Extraction Of Planting Structure And Analysis Of Spatial-temporal Distribution Characteristics In Irrigated Areas Based On Sentinel Data

With the continuous development of social economy and urbanization,people’s demand for crops has increased.In order to meet people’s basic living needs,it is necessary to clarify the distribution of arable land and crop planting structure.Crop planting structure can intuitively reflect the information of crop species,quantity and spatial distribution,provide data support for the adjustment of planting structure and the calculation of irrigation water,and has great significance for improving the comprehensive agricultural production capacity and developing modern agricultural bases.Taking Qingyijiang Irrigation Area in Leshan City of Sichuan Province as the research area,this paper uses maximum likelihood method and decision tree classification method to extract and analyze the crop planting structure of Qingyijiang Irrigation Area by multi-source and multi-temporal Sentinel satellite data,studies the proportion and distribution of crop planting,and compares Sentinel-1 and Sentinel-2 data and Sentin Sentinel-2 data.The accuracy of el-1 and Sentinel-2 bands in crop classification is analyzed,and the extraction accuracy of crop structure is analyzed based on the extracted crop planting area.Finally,the spatial and temporal variation characteristics of crop structure in irrigation area are analyzed based on GIS according to the optimal extraction results.The main conclusions are as follows:(1)The classification accuracy of Sentinel-1 and Sentinel-2 combination bands is higher than Sentinel-1 data and Sentinel-2 data.The average classification accuracy is over 88%,and the average Kappa coefficient is 0.835.This shows that the classification accuracy of Sentinel-2 data is higher than Sentinel-1 data,while the classification accuracy of decision tree classification is higher.The accuracy of the results is higher than that of the maximum likelihood method,which indicates that decision tree is more suitable for crop classification,and the maximum likelihood method has greater error in crop classification.(2)The results of crop classification using Sentinel-1 and Sentinel-2 combination bands are the best,which can reflect the basic situation of crop planting structure in irrigation area.Among spring crops,rice planting area was 11801.01ha,accounting for 47.9%of spring crops;spring vegetable planting area was 3155.11 ha,accounting for 12.8%of spring crops;corn planting area was 3959.57 ha,accounting for 16.1%of spring crops;spring potato planting area was 11265.67ha,accounting for 23.2%of spring crops.In spring crops,rapeseed planting area was 317.86 ha,accounting for 23.1%of spring crops;vegetable planting area was 8179.44 ha,accounting for 60.6%of spring crops;potato planting area was 202.55 ha,accounting for 16.3%of spring crops.(3)The planting area of spring crops is large.Rice accounts for a large proportion of spring crops,followed by spring potatoes and maize,and spring vegetables account for a small proportion of spring crops;rice mainly distributes in the northern,central and southern areas of the irrigation area,spring vegetables and maize mainly distribute in the south of Qingyi River,spring potatoes distribute in bulk in the northern and southern areas of the irrigation area;spring crops have a relatively distinct distribution pattern.The distribution of crops is concentrated.For spring crops,the proportion of Xiaochun vegetable is larger,followed by rapeseed,Xiaochun potato is less,and the distribution is more discrete;rapeseed is relatively centralized planting,distributed in different size planting areas;Xiaochun vegetable is mainly distributed on both sides of Qingyi River,scattered in other places;Xiaochun potato is mainly distributed in the southern part of the irrigation area,more in the southwest mountain area.Figure 31 table 16...