Late Freezing Injury Monitoring And Yield Prediction Of Winter Wheat In Henan Province Based On Internet Of Things And Remote Sensing

Winter wheat is Chinese main food crop,and its production is crucial to Chinese food security.Henan Province is one of the core areas of grain production in China,its production accounts for more than 25%of total wheat production in the country.In recent years,late frost damage frequently occurs,which has a great impact on winter wheat yield and quality,even leads to the loss of yeild.Therefore,it is important to monitor and predict the winter frost damage of winter wheat in Henan Province.Based on the Internet of Things technology and remote sensing technology,this paper selects the appropriate spatial interpolation method to construct the temperature surface of Henan Province.Based on the constructed temperature surface,it monitors whether winter wheat damage occurs in winter wheat in Henan Province,and combines extreme low temperature(FDD)to quantify the winter frost damage of winter wheat,to the extent that it constructs the winter wheat night frost damage index(F).Based on the EDD(extreme degree-day),GDD(growth degree-day),winter wheat night frost damage index(F)and NDVI(normalized difference vegetation index),the regression model was constructed to produce winter wheat yield.It is predicted that winter wheat frost damage monitoring and yield prediction will be realized,which will provide a new means for intelligent decision support of winter wheat production.The main research and results of this paper are as follows:(1)Using the IoT(Internet of Things)sites deployed in Henan Province and the weather data of the weather forecast,spatial interpolation of the temperature in Henan Province is carried out by combining four interpolation methods:Kriging interpolation,pan Kriging interpolation,spline interpolation and inverse distance weight interpolation.Comparing the spatial interpolation methods of various regions in Henan Province,it is found that the interpolation results of the spline interpolation method on any terrain are not optimal;the Kriging interpolation method is better than the other three methods on the hilly terrain;the Pan Kriging interpolation method achieved better interpolation results on the basin topography;the anti-distance weight method has higher interpolation accuracy in the transition zone between the plain,the mountainous area and the basin to the plain than the other three methods.According to the main topography of each region,the appropriate spatial interpolation method is selected to establish the temperature surface of Henan Province,and finally the real-time monitoring of the temperature in Henan Province is realized.(2)The greening period and jointing period of winter Wheat were calculated by the temperature of each hour at 24 hours a day.Compared with the actual time,it was found that the error of the greening period and the jointing stage calculated by using the 24-hour temperature was smaller than the previous method.The growth degree-day(GDD)and extreme degree-day(EDD)of winter wheat were calculated by combining the data of minimum temperature and maximum temperature.The concept of extreme low temperature(FDD)of winter wheat was optimized by EDD concept,and then the index F of winter wheat frost damage was determined.The winter wheat frost damage index and spatial interpolation method were finally determined.In the monitoring of winter wheat night frost damage,the results of winter wheat growth period and combined temperature were used to judge whether winter wheat had late frost damage.(3)Winter wheat was predicted by using three regression models for winter wheat yield,including sum of growth degree-day(SGDD),sum of extreme degree-day(SEDD),normalized difference vegetation index(NDVI)and night frost damage index(F).The experimental results show that the regression model with SGDD,SEDD and NDVI is more suitable than the two models with SGDD and SEDD,which proves that NDVI is a sensitive factor for yield prediction.Taking Shangqiu City as an example,the four dependent variables of SGDD,SEDD,NDVI and F were used to predict the yield of winter wheat,the results showed that the model of the late frost injury index F was added to the three production prediction models in Shangqiu City.