Assessment Of Rice Water Productivity Based On MODIS Data In Bielahong River Irrigation Area Of Heilongjiang Province

Food and water resources are important factors related to human survival and social stability.Rice is one of the most important food products in my country,and its planting and production process consumes a lot of water.However,at present,the global water resources are scarce.With the development of the economy,the competition for water use in various industries is fierce.Therefore,improving crop water productivity can effectively alleviate the inter-industry conflict.Water conflict.In this context,this paper takes the Bielahong River Irrigation District in Heilongjiang Province as the research area,and uses the MODIS remote sensing data in 2019 and the measured yield data of 83 sampling points as the main data sources to draw a spatial distribution map of water productivity in the study area in 2019,and draws the spatial distribution map of water productivity in the study area in 2019.The water-saving potential of the study area was analyzed with water productivity as an index,which provided a theoretical reference for rational regulation of agricultural water resources and food safety production.The main results of this study are as follows:(1)Rice planting area and its spatial distribution.The NDVI,EVI,LSWI and RVI time series data sets of the study area were constructed using MOD09A1 data,and the SG filtering of Timesat software was used to denoise them,and four main ground objects(rice fields,forest land,wasteland,and building land)were selected in the study area.Several sampling points are used as training sample areas,and the time series curves of 4 remote sensing indices of 4 ground objects are extracted.According to the different characteristics of ground objects in different periods on the remote sensing index,appropriate thresholds are selected to establish a decision tree model for rice planting area.The rice planting area in the study area was obtained,and the spatial distribution map of the rice planting area was obtained.On this basis,the land use classification provided by the high-resolution satellite combined with the confusion matrix method is used to evaluate the accuracy of the extraction results.The results show that the Kappa coefficient reaches 0.71,which is highly consistent,and the total classification accuracy is 88.75%,the classification effect is also better.(2)Rice yield estimation and its spatial distribution.The correlation coefficient analysis(r)method,the variable projection importance analysis(VIP)method and the out-of-bag data importance(OOB)analysis method were used to analyze the four growth stages(tillering stage,heading stage,booting stage and milk maturity stage)of rice.The sensitivity of different bands and spectral indices to rice yield,the characteristic bands and indices were screened out,and then combined with three statistical regression methods,random forest(RF),support vector machine(SVM)and partial least squares(PLS),9 were constructed.Rice yield estimation models:r-RF,r-SVM,r-PLS,VIP-RF,VIP-SVM,VIP-PLS,OOB-RF,OOB-SVM,OOB-PLS.The results show that the same index screening method has different degrees of fit for different models,OOB and RF fit better,VIP and r fit better with PLS,and r fit better with SVM;in addition,the OOB-RF model is the most fit among the 9 models.Excellent,the model validation coefficient of determination is0.742,and the root mean square error is 206 kg/hm~2.Finally,the OOB-RF model was used to estimate the rice yield in the study area pixel by pixel,and the spatial distribution map of rice yield in the study area was drawn.(3)Study on water productivity and water saving potential of rice.Using the MOD16A2 data combined with the spatial distribution map of rice planting area,the spatial distribution map of rice field evapotranspiration in the study area was drawn,and then the spatial distribution map of rice water productivity was drawn using the spatial distribution map of rice yield.By analyzing the sensitivity of water productivity to yield and evapotranspiration,it was found that the decrease of rice evapotranspiration was most sensitive to rice water productivity.Then,the water-saving potential of the study area was calculated and analyzed using the water productivity as an index,and it was found that the study area had a large water-saving potential.Therefore,water-saving measures to reduce evapotranspiration were proposed to improve the water productivity of the study area,so as to achieve sustainable agricultural water resources.purpose of development.In this study,water productivity in The Bielahong River irrigation area was studied,which provided theoretical significance for the sustainable utilization of agricultural water resources and the analysis of water-saving potential of crops in China.