHOUSE
GASES EMISSIONS INCREASING
5. Changes in climates affect crops we grow...
There will be
winners…
Number of crops with more than 5% gain
…But much
more losers in
developing
countries
Number of crops with more than 5% loss
6. We need models to quantify
impacts and design effective
adaptation options
7. GCMs Statistical Downscaling
MarkSim
Dynamical downscaling:
Regional Climate Model
DSSAT
Statistical
Downscaling GLAM
EcoCrop
Effective
MaxEnt adaptation
options
Bias correction Any model
8. Crop Models
Probability
EcoCrop
Environmental gradient
MaxEnt Models based on crop niches
DSSAT
Models based on processes
GLAM
9. The Model: EcoCrop
– A simple algorithm to look at the broad niche of each species
only based on climate data
– Ten growing parameters to set up the model
• Absolute rainfall interval
• Absolute temperature interval
• Optimum rainfall interval
• Optimum temperature interval
• Length of the growing season
• Crop freezing temperature
– Use of climate data
• Statistical downscaling of GCMs (IPCC4)
• Present-day climates from WorldClim
= Interpolations of observed data, representative of 1950-2000
• 24 different climate models (GCMs) to sample uncertainties
10. The Model: EcoCrop
• So, how does it work?
It evaluates on monthly basis if there
are adequate climatic conditions
within a growing season for
…and calculates the climatic suitability of the
temperature and precipitation…
resulting interaction between rainfall and
temperature…
11. Common Bean Current Suitability
Kiling temperature (°C) 0 Growing season (days) 90
Minimum absolute temperature (°C) 13.55 Minimum absolute rainfall (mm) 200.0
Minimum optimum temperature (°C) 17.45 Minimum optimum rainfall (mm) 362.5
Maximum optimum temperature (°C) 23.05 Maximum optimum rainfall (mm) 449.5
Maximum absolute temperature (°C) 25.63 Maximum absolute rainfall (mm) 710.0
16. MaxEnt
Maximum Entropy Modelling
• Model predicting the potential distribution of a
species
• Statistical dowscaling method apply to climate data.
• Many modellers use the set of the bioclimatic variables
Maxent use the principle of the maximum entropy
Maxent use only presence point of specific species and
environmental variables
• One of the most accurate model for the prediction of
shifts in suitable growth ranges of species
17. MaxEnt Application on Kenyan coffee
Main coffee-producing areas in Kenya are
located in two areas:
- the central region around Mount Kanya
- in the Rift Valley in the west
- The most suitable areas: in the higher areas
of Bungoma, Embu, Kericho, Kiambu, Kirinyaga,
Kisii, Machakos, Meru, Muranga, Nithi, Nyamira,
Nyeri and Trans-Nzoia
New markets
Management
Alternatives
to tea
18. DSSAT
Decision Support System for Agrotechnology Transfer
• Based on crop processes
• Integrate the interaction of weather, soil,
management and genetic factors
• Prediction of yields, plant phenologic stages, plant
weight,harverst date, water soil quantity, N
quantity…
• Current & Future predictions
• Need precise and daily data
19. DSSAT predicts yields
Jones and Thornton, 2003
Maize Yield negatively impacted by CC in most areas in Africa
Need effective adaptation options
20. DSSAT predicts yields
In 2055: Maize Yield
would be negatively
impacted by CC in
most areas in Ethiopia
Need to develop
adaptation strategies
Jones and Thornton, 2003
21. GLAM - General Large Area Model
Challinor et al. (2004)
• Designed at climate model scale to capitalize on
known large-scale relationships between climate and
crop yield, thus avoiding over-parameterization.
Uses grid-scaled
agricultural statistics To simulate yields at
to simulate yields climate model scale
Large-area models are able
to reproduce large-scale
historical yield responses to
climate and inter-annual
variability
Observed peanut yields (kg/ha) Rate of simulated to observed
yields
22. Conclusions
• Different models exist for evaluating the impact of CC on crops
• The application of each model depends on what information we
have and what we want to know:
• Daily/Monthly data
• Crop suitability
• Yield
• Agricultural management practices
• Impact studies at agricultural level benefit from having climate
data of higher resolution
• Inform adaptation to stakeholders: policy makers, donors, other
researchers, but also farmers.
• Gaps: Need more research to understand better crop responses
to climate change To decrease uncertainties
23. Future research plans
• Input climate data quality and its effects on impact predictions
• Analysis of trial data to better understand crop responses to
environment (soil, nutrients, CO2, heat stress, drought stress,
and their interactions)
• Expansion of crop model parameterisation, including multi-Ag-
Model ensembles
• Impacts of future climate change and future variability on crop
yields
• Uncertainty quantification (crop and climate)
• Design of crop genotypic adaptation strategies (“ideotypes”)
and link with analogues to find useful germplasm
• Scale up adaptation strategies to national/ international level
(policy-making)
Existing different change in our world such as population growth, industrial revolution and non-environmentally friendly technologies and practices, which affect the global concentration of greenhouse gases in the atmosphere.
Those change are happening and are affecting crop around the world. We need to know how and how much climate change is going to have an impact on crops to be able to build adaptation strategy and decrease the potential impact of CC on crops and agricultural systems.
Grado de cobertura diff segun modelo. Y resultados tambien yield o suitability. Tambien difieren en escala espacio-temporal a la que se usan.
Remover este slide. Esta info la puedes DECIR en el slide siguiente
Speak about food security problem around the world. Here we can see that food security will be negatively affected by CC in West Afrcia, India with a decrease of crop suitability between 1 and 10 %.
Looking at regional level, for Africa, only sorghum will be positively impacted by CC, the 5 other crops would be negaltively affected. We can think about adaptation strategy such as cropping more sorghum in the future which will not be negatively impacted to CC according to our model.
JRV: puede remover, o convertir en grafico. Esto es demasiado texto. Puede decirlo, no necesidad de escribirlo Statistical dowscaling method apply to climate data to produce 1km resolution surfaces of the monthly mean of max, min temperature and monthly precipitation. Interpolation between centroids of the GCM grid (same to produce WorldClim) and add the predicted climate anomaly for the respective grid cell to the WorldClim data points.
Remover. Esta informacion lo puedes dar en slide siguiente. No necesidad de tenerlo escrito
Recuerda decir: GLAM has been successfully used in various applications in India (groundnut), China (wheat), Brazil (maize), the Sahel (sorghum and groundnut), Nigeria (sorghum, maize, groundnut), and globe (soybean, maize, wheat) GLAM is able to simulate inter-annual variability in crop yield…as well as picking out areas where climate extremes are likely to affect crops and assesing how a change in crop avreity can be used to adapt to these changes. It is designed for use with regional and global climate model output. It is similar to DSSAT with the benefits of empirical models in order to simulate yields.
We need to work to fill the gaps about how crops are responding to CC ; for example how crops are responding to dry period? … We need to understand better all crop processes to different climate to decrease uncertainties in crop modelling.