全球250m陆地生态系统净生产力遥感大数据估算产品（2001-2020年）

净生态系统生产力（NEP）是陆地生态系统中一个非常重要的参数，在不考虑干扰造成的其他碳通量的前提下，代表生态系统与大气之间的净生态系统交换（NEE）。在该NEP产品中，我们构建了了以通量站点实测NEP为目标变量、遥感产品和气象数据为驱动变量的训练数据集。并进一步优化了这些数据集，以解决空间分布方面的问题，最终通过应用随机森林方法训练模型并生成 2001-2020 年的全球 NEP 产品。

Net Ecosystem Productivity (NEP) is a critical parameter in terrestrial ecosystems. Excluding other carbon fluxes caused by disturbances, it represents the Net Ecosystem Exchange (NEE) between the ecosystem and the atmosphere. In this NEP product, we constructed a training dataset using in-situ NEP measured at flux towers as the target variable, and remote sensing products and meteorological data as driving variables. We further optimized this dataset to resolve issues related to spatial distribution, and finally trained a model via the random forest method to generate a global NEP product covering the period from 2001 to 2020.