Spatialized Application of Remotely Sensed Data Assimilation Methods for Farmland Drought Monitoring Using Two Different Crop Models

Dragon 3 Final Results and Dragon 4 Kick-Off, Wuhan, China, 4-8 July 2016The aim of this work was to develop a tool to evaluate the effect of water stress on yield losses at the farmland and regional scale, by assimilating remotely sensed biophysical variables into crop growth models. Biophysical variables were retrieved from HJ1A, HJ1B and Landsat 8 images, using an algorithm based on the training of artificial neural networks on PROSAIL. For the assimilation, two crop models of differing degree of complexity were used: Aquacrop and SAFY. For Aquacrop, an optimization procedure to reduce the difference between the remotely sensed and simulated CC was developed. For the modified version of SAFY, the assimilation procedure was based on the Ensemble Kalman Filter. These procedures were tested in a spatialized application, by using data collected in the rural area of Yangling (Shaanxi Province) between 2013 and 2015. Results were validated by utilizing yield data both from ground measurements and statistical survey