祁连山区域30m土地覆盖分类产品数据集（1990-2020年，每5年一期）

本数据集为祁连山区域1990-2020年的30m土地覆盖分类产品（每5年一期）。该数据集使用长时间序列的30m陆地卫星序列数据生成，具有以下特点：1）利用地理分区和分层分类决策树策略，并通过大量目视解译进行完善，精度高；2） 以2015年为基准，其他年份采用变化检测方法生产，各期产品间一致性高；3） 主要生产过程在谷歌地球引擎（GEE）云平台上进行，效率高。该数据集包含了耕地、林地、草地、灌丛、湿地、水体、不透水层、裸地和冰川积雪，共9大类，总体精度优于90%。

This dataset is a 30m land cover classification product for the Qilian Mountains region spanning 1990 to 2020, with each phase released every 5 years. It is generated using long-time-series 30m Landsat satellite data, and has the following characteristics: 1) It adopts a geographic zoning and hierarchical classification decision tree strategy, and is refined through extensive visual interpretation, achieving high accuracy; 2) It takes 2015 as the baseline year, while product phases for other years are generated via change detection methods, ensuring high consistency across all phases; 3) The core production workflow is conducted on the Google Earth Engine (GEE) cloud platform, delivering high efficiency. This dataset includes 9 major land cover categories: cropland, forestland, grassland, shrubland, wetland, water body, impervious surface, bare land, and glacier and snow cover, with an overall accuracy exceeding 90%.

搜集汇总

背景与挑战

背景概述

该数据集为祁连山区域1990-2020年每5年一期的30米土地覆盖分类产品，采用高精度分类方法生产，包含9大类土地覆盖类型，总体精度优于90%。数据集以2015年为基准，通过变化检测方法确保各期产品间的高一致性。