Emulators provide a reliable, more computationally efficient alternative to globally gridded crop models, and can advance integrated assessment research addressing land-use change.
Regional and global changes in crop yields impact land-use change, with implications for carbon sources and sinks and the energy balance and hydrological feedbacks to the Earth system. To assess the likely impact of climate change on crop yields, researchers typically run a combination of climate and globally-gridded crop models that project how yields of maize, wheat and other key crops will change over time. The suite of models commonly used to simulate crop yields are computationally intensive and produce projections that vary significantly, indicating structural uncertainty. To generate projections that account for wide-ranging modeling uncertainty with far less computational resources, a new toolset of statistical emulators has been developed.
Extending an earlier study focused on maize yields, this research provides a computationally efficient way to represent the impact of climate change on crop yields and land-use change within an integrated assessment model.
Process-based crop models can simulate a wide range of weather and environmental conditions, but are computationally demanding. Statistical models, which are based on observed yield data, are much more efficient, but are hampered by incomplete data sets: crops are only grown under conditions where they do reasonably well most of the time, and hence these models are ill-equipped to estimate the impacts of climate change scenarios well outside the bounds of observation. A third approach is to combine the best of both methods, “training” a statistical model to make reasonably accurate predictions based on the output of a process-based model, but predictions from more than one process-based model must be considered to account for uncertainty in the impact of climate change on crop yields. To that end, Elodie Blanc, a research scientist at the MIT Program for the Science and Policy of Global Change, has trained five simple statistical models to accurately replicate the outcomes of five process-based, globally gridded crop models under diverse climate conditions. Using the statistical models to predict the responses of maize, rice, soybean and wheat yields to climate change-driven variations in temperature and precipitation, Blanc found good agreement between predictions from the process and statistical models. The research, which appears in Agricultural and Forest Meteorology, draws upon a previous collaboration in 2015 with Benjamin Sultan of the University Pierre and Marie Curie in Paris.
BER PM Contact
MIT Joint Program on the Science and Policy of Global Change
The study was funded by the U.S. Department of Energy (DOE) Office of Science under the grant DE-FG02-94ER61937, the U.S. Environmental Protection Agency and other government, industry and foundation sponsors of the MIT Joint Program.
E. Blanc, 2017: Statistical emulators of maize, rice, soybean and wheat yields from global gridded crop models, Agriculture and Forest Meteorology, doi: 10.1016/j.agrformet.2016.12.022.
Photo: Wheat field in Burgundy, France (Source: Myrabella)