Quantifying the GCM-related uncertainty for climate change impact assessment of rainfed rice production in Cambodia by a combined hydrologic - rice growth model

Abstract

The effects of climate change on agriculture are a major concern for global food security. In this study, the impacts of climate change on rainfed rice production in the granary of Cambodia were examined on a basin scale by developing and applying a combined model consisting of a crop model and a basin-scale distributed hydrological model. The response of rice production to soil-water availability was simulated for past (1981–2000) and future (2041–2060, 2081–2100) periods. From 34 general circulation models (GCMs) that participated in the Coupled Model Intercomparison Project Phase 5 (CMIP5), 5 GCMs were selected by evaluating monthly rainfall in the past. Although annual rainfall was projected to increase by all five selected GCMs, notable decreases in rainfed rice production were projected with 3 GCMs, while small changes were projected with the other 2 GCMs. The main factor restricting future rice production was soil water availability, brought by the projected change in the seasonal distribution of rainfall and the projected more severe dry spells in the early monsoon season. The results suggest the importance of the selection and bias correction of GCMs to force rice crop models and of the simulation of soil water flow on a basin scale for the assessment of rain-fed rice production. In particular, improvements in projections of rainfall amounts over shorter periods rather than annual or seasonal periods, which fit within the time scales of rice plant growth, were suggested to be important.