Forecasting rice crop yields with time-series MODIS EVI data: A case study from Mekong delta, Vietnam

The Mekong Delta known as the rice bowl of Vietnam annually produces approximately half of the country's rice production. This region plays an important role in the country's economy for food supply and rice export. The impacts of climate change have however caused declined rice production. Forecasting rice crop yields is thus an important activity to provide agricultural planners with useful information so that they can devise better strategies to ensure food security and rice exports for the country. The main objective of this study was to develop a model for estimating rice crop yields in the study area using enhanced vegetation index (EVI) derived from the multi-temporal Moderate Resolution Imaging Spectroradiometer (MODIS) data. The data were processed for 2006-2007 through four main steps: (1) constructing time-series EVI data, (2) noise filtering of the time-series EVI data using the wavelet transform, (3) developing a rice-crop yield model, and (4) result verification using the root mean square error (RMSE) and standard error of prediction (SEP). The predicted results compared with the rice yield statistics indicated that the model produced the close agreement between predicted and yield statistics. This study demonstrates the validity of the developed model using EVI for forecasting rice crop yields in the Mekong Delta prior to the harvesting period. Copyright