Study On Assimilation And Application Of WOFOST Crop Growth Model Based On Remote Sensing Information

Grain yield security is closely related to the national economy and people’s livelihood.Crop yield estimation information plays a key role in the macro-control of the national economy,providing an important basis for optimizing intelligent decisionmaking.With the development of starry and multi-layer agricultural technology,remote sensing agricultural information,crop yield estimation information and intelligent irrigation decision-making play an important role in agricultural management and food security.Therefore,this study relies on maize growth model WOFOST,and utilizes the advantages of remote sensing observation technology to construct a maize yield prediction and evaluation model,and applies it to the formulation of irrigation system in actual production.This study took spring corn as the research object and WOFOST crop growth and development model as the basis.On this basis,localization of WOFOST crop growth model was studied,and parameter localization experiments were carried out by using field investigation and measurement data in Tuzuo Banner experimental area,literature data related to spring corn planting in Tuzuo Banner,and NASA Power open meteorological data.At the same time,the experimental data of landsat-8 remote sensing images in the study area were obtained,and the data were pre-processed by radiometric calibration and atmospheric correction.The inversion of area index LAI and soil water SM during the middle period of corn growth was conducted based on prosail model lookup table method and SCWI model respectively.By studying the collected remote sensing data and sensor data,the leaf area index(LAI)and soil water content(SM)in the growth process of maize were obtained.The simulated data of the model were corrected through the integrated Kalman filter ENKF and the four-dimensional variational 4DVAR algorithm.Finally,the hydrology year of Tuzuo Banner was divided by Pearson III frequency distribution function,and the effective daily rainfall of the growth period was calculated.The daily water requirement of the whole growth stage was obtained by Penman-Monteith method and crop coefficient method,so as to obtain the daily net irrigation amount of spring corn in the study area.Based on the WOFOST-ENKF model proposed in this thesis,the irrigation simulation of representative sites is carried out.The main conclusions are as follows:(1)Through experimental comparison,the RMSE and RRMSE of the Wofo STEn KF assimilation model based on remote sensing and the Wofo ST-4DVAR assimilation model based on remote sensing are lower than that of the WOFOST model.WOFOSTENKF model has better prediction accuracy than WOFOST-4DVAR model.(2)The WOFOST-ENKF assimilation model based on remote sensing proposed in this thesis is feasible for the simulation of spring corn irrigation scheme in the study area.The irrigation simulation results of primary and secondary key irrigation growth periods are applied to yield prediction and water use efficiency evaluation.The results show that the irrigation simulation of secondary key irrigation growth periods is the optimal irrigation scheme.