2020-2024年王朗典型山地总初级生产力参考“真值”产品

该产品是以王朗地区典型山地生态系统为研究区，基于WRF-LES高分辨率气象模拟与EC-LUE光能利用率模型融合研发的参考“真值”产品。模型利用WRF-LES模式模拟输出的温度、短波辐射、潜热与感热通量等气象数据，结合ANUSPLIN插值方法引入高分辨率DEM数据作为协变量将气象数据精细化至30米空间尺度。同时，结合30米高分辨率地表反射率遥感数据计算光合有效辐射比率FPAR，最终通过EC-LUE模型反演生成2020-2024年5×5平方公里范围内的时空连续总初级生产力产品。该产品覆盖王朗山地区域，空间分辨率为30米、时间分辨率为5天,时间跨度为2020-2024年。产品精度检验结果表明，模型反演GPP与站点实测值的R²达0.83， RMSE为0.68 gC/m²，达到了项目设计要求（误差＜10%）。(协议共享)

This reference "ground truth" product is developed for the typical mountain ecosystem study area in the Wanglang region, by fusing high-resolution meteorological simulations from the WRF-LES model and the EC-LUE light use efficiency model. The model utilizes meteorological data including air temperature, shortwave radiation, latent and sensible heat fluxes output by the WRF-LES model, and combines the ANUSPLIN interpolation method with high-resolution DEM data as a covariate to downscale the meteorological data to a spatial resolution of 30 meters. In addition, the Fraction of Photosynthetically Active Radiation (FPAR) is calculated using 30-meter high-resolution surface reflectance remote sensing data, and finally the spatiotemporally continuous Gross Primary Productivity (GPP) product spanning 2020 to 2024 within the 5×5 km² study area is generated via inversion with the EC-LUE model. This product covers the entire Wanglang Mountain region, with a spatial resolution of 30 meters, a temporal resolution of 5 days, and a time span from 2020 to 2024. Accuracy validation results of the product show that the coefficient of determination (R²) between the GPP retrieved by the model and in-situ site measurements reaches 0.83, with a Root Mean Square Error (RMSE) of 0.68 gC/m², meeting the project design requirements (estimation error < 10%). (Shared under agreement)

搜集汇总

背景与挑战

背景概述

该数据集是2020-2024年王朗典型山地总初级生产力的参考“真值”产品，基于WRF-LES气象模拟和EC-LUE模型生成，空间分辨率为30米，时间分辨率为5天，覆盖5×5平方公里区域。数据集通过整合高分辨率气象和遥感数据，提供了时空连续的GPP估计，验证结果显示与实地测量值高度一致（R²=0.83，RMSE=0.68 gC/m²），误差低于10%，适用于山地生态系统研究。