A simple method to extract tropical monsoon forests using NDVI based on MODIS data: A case study in South Asia and Peninsula Southeast Asia
文献类型：SCI/SSCI论文
| 作者 | Lin S.; Liu, R. G. |
| 发表日期 | 2016 |
| 关键词 | monsoon forest Moderate Resolution Imaging Spectroradiometer (MODIS) Normalized Difference Vegetation Index (NDVI) amplitude threshold classification land-cover classification seasonal forests carbon balance dry forests index agreement dynamics thailand regions images |
| 英文摘要 | Distribution of monsoon forests is important for the research of carbon and water cycles in the tropical regions. In this paper, a simple approach is proposed to map monsoon forests using the Normalized Difference Vegetation Index (NDVI) derived from the Moderate Resolution Imaging Spectroradiometer (MODIS) data. Owing to the high contrast of greenness between wet season and dry season, the monsoon forest can be easily discriminated from other forests by combining the maximum and minimum annual NDVI. The MODIS-based monsoon forest maps (MODMF) from 2000 to 2009 are derived and evaluated using the ground-truth dataset. The MODMF achieves an average producer accuracy of 80.0% and the Kappa statistic of 0.719. The variability of MODMF among different years is compared with that calculated from MODIS land cover products (MCD12Q1). The results show that the coefficient of variation of total monsoon forest area in MODMF is 7.3%, which is far lower than that in MCD12Q1 with 24.3%. Moreover, the pixels in MODMF which can be identified for 7 to 9 times between 2001 and 2009 account for 53.1%, while only 7.9% of MCD12Q1 pixels have this frequency. Additionally, the monsoon forest areas estimated in MODMF, Global Land Cover 2000 (GLC2000), MCD12Q1 and University of Maryland (UMD) products are compared with the statistical dataset at national level, which reveals that MODMF has the highest R (2) of 0.95 and the lowest RMSE of 14 014 km(2). This algorithm is simple but reliable for mapping the monsoon forests without complex classification techniques. |
| 出处 | Chinese Geographical Science |
| 卷 | 26 |
| 期 | 1 |
| 页 | 22-34 |
| 语种 | 英语 |
| ISSN号 | 1002-0063 |
| DOI标识 | 10.1007/s11769-015-0789-3 |
| 源URL | [http://ir.igsnrr.ac.cn/handle/311030/43074]  |
| 专题 | 地理科学与资源研究所_历年回溯文献 |
| 推荐引用方式 GB/T 7714 | Lin S.,Liu, R. G.. A simple method to extract tropical monsoon forests using NDVI based on MODIS data: A case study in South Asia and Peninsula Southeast Asia. 2016. |
入库方式： OAI收割
来源：地理科学与资源研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。