2000-2020全球1km分辨率蒸散发融合数据集

全球陆表蒸散发是全球水循环的关键环节，是能量平衡与物质平衡的结合点，且与全球地表碳收支密切相关，具有全球流域水文-生态过程耦合的重要纽带作用，准确估算全球陆表蒸散发不仅对于研究地球系统和全球气候变化具有重要意义，而且对于全球水资源有效开发利用、农业水资源综合管理、全球旱情监测和预测等方面具有十分重要的应用价值。项目团队采用全球645个涡动相关仪观测的实际蒸散发数据，对现有多种全球陆表蒸散发产品数据集，在不同下垫面类型、气候带和高程等条件下，使用平均绝对误差（MAE）、平均绝对百分比误差（MAPE）、均方根误差（RMSE）相对均方根误差（RRMSE）、相关系数（R）和一致性指数（d）共计六种指标进行时空尺度验证分析，筛选出各个条件下以各种指标验证时表现最优的全球陆表蒸散发数据产品，统计后筛选出综合表现最优的两种或三种全球陆表蒸散发数据产品，采用Taylor Skill Score加权与均值等方法进行数据融合，构建得到了全球1km分辨率的长时间序列月度实际蒸散发融合数据集（2000-2020年）；并采用独立的全球不同区域与涡动相关仪观测的实际蒸散发数据进行验收保证该数据的一致性。该数据的单位是毫米/月，使用前应除以100；每个月的GeoTIFF文件分为两部分，第一部分名为SynthesizedET_YYYY_MM-0000000000-0000000000.tif，第二部分名为SynthesizedET_YYYY_MM-0000000000-0000032768.tif；NoData值被分配为-1。其中YYYY是指年份（从2000年到2020年），MM是指月份（01, 02, 03, 04, 05, 06, 07, 08, 09, 10, 11, 12）。

Global land surface evapotranspiration (ET) is a critical component of the global water cycle, a junction of energy and mass balances, and closely linked to global surface carbon budgets. It also serves as a key link coupling hydrological and ecological processes in global river basins. Accurate estimation of global land surface ET is not only of great significance for studying the Earth system and global climate change, but also has crucial application value for the effective development and utilization of global water resources, comprehensive management of agricultural water resources, global drought monitoring and prediction, and other related fields. The project team utilized 645 global in-situ actual ET datasets measured by eddy covariance systems to perform spatial-temporal scale validation and analysis on multiple existing global land surface ET product datasets under conditions of different underlying surface types, climate zones and elevations, using a total of six indicators: Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE), Root Mean Square Error (RMSE), Relative Root Mean Square Error (RRMSE), Correlation Coefficient (R), and Index of Agreement (d). The team screened out the optimal global land surface ET data products for each condition under each individual indicator, then selected the top two or three products with the best comprehensive performance through statistical analysis. Data fusion was conducted via methods such as Taylor Skill Score weighting and arithmetic averaging, and a long-time series monthly actual ET fused dataset (2000–2020) with a global 1 km resolution was constructed. Independent actual ET datasets measured by eddy covariance systems from different global regions were used for validation to ensure the consistency of this dataset. The unit of this dataset is millimeters per month (mm/month), and it should be divided by 100 before use. Each month's GeoTIFF files are split into two parts: the first part is named SynthesizedET_YYYY_MM-0000000000-0000000000.tif, and the second part is named SynthesizedET_YYYY_MM-0000000000-0000032768.tif. The NoData value is assigned as -1. Here, YYYY refers to the year ranging from 2000 to 2020, and MM refers to the month (01, 02, 03, 04, 05, 06, 07, 08, 09, 10, 11, 12).