Long term high-resolution national precipitation (P), soil moisture (SM), and snow water equivalent datasets are necessary for predicting floods, droughts, and assessing the impact of climate change on river flow in China. The current long-term daily or sub daily datasets of P, SM, and SWE are limited by rough spatial resolution or lack of local correction. Although the SM and SWE data obtained from nationwide hydrological simulations have good spatial resolution and utilize local forcing data, hydrological models cannot be directly calibrated using SM and SWE data.
In this study, we generated a daily 0.1°dataset of P, SM, and SWE in China from 1981 to 2017. Cross validation was performed using spatial and temporal segmentation of CMPA data, and the median Kling Gupta efficiency (KGE) of reconstructed P was 0.68 for all grids on a daily scale. At the daily scale, the median KGE of SM in calibration for all grids is 0.61. For the grids of two regions with abundant snow cover, the median SWE KGEs calibrated for the Songhua Basin, Liaohe Basin, and Northwest Continental Basin are 0.55 and -2.41 on a daily scale, respectively.
Overall, the P and SM performance of the reconstructed dataset in South and East China is better than that in North and West China, and the SWE performance in Northeast China is better than that in other regions. As the first long-term daily dataset of 0.1°P, SM, and SWE, combining information from local observations and satellite based data benchmarks, this reconstruction product is of great value for future national hydrological process investigations.
Use global background data and local on-site data as mandatory inputs, using satellite data as the reconstruction benchmark.
Merge P data from 1981 to 2017 between global 0.1°and local 0.25°, and reconstruct the historical P of 0.1°. China Consolidated Precipitation Analysis (CMPA) using Stacked Machine Learning Models from 2008 to 2017. The reconstructed P data is used to drive the HBV hydrological model to simulate SM and SWE data from 1981 to 2017. SM simulation is calibrated using soil moisture active passive level 4 (SMAP-L4) data. The SWE simulation was calibrated using the China National Satellite Snow Depth Dataset (Che and Dai, 2015) and the Moderate Resolution Imaging Spectrometer (MODIS) snow cover data.
The data quality is good.
This work is licensed under a
Creative
Commons Attribution 4.0 International License.
CSTR:11738.11.NCDC.ZENODO.DB3946.2023
10.5281/zenodo.5811099