| 基于GEE的东南亚主产区橡胶林分布遥感提取 |
| |
| 引用本文： | 李伟光,张京红,刘少军,陈小敏,邹海平,白蕤,吕润.基于GEE的东南亚主产区橡胶林分布遥感提取[J].热带作物学报,2021,42(11):3345-3350. |
| |
| 作者姓名： | 李伟光 张京红 刘少军 陈小敏 邹海平 白蕤 吕润 |
| |
| 作者单位： | 1. 海南省气候中心,海南海口 5702032. 海南省南海气象防灾减灾重点实验室,海南海口 5702033. 海南省气象科学研究所,海南海口 570203 |
| |
| 基金项目： | 中国气象局创新发展专项(CXFZ2021J070);海南省自然科学基金项目(421MS101);国家自然科学基金项目(41765007);南海气象防灾减灾重点实验室开放基金(SCSF201808) |
| |
| 摘 要： | 为遥感监测全球橡胶主产区橡胶的长势,选取橡胶产量最大的泰国、马来西亚、印度尼西亚三国,开展基于Google Earth Engine（GEE）的橡胶林分布遥感提取研究。通过目视解译选择典型样本区,根据Landsat 7多波段光谱特征、MODIS NDVI反映的植被物候特征建立分类回归树CART分类模型提取橡胶林分布。精度评价显示模型总体分类精度为95.8%,Kappa系数为0.94,生产者精度达到94.8%,用户精度为88.2%,达到较高水平。模型提取结果显示：橡胶林在泰国中部、南部半岛,马来半岛的东部和南部地区,苏门答腊岛分布较为集中,而泰国北部、加里曼丹岛及其他岛屿橡胶林相对稀疏。
|
| 关 键 词： | 东南亚 橡胶 分布 GoogleEarthEngine |
| 收稿时间： | 2021-01-20 |
GEE-based Extraction of Rubber Forest Distribution in Main Producing Areas of Southeast Asia |
| |
| LI Weiguang,ZHANG Jinghong,LIU Shaojun,CHEN Xiaomin,ZOU Haiping,BAI Rui,LYU Run.GEE-based Extraction of Rubber Forest Distribution in Main Producing Areas of Southeast Asia[J].Chinese Journal of Tropical Crops,2021,42(11):3345-3350. |
| |
| Authors: | LI Weiguang ZHANG Jinghong LIU Shaojun CHEN Xiaomin ZOU Haiping BAI Rui LYU Run |
| |
| Institution: | 1. Hainan Climate Center, Haikou, Hainan 570203, China2. Hainan Key Laboratory of Meteorological Disaster Prevention and Mitigation in South China Sea, Haikou, Hainan 570203, China3. Hainan Institute of Meteorological Sciences, Haikou, Hainan 570203, China |
| |
| Abstract: | In order to monitor the growth of rubber forest in the main rubber foresttree producing areas of the world by remote sensing, we carried out the remote sensing extraction of rubber forest distribution based on Google Earth engine (GEE) in Thailand, Malaysia and Indonesia, which are the three largest rubber producing countries in the world. The typical sample areas were selected through visual interpretation, and the classification regression tree CART classification model was established based on the multi-band spectral albedo of Landsat 7 and the vegetation phenological characteristics reflected by MODIS NDVI. The overall classification accuracy of the model was 95.8%, the Kappa coefficient was 0.94, the producer accuracy was 94.8%, and the user accuracy was 88.2%, reaching a high level. The extraction results show that rubber plantations are concentrated in central and southern peninsula of Thailand, the eastern and southern parts of Malay Peninsula and Sumatra Island, while rubber plantations are relatively sparse in northern Thailand, Kalimantan Island and other islands. |
| |
| Keywords: | Southeast Asia rubber distribution Google Earth Engine |
|
| 点击此处可从《热带作物学报》浏览原始摘要信息 |
| 点击此处可从《热带作物学报》下载免费的PDF全文 |
|
