Study On Irrigation Zoning Method And System Based On Spatial Distribution Of Crop Water Demand

With the continuous growth of population,the continuous improvement of living standards and the aggravation of natural disasters such as drought caused by climate change,water shortage is becoming increasingly serious and has become a major factor restricting the sustainable development of the world economy and society.Agriculture is the largest user of water resources,and irrigation water occupies the main part in the consumption of agricultural water resources.In order to improve the effective utilization rate of agricultural irrigation water,both at home and abroad are actively exploring efficient and water-saving irrigation technology.Zoning irrigation is an important idea in realizing water-saving in agricultural irrigation,which can achieve the purpose of water-saving by implementing different irrigation volumes for different regions.In this paper,uav multi-spectral remote sensing monitoring technology for spatial distribution of crop water demand and farmland irrigation zoning method were studied,and a farmland irrigation zoning system based on WebGIS was established.The main research contents and results are as follows:(1)An unmanned aerial vehicle(UAV)multi-spectral remote sensing monitoring model for spatial distribution of crop water demand was constructed.The penman-Montes formula was used to calculate the crop water requirement,and the model of crop coefficient Kc was established by uav remote sensing,so as to obtain the spatial distribution of crop water requirement.Based on the spectral characteristics of crop canopy,using uav multi-spectral remote sensing data,using unary linear regression,polynomial regression and exponential regression,the relationship models of crop coefficient Kc and 6 planting cover indexes at different growth stages were established.The results show that:The unary linear regression Kc model constructed by using ratio vegetation index(RVI)had the highest accuracy(R~2=0.9332)in the rapid growth period,and the exponential regression Kc model constructed by using normalized differential vegetation index(NDVI)had the highest accuracy(R~2=0.7409)in the middle growth period.The polynomial regression Kc model constructed by using enhanced vegetation index(EVI)has the highest accuracy(R~2=0.7805)in the late growth stage.(2)An image resampling method for spatial distribution of crop water demand based on pixel clustering was proposed.The nearest neighbor method and pixel clustering method were used to ressample the pixels of the spatial distribution map of crop water demand at three ground resolution scales of 1 m,3 m and 5 m,and the coefficient of variation of pixel values was used as the evaluation index of resampling rationality.The results show that the coefficient of variation based on the nearest neighbor method is 20.42%,24.25%and 27.22%respectively,which are all larger than the coefficient of variation of the original pixel values.The coefficients of variation based on pixel clustering are 13.96%,11.92%and13.91%respectively,which are all smaller than the coefficients of original image.The image resampling method can better retain the distribution information of the original crop water demand and is more suitable for resampling the image of crop water demand before the irrigation partition.(3)A crop irrigation zoning method based on K-means algorithm was proposed.The k-means algorithm was used to partition the resampling images based on pixel clustering with cluster numbers of 2,3 and 5,and the results were evaluated and verified by the total deviation of water demand.The results show that under the condition of 1 m resolution,the total water demand deviations of the three clusters are 8.5～18.7 m~3(July 26)and 10.1～20.2m~3(August 12).Under the condition of 3 m resolution,the total water demand deviation of the three clusters was 3.9～14.9 m~3(July 26)and 4.3～16.7 m~3(August 12).Under the condition of 5 m resolution,the total water demand deviation of the three cluster number zones was 3.9～11.3 m~3(July 26)and 3.9～12.2 m~3(August 12).With the decrease of the resampling scale of the spatial distribution map of crop water demand,the deviation between the calculated total farmland water demand and the actual water demand decreases gradually.At three resampling scales of 1 m,3 m and 5 m,the deviation of total farmland water demand gradually decreased with the increase of the number of irrigation zones.When the number of irrigation zones was 5,the deviation range of total farmland water demand was 3.9～10.1 m~3.(4)An irrigation zoning system based on the spatial distribution of crop water demand was developed.SQL Server is selected as the database management software of the system.According to the data requirements,the structure of the system database is designed,and the database including meteorological data,uav multi-spectral image data,field crops and soil information is completed.Arc GIS Server was used to display,scale and overlay the spatial distribution map of crop water demand.C#,CSS,HTML,Java Script and other technologies were used to realize the calculation of crop water demand,ressampling,irrigation zoning and other functions,and developed a farmland irrigation zoning WebGIS system based on the spatial distribution of crop water demand...