Research on remote sensing information and WheatSM model-based winter wheat yield estimation

The research was implemented in a planted area dominated by winter wheat in the Hebi city of Henan Province driving method and EnKF assimilation method were used to coupling the optimized and reconstructed MODIS LAI time series data and the rigorously calibrated wheat growth model WheatSM(Wheat Growth and Development Simulation Model) developed according to the planting varieties of the winter wheat region in North China respectively, the research on winter wheat yield estimation was conducted at field scale and regional scale from 2013 to 2016. The results show that : (1) The coupling simulation precision was inferior to the pre-coupled simulation precision at field scale simulation.(2) When the grid was used for regional scale simulation, the simulation accuracy of 1 km resolution grid point is higher than 5km resolution.(3) At two kinds of grid resolutions, EnKF assimilation simulation has the highest accuracy, followed by the driving method, and the simulation precision before coupling is the lowest. The RMSE of the total output of the EnKF assimilation simulation area and the actual total output at the 1 km resolution grid is 15.1. Assimilation of remote sensing information at fine grid resolution can improve the precision of regionalization application of the WheatSM model.