Font Size: a A A

Remote Sensing Monitoring Of Temporal And Spatial Changes Of Crops In Guigang City

Posted on:2021-02-22Degree:MasterType:Thesis
Country:ChinaCandidate:Q LiangFull Text:PDF
GTID:2393330605966450Subject:Cartography and Geographic Information System
Abstract/Summary:PDF Full Text Request
Guangxi is a structural grain shortage area.Although rice production can meet the population demand,corn,beans and other crops are relatively scarce.Guigang City has the largest alluvial plain in Guangxi-Xunyu plain.Good geographical location and climatic conditions have created this fertile land.The superior planting environment makes Guigang become one of the key areas to promote grain production and efficiency.The analysis of the changes of planting area,planting structure and influencing factors in Guigang is not only the way to master grain production accurately,but also the key to optimize grain production.In order to understand the distribution of crops in Guigang City,the spatial-temporal data fusion model was used to HJ?NDVI and MODIS?NDVI data is fused to obtain high spatial-temporal resolution NDVI fusion data.On this basis,the high spatial-temporal resolution NDVI data set is fitted and reconstructed and phenological parameters are extracted by combining with TIMESAT software.In the En MAP-Box plug-in,the random forest classification method is used to extract agricultural information.Through analyzing the spatial distribution characteristics of crops,the reasons that affect the change of crop distribution are found,Based on this,the index system of crop impact assessment is constructed,and the influence of impact factors on the spatial distribution of crops is calculated.The research is divided into five parts.The first part is to preprocess MODIS data and environmental satellite data to obtain NDVI data,and to use the fusion model of ESTARFM to HJ?NDVI and MODIS?NDVI data sets with high spatial-temporal resolution are obtained by space-time fusion.In the second part,the high spatial-temporal resolution NDVI data set is reconstructed and the phenological parameters are extracted in the software TIMESAT.In the third part,with the help of En MAP-Box plug-in,the random forest classification method is used to extract crop planting information and verify the accuracy.The fourth part is the analysis of temporal and spatial changes of crops.In the analysis of temporal and spatial changes of crop quantity,the total area of crops and the area of crops in districts and counties are analyzed,and the geographical and spatial changes and distribution characteristics of crop planting area are studied from the village level and the county level respectively.The fifth part is the analysis of the driving force of crop planting change.The causes of the spatiotemporal changes of crops are classified into three categories: natural factors,social factors and economic factors,from which the evaluation index system of crop impact is constructed.The main driving factors that affect the distribution of crops are obtained through the least square linear regression model and the geological detector.The results show that:(1)In this paper,the high spatial-temporal resolution NDVI data set can be obtained by the fusion model of ESTARFM,and the S-G fitting and reconstruction algorithm is used in the software of TIMESAT to process the fitting and reconstruction of the high spatial-temporal resolution NDVI data set and calculate the phenological characteristics,so as to obtain the results and phenological characteristics information of the NDVI time series reconstruction of Guigang City.(2)In this study,the phenological characteristics of NDVI data sets with high spatial-temporal resolution in 2009,2013 and 2018 were analyzed with the help of TIMESAT software,and 11 phenological characteristics were analyzed.It is found that the data of three phenological indexes in 2009 are scattered: the beginning of growth season,the length of growth season and the small score of growth season.In 2013 and 2018,the data of four phenological indexes are scattered: the beginning of growth season,the end of growth season,the length of growth season and the middle of growth season,which can be combined with the scattered phenological indexes for crop classification.(3)In this study,the random forest classification algorithm was used to extract the crops in Guigang City in 2009,2013 and 2018.The total area of crops in Guigang City in 2009 was 437800 hectares,the total area of crops in 2013 was 458000 hectares,and the total area of crops in 2018 was 476800 hectares.(4)The influencing factors of crop planting can be summarized as natural factors,social factors and economic factors.Natural factors are the basis of crop planting,good landform environment and abundant sunshine and water are conducive to the growth of crops,social factors affect the planting trend of crops through market orientation,guidance of agricultural related policies,supporting and improvement of agricultural infrastructure and investment of agricultural materials,and economic factors affect the growth of crops from the price and income level of agricultural products Development direction and planting target of agriculture.(5)The distance to the railway,the average gradient and the per capita disposable income of urban residents have great influence on the spatial heterogeneity of crops.Traffic conditions can speed up the transportation of crop planting materials and improve the circulation speed of agricultural products.Crops can be planted in areas with a slope of less than 25 ° and agricultural activities cannot be carried out in areas with a slope of more than 25 °.The income of urban residents affects the sales of agricultural products.The higher the income,the greater the demand for agricultural products,the greater the planting intensity of crops,and vice versa.(6)The crops in Guigang City are mainly distributed in the middle of the plain,on both sides of the river,and gradually reduce to the mountain areas on both sides,while the crops in the mountainous and hilly areas are less scattered.Combined with the growing conditions and environment of crops,rice and sugarcane can be planted more intensively in areas with flat terrain and contiguous plots,and vegetables can be planted more intensively in suburban and peripheral areas of cities and towns.Corn,cassava and other dry land crops can be planted in areas close to mountains and hills and far away from water sources.Peanut crops are preferentially planted in the area with a large proportion of sandy soil in the middle of Guigang City.
Keywords/Search Tags:Crop extraction, ESTARFM Algorithms, Time series of NDVI, phenological parameters, Random forest
PDF Full Text Request
Related items