Rice yield estimation with satellite data

Information on rice yield is thus critical for agronomic planners to estimate rice production in respect to successful strategies to address food security and rice grain export issues. This study aimed at modelling rice yields based on rice crop phenology with Moderate Resolution Imaging Spectroradiometer (MODIS) satellite data in the Mekong River Delta, Vietnam. The data processed for 2011 comprised of three main steps: (1) constructing the filtered time-series enhanced vegetation index (EVI) data, (2) model rice crop yields, and (3) verify the yield model. The comparison results between the satellite-based estimated yields and government's yield statistics indicated a significant relationship between the two datasets (p-value < 0.001) with the correlation coefficients (R²) of 0.66 to 0.46 for spring-winter and summer-autumn rice crops, respectively. The RMSE and MAE values for winter-spring and summer-autumn crops were lower than 10%, indicating the consistency between estimated yields and statistics.