Abstract
Waterlogging constrains crop yields in many regions around the world. Despite this, key drivers of crop sensitivity to waterlogging have received little attention. Here, we compare the ability of the SWAGMAN Destiny and CERES models in simulating soil aeration index, a variable contemporaneously used to compute three distinct waterlogging indices, denoted hereafter as WI (Destiny), WI(ASD1), and WI(ASD2). We then account for effects of crop growth stage and soil temperature on waterlogging impact by introducing waterlogging severity indices, WI (Growth), which accommodates growth stage tolerance, and WI (Plus), which accounts for both soil temperature and growth stage. We evaluate these indices using data collected in pot experiments with genotypes "Yang mai 11" and "Zheng mai 7698" that were exposed to both single and double waterlogging events. We found that WI (Plus) exhibited the highest correlation with yield (-0.82 to -0.86) suggesting that waterlogging indices which integrate effects of temperature and growth stage may improve projections of yield penalty elicited by waterlogging. Importantly, WI (Plus) not only allows insight into physiological determinants, but also lends itself to remote computation through satellite imagery. As such, this index holds promise in scalable monitoring and forecasting of crop waterlogging.