文献类型：期刊文献

英文题名：tSIM: a novel temperate Savanna Identification Method integrating UAV and High-resolution satellite

作者：Li, Xiaoya[1,2] Duan, Tao[3] Yang, Kaijie[1,2] Yang, Bin[4] Wang, Chunmei[5] Tian, Xin[6] Lu, Qi[1,2] Wang, Feng[1,2]

第一作者：Li, Xiaoya

机构：[1] Institute of Ecological Conservation and Restoration, Institute of Desertification Studies, Chinese Academy of Forestry, Beijing, 100091, China; [2] Institute of Great Green Wall, Dengkou County, Inner Mongolia, Bayan Nur, 015200, China; [3] Satellite Application Centre for Ecology and Environment, MEE, Beijing, 100094, China; [4] School of Electronics and Information Engineering, Wuxi University, Jiangsu, Wuxi, 214105, China; [5] Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing, 100094, China; [6] Institute of Forest Resource Information Techniques, Chinese Academy of Forestry, Beijing, 100091, China

年份：2024

外文期刊名：SSRN

收录：EI(收录号：20240169878)

语种：英文

外文关键词：Ecosystems - Forestry - Geographical distribution - Regression analysis - Remote sensing - Satellite imagery - Vegetation

摘要：Savanna is recognized as essential for regulating net primary productivity and terrestrial carbon cycle. It plays a crucial role in supporting a substantial portion of rangelands and livestock, thereby contributing significantly to human livelihoods. Temperate savanna shares a wide geographical distribution and holds significant natural and socio-economic importance. Due to the high spatial heterogeneity induced by tree-grass fusion in savanna, it is difficult to identify woody and herbaceous vegetation at the fine scale even if we used the high-resolution satellite imagery. Therefore, the precise identification of savanna by remote sensing poses significant challenges at a regional scale. The aim of this study is to develop a new temperate Savanna Identification Method (tSIM) and map the temperate savanna on the sandy land in Northeastern China. We estimated the fractional woody vegetation cover (FWVC) and fractional herbaceous vegetation cover (FHVC) by integrating Very High Resolution (VHR) UAV imagery with high-resolution satellite imagery of Sentinel-1 and Sentinel-2 data using Random Forest regression and Classification and Regression Tree algorithm (CART) algorithms. Training and validation datasets were collected from VHR UAV imagery covering the ground area of 50 km2. The proposed method achieved with the overall accuracy of 0.82 and Kappa coefficient of 0.63 for savanna identification. The temperate savanna on the sandy land in the Northeastern China were mapped at spatial resolution of 0.5 ha. We found that temperate savanna in Northeastern China is mainly distributed in the transition zone from mountainous to plain areas, as well as the Horqin Sandy Land, the Otindag Sandy Land, and Hulunbuir Sandy Land, covering a total area of 156,000 km2. This map enhances our understanding of the spatial extent and area of temperate savanna in North China. It offers essential geospatial information for developing conservation strategies for temperate savanna ecosystems. The derived tSIM provides a new insight to explore the extent of regional and global savanna in the future. ? 2024, The Authors. All rights reserved.