2001-2020年全球500m水体逐日产品

本数据以内陆水体为研究对象，在充分理解其光谱、时间、地形特征的基础上，利用MODIS逐日反射率数据，基于一套自动的逐日水体提取算法流程，该算法流程包括了单景影像水体提取、水体分类后处理及时间序列处理。利用FROM-GLC训练和验证样本，以及Landsat影像对MODIS动态水产品进行验证，验证结果表明生产者和用户精度均在85%以上。

This dataset takes inland waters as its core research subject. Based on a full understanding of their spectral, temporal and topographic characteristics, MODIS daily reflectance data was adopted to develop an automated daily water body extraction workflow, which covers single-scene image water body extraction, post-classification processing of water bodies and time-series processing. The MODIS dynamic water product was validated using FROM-GLC training and validation samples as well as Landsat images. The validation results demonstrate that both the producer's accuracy and user's accuracy are above 85%.

搜集汇总

背景与挑战

背景概述

该数据集是2001年至2020年全球范围的500米分辨率逐日水体产品，基于MODIS反射率时间序列的自动算法生成，包括单日水体提取、后处理和时间序列分析步骤。验证精度超过85%，确保了数据的可靠性，适用于全球水体动态监测和研究。