喜马拉雅山脉树冠覆盖度数据集（1990-2020）

森林是陆地生态系统的重要组成部分，约占全球陆地面积的三分之一，在调节气候、为物种提供栖息地及维持全球生态平衡中发挥关键作用。树冠覆盖度是树木在单位面积上的垂直投影面积，能够有效反映森林的水平结构与覆盖状况，是监测森林变化和评估生态功能的重要参数。本研究选取250 m MODIS VCF Tree Cover数据为训练样本，并将 Landsat 地表反射率数据重采样到250 m，构成了训练样本数据集，将训练好的模型运用到原始30 m分辨率的 Landsat 数据上。基于机器学习方法获得了1990~2020年尺度的30m空间分辨率的树冠覆盖度数据。利用激光雷达GEDI冠层覆盖度数据评估了提取的树冠覆盖度，发现基于 Landsat 的喜马拉雅山脉树冠覆盖度估计值与激光雷达GEDI冠层覆盖度数据具有很强的一致性，R2达到 0.7。该数据集可为山地森林动态变化监测、气候变化响应研究以及区域森林资源管理提供科学支撑。

Forests are vital components of terrestrial ecosystems, accounting for approximately one-third of the global land area, and play a critical role in climate regulation, providing habitats for species, and maintaining global ecological balance. Tree canopy cover refers to the vertical projected area of trees per unit area, which effectively reflects the horizontal structure and coverage status of forests, and is an important parameter for monitoring forest changes and evaluating ecological functions. In this study, 250 m MODIS VCF Tree Cover data was selected as training samples, and Landsat surface reflectance data was resampled to 250 m to construct the training sample dataset. The trained model was then applied to the original 30 m resolution Landsat data. Tree canopy cover data with 30 m spatial resolution from 1990 to 2020 was obtained using machine learning methods. The extracted tree canopy cover data was evaluated using LiDAR GEDI canopy cover data. It was found that the Landsat-based estimates of tree canopy cover in the Himalayas had a strong consistency with the LiDAR GEDI canopy cover data, with an R² of 0.7. This dataset can provide scientific support for mountain forest dynamic change monitoring, climate change response research, and regional forest resource management.

搜集汇总

背景与挑战

背景概述

该数据集提供了喜马拉雅山脉1990年至2020年间的30米空间分辨率树冠覆盖度信息，基于MODIS和Landsat数据通过机器学习方法生成，并经过激光雷达GEDI数据验证（R2=0.7），准确性较高，适用于森林动态监测、气候变化研究和区域资源管理。