Improving Simulations of Rice in Response to Temperature and CO2

Authors: LI, SANAI - Us Forest Service (FS); Fleisher, David; Timlin, Dennis; Barnaby, Jinyoung; SUN, WENGUANG - University Of Nebraska; WANG, ZHUANGJI - University Of Maryland; Reddy, Vangimalla

Submitted to: Agronomy Journal
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 11/21/2022
Publication Date: 11/23/2022
Citation: Li, S., Fleisher, D.H., Timlin, D.J., Barnaby, J.Y., Sun, W., Wang, Z., Reddy, V. 2022. Improving Simulations of Rice in Response to Temperature and CO2. Agronomy Journal. 12(2):2927. https://doi.org/10.3390/agronomy12122927.
DOI: https://doi.org/10.3390/agronomy12122927

Interpretive Summary: Rising air temperatures are outpacing positive effects of atmospheric carbon dioxide concentration on crops such as rice. As a result, rice growers are concerned about their yields in many rice production regions throughout the world. Mathematical models called 'crop' models are frequently used to study current and future effects of these climate factors on rice in order to help farmers adapt. However, improvements to these models still need to be made and tested to improve confidence in their predictions. In this study, we modified a popular rice model with equations that improve predictions of heat stress and crop growth rate to temperature and carbon dioxide. More than twenty sets of data were used to evaluate these changes. We found that these improvements improved accuracy of the model by as much as 40% over the original version. These changes to the model can be easily implemented by other modeling groups to improve their own decision support tools. Given the importance of rice as a main food source for almost half the world's population, these improvements are highly relevant for global food security research and help improve confidence when studying ways to adapt rice production systems to a warming environment.

Technical Abstract: Crop models are frequently used to assess impact of climate change responses, such as rising air temperature (T) and atmospheric carbon dioxide concentration (CO2) on yield and developmental responses. However, evaluation of model performance against empirical data is crucial to establish confidence in such predictions, particularly for rice (Oryza sativa L.), one of the world's most important cereal crops. In this study, over twenty sets of soil-plant-atmosphere-research (SPAR) growth chamber data covering responses of three cultivars to a range of T and two CO2 levels were used to evaluate the ORYZA rice model. Two versions were evaluated, the original form of the model (ORYZA-or) and a previously modified version that used a coupled leaf-level gas exchange algorithm with an energy balance (ORYZA-gas) to simulate photosynthesis and transpiration rates. A modified heat stress method to account for spikelet sterility was also evaluated within both model versions. The ORYZA-gas model consistently exhibited better metrics for above-ground biomass, grain yield, and canopy photosynthetic rate across all datasets and cultivars. For these responses, root mean square error (RMSE) improved by about 1300 and 200 kg ha-1 and 3 mol CO2 m-2 season-1 between model versions. While the gas exchange modification was associated with improved accuracy for most results, the heat sterility modification was primarily responsible for improving grain yield, with a reduced RMSE by about 700 kg ha-1 for either model version. These results indicate the importance of improving both heat sterility functions as well as carbon assimilation methodology in rice models that incorporate direct responses to air temperature and/or CO2 concentration. Accounting for cultivar differences in heat sterility will be an important component to incorporate in tools for rice climate assessments as well.