five

Mapping of the ecosystem banks of Amu Darya (Uzbekistan)

收藏
DataCite Commons2023-11-20 更新2025-04-09 收录
下载链接:
https://dataverse.cirad.fr/citation?persistentId=doi:10.18167/DVN1/B6MAVG
下载链接
链接失效反馈
官方服务:
资源简介:
In 2011, the Government of Uzbekistan established the Lower Amudarya Biosphere Reserve (LABR). This protected area aims to conserve the tugay forest: a natural area specific to the stream banks of permanent rivers in Central Asia. The tugay is a riparian forest gallery over several thousand kilometers long, and a few kilometers large along the banks. These forests are located in an ecosystem dominated by a cold arid climate. Forests survive only to the presence of a permanent watercourse: the Amu Darya. Forests are home to flora and fauna specific to wetlands that need protection from the surrounding anthropogenic pressure. Our objective was to define the land cover in the LABR in order to better inderstand the interactions between different actors and factors of land use changes to policymakers and institutions. For the technicians in charge of the LABR management and planning, this mapping must provide them with a functional tool to operate in this forest environment and plan their activities. To carry out a detailed mapping of ecosystems, we analyzed different satellite images: a Sentinel-2 time serie (10 m) acquired between the 01/07/2018 and the 16/07/2019, and a Very High Resolution Spatial (VHRS) image SPOT6 (1.5 m) acquired the 13/07/2019. The methodology is based on an Object Oriented Segmentation of the VHRS image then a classification of each polygon (Random Forest algorithm) based on the SPOT6 image, the Sentinel2 time series, the SRTM at 30m and many indices calculated from these images (NDVI, texture indices, slope, ...). The algorithm was trained via a set of data acquired in the field at different seasons (November 2018 and June 2019) supplemented with data obtained by photo-interpretation. Manual corrections by photo-interpretation were carried out in order to improve the result. The overall accuracy is 94.8%. The classification contains 12 land cover classes: Crop, Open forest, Middle forest, Dense forest, Open steppe, Middle steppe, Dense steppe, Desert steppe, Bare soil, Built-up surface, Rock and Water.
提供机构:
CIRAD Dataverse
创建时间:
2021-02-02
5,000+
优质数据集
54 个
任务类型
进入经典数据集
二维码
社区交流群

面向社区/商业的数据集话题

二维码
科研交流群

面向高校/科研机构的开源数据集话题

数据驱动未来

携手共赢发展

商业合作