Production Of Agricultural Forecast Based On WOFOST Model And Remote Sensing Data

Remote sensing technology emerged in the 1960s.With the development of the times,remote sensing technology is also gradually developed and improved.The data obtained by satellites is also constantly updated.Today’s remote sensing satellite data has high accuracy.This high-precision data enables remote sensing technology to blossom in the fields of agriculture,waters,environment,resource protection,hydrology,climate,geology,etc.Further promote the development of our country.Food is an important resource of a country.Throughout our history,our country has been a major agricultural country since ancient times.The southern part of our country has the Yangtze River,one of the three major river basins in the world.The unique geographical conditions are a good foundation for the country’s grain production.Although my country’s grain output is high,the grain problem is still an important issue in my country due to its large population.Therefore,the forecast of grain output is very important.Accurate prediction of grain output is of great significance to my country’s grain security,structural adjustment of crop planting,and formulation of grain import and export trade countermeasures.For now,the remote sensing technology is mainly used for large-scale regional production estimation.Remote sensing satellite data is easily affected by the climate environment,and the quality of image data is affected by clouds and other factors,making most data unusable.Therefore,most of the current yield estimation methods are based on single-period remote sensing data.Single-period remote sensing data cannot fully reflect the growth status of crops,and yield estimation methods have certain limitations.This article is based on remote sensing data and WOFOST model(vegetation growth model)to estimate the yield of rapeseed in Hubei Province for multiple periods.The content is as follows:(1)The important basis of rape yield estimation is the establishment of yield model.In order to obtain a more accurate yield model,this experiment analyzed and modeled the experimental data of the crop experimental base of Huazhong Agricultural University,and obtained the relationship model between rape Lai and yield.(2)There is a strong dependence on the change of Lai in the growth process of rape,but it is difficult to obtain the temporal Lai(leaf area index)data of rape.In this experiment,WOFOST model combined with daily meteorological data assimilation simulation is used to obtain Lai in the process of rape growth.The simulated Lai time series is fitted and compared with MODIS Lai to ensure that the simulated Lai time series curve is consistent with the change curve of MODIS Lai time series curve.Assimilating the model through remote sensing data can make the simulation results of rape consistent with the actual rape growth data.(3)The yield estimation of large-scale region can be realized ultimately by relying on high-precision remote sensing data.The realization of large-scale regional yield estimation is based on the yield estimation model and the correlation model between remote sensing data and crop growth state.The remote sensing data used in this experiment is gaogao-1 satellite data.Through the combination of multi period high score data and multi period Lai modeling,the GF-1 data estimation rape Lai model is obtained.(4)The advance prediction of rape Lai is conducive to the prediction of crop yield,the adjustment of rape planting strategy and the improvement of production.In order to predict the Lai of rape in advance,we use the long and short memory neural network(LSTM)model to simulate the Lai time series curve,and then predict the Lai20 days after a certain period of time.The Lai obtained in advance can further predict the rape yield in advance.