关键词: 多源遥感数据, 橡胶林识别, 空间分布特征, 种植适宜区, 瑞丽市

Abstract:

Taking Ruili City,Dehong Prefecture,Yunnan Province as the study area,this paper analyzes the spatial distribution characteristics of the rubber forest and its suitable planting areas.Sentinel-2 multi-spectral remote sensing images and MOD13Q1 were used as data sources to identify the rubber forest with the method of supporting vector machine classification.The results show that 1) the identification ability of the combined Sentinel-2 and MOD13Q1 NDVI data for rubber forest was significantly better than using only one kind of Sentinel-2 multi-spectral image or MOD13Q1-NDVI time series data,which achieved high precision identification of rubber forest.The overall identification accuracy is 91.41% and the Kappa coefficient is 0.76.2) By the end of 2019,the rubber planting area in Ruili was 7693.83 hm²,with the elevation range of the main planting area being 682-900 m,accounting for 83.61% of the total planting area;the slope of the main planting area was 5-15°,accounting for 84.61% of the total planting area.3) At present,6,433.11 hm² of rubber was planted in the suitable areas in Ruili while 1,260.72 hm² of rubber was not,accounting for 83.61% and 16.39%,respectively.

Key words: multi-source remote sensing data, rubber plantation identification, spatial distribution characteristics, suitable planting areas, Ruili City