全国2019-2023年高时空分辨率碳排放清单产品

本清单产品是采用“自上而下”大气反演方法，结合多源数据与机器学习技术，生成的全国格网化月度碳排放总量空间分布产品。产品融合了卫星遥感反演的温室气体与XCO₂、NO₂等痕量气体浓度数据、ERA5气象场数据以及夜间灯光、人口密度等社会经济活动代理指标，通过构建机器学习模型反演得到2019年至2023年中国的碳排放空间格局。产品提供0.25度（约25公里）和1公里两种空间分辨率版本，以满足不同尺度研究的需求。反演结果与清华大学MEIC清单等主流的“自下而上”排放清单在区域总量和空间分布趋势上表现出良好的一致性，并附有定量的不确定性评估图层。本清单产品为科学界和决策部门提供了一种独立于传统统计核算的碳排放监测视角，适用于国家碳收支的宏观评估、排放趋势的动态分析、不同行业和区域的排放贡献解析，以及各类气候政策与减排情景的模拟与效果预评估。清单产品为栅格格式。(协议共享)

This inventory product is a national gridded monthly total carbon emission spatial distribution product developed using the "top-down" atmospheric inversion method, combined with multi-source data and machine learning techniques. It integrates satellite remote sensing-retrieved greenhouse gas and trace gas concentration data such as XCO₂ and NO₂, ERA5 meteorological field data, and socioeconomic activity proxy indicators including nighttime light and population density, and inverts the spatial patterns of carbon emissions in China from 2019 to 2023 by constructing machine learning models. The product provides two spatial resolution versions: 0.25-degree (approximately 25 km) and 1-kilometer, to meet the requirements of multi-scale research. The inversion results show good consistency with mainstream bottom-up emission inventories such as the Tsinghua University MEIC Inventory in terms of regional total emissions and spatial distribution trends, and are accompanied by quantitative uncertainty assessment layers. This inventory product offers the scientific community and decision-making departments a carbon emission monitoring perspective independent of traditional statistical accounting, and is applicable to macro-assessment of national carbon budgets, dynamic analysis of emission trends, analysis of emission contributions from different industries and regions, as well as simulation and pre-evaluation of various climate policies and emission reduction scenarios. The inventory product is in raster format. (Shared under agreement)