内蒙古呼伦贝尔草原30 m/5 d GPP数据集(2023年)  

A dataset of 5-day GPP with a resolution of 30 meter for the Hulunbuir grasslands in Inner Mongolia(2023)

作　　者：刘佳 鲁维 崔晨曦 李豪 胡立知 胡云锋[1,2] LIU Jia;LU Wei;CUI Chenxi;LI Hao;HU Lizhi;HU Yunfeng(State Key Laboratory of Resources and Environmental Information System,Institute of Geographic Sciences and Natural Resources Research,Chinese Academy of Sciences,Beijing,100101,P.R.China;College of Resources and Environment,University of Chinese Academy of Sciences,Beijing,100048,P.R.China;College of Geographic Sciences,Inner Mongolia Normal University,Hohhot,010010,P.R.China;College of Civil Engineering and Architecture,China Three Gorges University,Yichang,443002,P.R.China)

机构地区：[1]中国科学院地理科学与资源研究所,北京100101 [2]中国科学院大学,北京101408 [3]内蒙古师范大学,呼和浩特010018 [4]三峡大学,湖北宜昌443002

出　　处：《中国科学数据(中英文网络版)》2025年第1期140-156,共17页China Scientific Data

基　　金：国家重点研发计划(2021YFD1300501);国家自然科学基金(42371304);资源与环境信息系统国家实验室自主创新重点项目(KPI011)。

摘　　要：总初级生产力(Gross Primary Productivity,GPP)是生态系统生产力的重要指标,反映光合作用下有机物的生成量。GPP水平的下降通常预示着生态退化。为了实现对GPP的高精度监测和量化,本研究基于Landsat-Sentinel协调数据(HLS数据)、ERA5-Land气象再分析数据和全球草地通量站点数据,开发了呼伦贝尔草原30米空间分辨率、5天时间分辨率的GPP数据集。首先,应用Savitzky-Golay滤波法对归一化植被指数(NDVI)进行时序重建,从而获得连续、无缝的植被动态数据。随后,通过对比4个GPP反演模型(MOD17、CFIX、CILUE和EC-LUE),评估各模型在呼伦贝尔草原生态系统模拟中的表现,从而确定精度表现最优的模型。结果表明:基于EC-LUE的改进模型在草地GPP反演中效果突出,模型模拟数据与通量站观测数据的相关系数达0.80。本研究生成了2023年3月至10月呼伦贝尔草原高时空分辨率GPP数据集。本数据集将为草原生态系统的动态监测、碳循环研究以及草地资源管理和畜牧业应用等提供数据支持。Gross Primary Productivity(GPP)is a key indicator of ecosystem productivity,reflecting the amount of organic matter generated through photosynthesis.A decline in GPP often indicates ecological degradation.To enable precise monitoring and quantification of GPP,this study developed a dataset of 5-day GPP with a resolution of 30 meter,based on Landsat-Sentinel harmonized data(HLS),ERA5-Land meteorological reanalysis data,and global grassland flux site data.First,the Savitzky-Golay filter was applied to reconstruct the time series of the Normalized Difference Vegetation Index(NDVI),resulting in continuous and seamless vegetation dynamic data.Subsequently,four GPP inversion models(MOD17,CFIX,CILUE,and EC-LUE)were compared to assess their performance in simulating GPP in the Hulunbuir grassland ecosystem.The results showed that the improved EC-LUE model outperformed others,achieving a correlation coefficient of 0.80 between model simulations and flux station observations.Based on this model,a GPP dataset with a high spatiotemporal resolution was generated for the Hulunbuir grasslands from March to October 2023.This dataset provides valuable support for dynamic monitoring of grassland ecosystems,carbon cycle research,grassland resource management,and livestock applications.

关 键 词：草原 GPP 精度评估 EC-LUE 呼伦贝尔 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置] S812[自动化与计算机技术—控制科学与工程]