Rubber Forest Identification And Suitability Assessment Of Rubber-Puzzle Plantation Land Based On Multi-source Remote Sensing Data

Natural rubber(Hevea brasiliensis)is an important industrial raw material and strategic material in China,and rubber trees are the only source of raw materials for natural rubber production.In recent years,as rubber prices continue to be low,the existing rubber plantations are unstable due to the lack of scientific management causing many problems such as idle waste of space under rubber gardens,and the single rubber tree cultivation model has been seriously impacted,and the only way to effectively stabilize the rubber industry is to increase land production value by intercropping cash crops under rubber forests.Scientific evaluation of the suitability of rubber intercropping compound planting is the fundamental guarantee for the development of rubber undergrowth intercropping economy,meanwhile,rubber forests are mainly distributed in tropical alpine cloudy areas,and the precise identification of their spatial distribution also determines the accuracy of rubber intercropping compound planting suitability evaluation.To summarize,in order to grasp the spatial distribution pattern of rubber forest more scientifically and comprehensively and to use the vast space under rubber forest more reasonably and effectively,this study selects Baisha County of Hainan Province,a typical rubber plantation county with complex topography and cloudy fog,as the study area,and takes the rubber-puzzle(Alpinia oxyphylla)intercropping system,a typical mature rubber-medicine complex plantation model,as the study Based on multi-source remote sensing data and Google Earth(GE)high-definition historical images and other image data,different methods of comparative analysis were adopted using ENVI and other software platforms to finally extract the precise distribution of the current rubber plantation in Baisha County.On this basis,five major categories of element indicators,such as geographic location,topography,climate,soil properties and soil nutrients,were selected to establish a rubber-puzzle compound plantation land suitability evaluation index system by using hierarchical analysis method,and the suitability evaluation study of rubber-puzzle plantation land was carried out by combining GIS technology,and the main research results are as follows.(1)In this study,LANDSAT 8 OLI remote sensing images of Baisha County from March to August 2019 were selected based on rubber forest phenological characteristics,and three supervised classification methods were selected: minimum distance classification method,neural network classification method and maximum likelihood method,and the addition of Google Earth image data with high-definition resolution made the selection of sample data more abundant and convenient,and after the comparative analysis of image classification results,it was determined to use The neural network classification + GE image recognition has the highest accuracy,with the overall accuracy and Kappa coefficient reaching 92.17%and 90.39%,respectively.The area of the current rubber plantation in 2019 in Baisha County was obtained as 70,149.14 ha after removing small patches,which provides scientific guidance for the identification and extraction of rubber forest area in southern China,especially in the mountainous region of Hainan.(2)The typical "southern medicine" crop puzzle in rubber intercropping system was selected as the compound system in this study,and 17 evaluation index factors,such as average annual temperature,water coverage,soil texture,soil organic matter and effective soil phosphorus,were selected from the major factors affecting the growth of puzzle,and the compound system was considered from various aspects.The evaluation criteria of system suitability zoning were used to construct a suitability evaluation system and model for rubber compound system planting land in Baisha County based on hierarchical analysis,which provided a theoretical basis for future suitability evaluation studies under approximate environmental conditions.(3)In this study,the suitability evaluation model constructed by the hierarchical analysis method was used to divide the rubber plantation in Baisha County into four different levels of planting areas: highly suitable,moderately suitable,barely suitable,and unsuitable,and the final results of the suitability zoning of the compound planting land in the study area were obtained by using GIS spatial analysis technology.The areas of suitable,moderately suitable,barely suitable and unsuitable zones in Baisha County are 366447 ha,307774ha,94041 ha and 12218 ha,respectively,and the proportion of the four different evaluation levels in the total area of rubber forest in Baisha County are 46.95%,43.53%,7.80% and 0.43%,respectively.