The Variation of Rice Models in Simulating Responses to Elevation of CO2 and Temperature.

The impacts of elevating CO₂ and temperature on rice production must be quantified for designing appropriate adaptation and mitigation strategies to improve rice production under climate change. Crop models are the tools to quantitatively evaluate the impacts. However, the quantification of the impacts may not be consistent with each other among crop models. We evaluated 13 rice crop models against multi-year field experimental data at four sentinel sites (representing different rice growing environmental conditions in Asia), and data from FACE (Free-Air CO₂ Enrichment) experiments at two sites. Predicted yield decreased with increasing temperature and increased with increasing [CO₂] at all sites, but the magnitude of predicted changes varied among the models. The responses to a temperature increase were site-specific, but were little influenced by atmospheric [CO₂]. Yield responses to increases in [CO₂] were more similar across sites than yield responses to increases in temperature. Averaged across all crop models, the yield increases due to elevated [CO₂] were similar to those observed in the rice FACE experiments, and had a diminishing return from increasing CO₂ as it was enriched up to double concentration. However, the interactions of CO₂ and temperature were not represented well by the models. Using the FACE data obtained from China and Japan, We are currently determining the sources of variation in the predicted responses among the crop models.