中国30米纯像元归一化植被指数图集用于植被覆盖度估算（2014_2018_2022）

本研究利用多角度遥感数据，生成了中国全域30米空间分辨率的纯植被像元归一化植被指数（Vv）和纯土壤像元归一化植被指数（Vs）地图。这些像元级的Vv和Vs地图可与植被指数（VI）模型结合，支持不同空间分辨率和大范围区域的植被覆盖度（FVC）快速精确估算。 产品采用多角度算法（MultiVI）生成，该算法有效解决了Vv和Vs固有的空间变异性问题，与传统统计方法相比显著提高了FVC估算精度。在不同实验场地与实地测量FVC验证显示，估算的FVC均方根偏差（RMSD）约为0.1，表明产品具有较高的可靠性。Using multi-angle remote sensing data, this study generated 30-meter spatial resolution maps of normalized difference vegetation index for pure vegetation pixels (Vv) and pure soil pixels (Vs) across the entire territory of China. These pixel-level Vv and Vs maps can be combined with vegetation index (VI) models to support rapid and accurate estimation of fractional vegetation cover (FVC) across large-scale regions with varying spatial resolutions. This dataset was produced using the multi-angle algorithm (MultiVI), which effectively addresses the inherent spatial variability of Vv and Vs, and significantly improves the accuracy of FVC estimation compared with conventional statistical methods. Validation against field-measured FVC at multiple experimental sites showed that the root mean square deviation (RMSD) of the estimated FVC is approximately 0.1, demonstrating the high reliability of this dataset.

搜集汇总

背景与挑战

背景概述

该数据集提供了中国2014、2018和2022年30米分辨率的纯像元归一化植被指数（NDVI）图，用于估算植被覆盖度（FVC）。它基于多角度遥感数据和MultiVI算法生成，有效减少了纯植被和纯土壤像素的空间变异性，与野外实测数据相比，FVC估算的均方根偏差约为0.1，具有较高的准确性和稳定性。数据集适用于大尺度生态环境监测和植被动态研究。