青藏高原30米叶面积指数产品（2023）

1）数据内容：本数据集包含2023年青藏高原地区30米分辨率叶面积指数遥感产品。2）数据来源及加工方法：利用Landsat时间序列数据，研发了基于物理机理模型（PROSPECT+SAIL模型）和机器学习方法（随机森林模型）的大区域30米叶面积指数产品生产算法。基于PROSPECT+SAIL模型结合Landsat光谱响应函数构建代表各种植被状况的数据集来训练随机森林模型，基于长时序的Landsat影像和训练好的随机森林模型反演得到长时序30米分辨率年最大合成叶面积指数。3）数据质量描述：精度验证结果表明产品的R2为0.75，root-mean-square error（RMSE）为1.16。4）数据应用成果及前景：叶面积指数高度综合了植被的水平覆盖状况和垂直结构，是植被冠层的重要结构参数，该数据集可为陆面过程模拟、资源调查、生态环境监测、全球变化研究等相关领域的研究和应用提供数据产品支撑。

1) Data Content: This dataset contains 30-meter resolution Leaf Area Index (LAI) remote sensing products over the Qinghai-Tibet Plateau in 2023. 2) Data Source and Processing Method: Using Landsat time-series data, an algorithm for producing large-scale 30-meter resolution LAI products was developed, which integrates the physically-based model (PROSPECT+SAIL model) and machine learning method (Random Forest model). A dataset representing various vegetation conditions was constructed based on the PROSPECT+SAIL model combined with Landsat spectral response functions to train the Random Forest model. Then, time-series 30-meter resolution annual maximum composite LAI was retrieved using long-term Landsat images and the trained Random Forest model. 3) Data Quality Description: The accuracy verification results show that the product has an R² of 0.75 and a root-mean-square error (RMSE) of 1.16. 4) Data Application Achievements and Prospects: Leaf area index highly integrates the horizontal coverage and vertical structure of vegetation, and is an important structural parameter of vegetation canopies. This dataset can provide data product support for research and applications in related fields such as land surface process simulation, resource investigation, ecological environment monitoring, and global change research.

搜集汇总

背景与挑战

背景概述

该数据集是2023年青藏高原地区的30米分辨率叶面积指数（LAI）遥感产品，基于Landsat时间序列数据，采用物理机理模型和机器学习方法生产，精度较高（R²=0.75，RMSE=1.16）。数据集适用于陆面过程模拟、资源调查和生态环境监测等领域，以开放获取方式共享，数据大小为8.65 GB。