全球294万条河段的天然径流模拟数据（GRADES）

全球294万条河段的天然径流量模拟数据产品，单位m3/s。本数据是基于VIC水文过程模式与RAPID矢量河网汇流模型模拟得到。其中陆面水文过程模式空间分辨率为0.25°，矢量汇流模式中的河网数据基于90-m MERIT Hydro水文矫正地形数据产品提取。产流部分经过基于机器学习得到的径流特征值进行参数率定，并基于多分位数径流特征值进行了格点尺度的产流偏差矫正，经全球1.4万个径流观测站点验证，数据产品具有较优的验证精度。2024年4月更新GRADES空间索引信息：MERIT_Basins矢量河网数据。径流数据中，rivid与MERIT_Basins中的ID为一一对应关系。更多信息请参考：Lin et al. (2019, WRR)。

A simulated natural streamflow dataset covering 2.94 million river segments globally, with the unit of m³/s. This dataset was generated based on the Variable Infiltration Capacity (VIC) land surface hydrological model and the RAPID vector river network routing model. The spatial resolution of the VIC land surface hydrological model is 0.25°, and the river network data used in the vector routing model was extracted from the 90-m MERIT Hydro hydrologically corrected topographic dataset. The runoff generation component was calibrated using runoff characteristic values derived from machine learning, and then subjected to grid-scale runoff bias correction based on multi-quantile runoff characteristic values. This dataset has achieved excellent validation performance when tested against 14,000 global streamflow observation stations. In April 2024, the GRADES spatial index was updated using the MERIT_Basins vector river network data. In the streamflow dataset, the "rivid" field has a one-to-one correspondence with the IDs in MERIT_Basins. For more information, please refer to Lin et al. (2019, WRR).

搜集汇总

背景与挑战

背景概述

全球294万条河段的天然径流模拟数据（GRADES）是基于VIC水文过程模式和RAPID矢量河网汇流模型模拟的高精度数据，时间覆盖1979-2013年，空间分辨率为0.25°，适用于全球范围内的水文研究。