Rice yield estimation based on MODIS EVI and measured data derived from statistical sampling plots at province level

Rice yield estimation by remote sensing is very important for making agricultural decisions and food security in China and worldwide. Based on the feasibility analysis of combination of MOD09GA and MYD09GA EVI(enhanced vegetation index), provincial unit rice yield was estimated using rice standard yield per area from statistical sampling survey plots. The results showed that more than 50%, 85% and 95% absolute difference between MOD09GA and MYD09GA EVI for rice sampling plots during rice growth period was less than 0.03, 0.08 and 0.1, respectively. The accuracies of estimating rice yield derived from quadratic nonlinear models and regression models were higher than those of other kind of models. The fitting and forecasting unit rice yield had a relative error of 2% and 5%, respectively. The scatter diagrams showed that the estimating paddy rice yield was closed to the observed data, which indicated that the estimating provincial unit rice yield based on MODIS EVI and measured data derived from statistical sampling plots was consistent with the statistical results.