Using seasonal climate forecasts to predict rice yield over Nepal's Terai

Skillful seasonal climate prediction combined with the dynamic crop model can help to optimize management and predict yield variability associated with interannual climate variability. This study’s goal was to examine skill of the state-of-the-art seasonal prediction systems (SPSs) for crop yield forecasting and management optimization. The Climate Forecast System Version 2 (CFS v2) and the Centro Euro-Mediterraneo sui Cambiamenti Climatici Version 1.5 (CMCC v1.5) SPSs’ hindcasts were examined. We used CFS v2’s daily output in the crop environment resource synthesis (CERES) Rice crop model of the Decision Support System for Agro-technology Transfer (DSSAT) v4.6 to simulate rice yield. We tested various options to improve predictability and found that the potential of CFS v2’s daily output for agricultural applications is limited by their skill. We also examined potential application of El Niño Southern Oscillation (ENSO) based forecasts and found that they have potential to increase yield and minimize N leaching. Before generalizing, these conclusions must be verified by the research station trials.