Use of CERES-Wheat model for wheat yield forecast in central Indo-Gangetic Plains of India

• DOI: 10.1017/s0021859604004022
• Publication Date: 2004-07-12T09:51:15Z
• Authors: A. S. NAIN; V. K. DADHWAL; T. P. SINGH
• Keywords: Uttar pradesh; Winter wheat; Yield gap

A methodology was developed for large area yield forecast using a crop simulation model and discrete technology trend, applied to the coherent wheat variability zones of Eastern Uttar Pradesh, India. The approach consisted three major steps: (a) prediction trend historical series region; (b) weather-induced deviation in CERES-Wheat relating simulated observed deviations from trend; (c) final by incorporating predicted yield. regression coefficients step were generated 10 years' data (1984/85–1994/95) reliability tested on set 5 independent (1995/96–1999/2000). results showed that this could capture year-to-year with reasonable accuracy. Root Mean Square Error (RMSE) between reported as 0·098 t/ha mean 2·072 (4·72%). However, RMSE slightly higher forecasting period comparison calibration period. use issuing pre-harvest effect upgrading also studied. forecasts made in-season weather up end February climatic-normal rest wheat-growing season, which good agreement yields. season 1999/2000 1994/95 updated 1998/99, fell latter case, 4·10 2·50%.