Introducing the AgMIP Project and improving crop models for response to climate factors

IRRI Seminar

K.J. Boote
University of Florida

1315-1415 H, Thursday, 6 December 2012
Havener Auditorium, IRRI

Abstract:

The Agricultural Model Intercomparison and Improvement Project (AgMIP) is a distributed climate-scenario simulation exercise for intercomparison of historical models and future climate change conditions. It involves the participation of modeling groups for multiple crops and world agricultural trade around the world.

The goals of AgMIP are to substantially improve the characterization of risk of hunger and world food security due to climate change and to enhance adaptation capacity in both developing and developed countries.

AgMIP has leadership teams in four areas: climate, crop modeling, agricultural economics, and IT. There is strong emphasis on using a multiple modeling approach, starting with multiple climate models, continuing with multiple crop models and multiple economic models to evaluate the uncertainty posed by and the impact of climate change on agricultural production and food security, both within regions and across the world.

The Crop Modeling Team is led by Ken Boote and Peter Thorburn and is responsible for coordinating the evaluation, intercomparison, and improvement of crop models prior to using the crop models for climate change assessment and adaptation. These efforts include comparing multiple crop models for their response to climate change factors to see how the models differ, and testing of crop models against field and controlled-environment chamber data on response to CO2, temperature and other factors.

Crop models can be tested at the level of processes (photosynthesis, transpiration, etc.) as well as at in-season and end-of-season biomass and yield. To verify that crop models are robust and accurate, it is important that crop models be compared to experimental data from field studies as well as data from controlled-environment studies that cover a range of factors (CO2, temperature, solar radiation, drought, and management).

Excellent experimental data and metadata on various crops are now available, and examples of metadata for CO2 response and temperature response will be discussed for rice, sorghum, soybean, peanut, and bean.

Rice is somewhat less CO2-responsive than soybean and peanut, and shows more down-regulation of leaf N concentration at elevated CO2. Rice and sorghum appear to respond similarly to changes in temperature, with optimum temperature for grain yield being at a mean of 25°C, with progressive reduction at higher temperature to a point of failure at a mean of 35°C (corresponding to Tmax/Tmin of 40/30°C).

On the other hand, soybean and peanut, while having a similar optimum, have a higher failure temperature mean of 40°C. Model comparisons of elevated temperature data will be discussed for soybean and peanut, using the CSM-CROPGRO model.

More testing and improvement of crop(?) models against CO2 and temperature data is needed. Unfortunately, climate impact modelers are running ahead with crop models that have not been tested well or at least improved. More emphasis and resources are needed on crop model improvement. IRRI’s leadership in rice multi-model intercomparison and improvement is commendable.