Assessments and corrections of GLDAS2.0 forcing data in four large transboundary rivers in the Tibetan Plateau and Northeast China

Wei Qi,Junguo Liu,Hong Yang,Deliang Chen,Lian Feng
DOI: https://doi.org/10.1029/2020ea001576
IF: 3.68
2021-11-23
Earth and Space Science
Abstract:GLDAS2.0 provides long-term fine resolution gridded hydro-meteorological datasets which are necessary for water related studies, particularly in some transboundary rivers that are partially without observation. Yet, GLDAS2.0 has only been validated at limited locations, and few studies have been conducted to develop approaches to correct the GLDAS2.0 data for transboundary rivers. This work assessed the GLDAS2.0 data and developed approaches to correct their uncertainties for studies in large transboundary rivers in the Tibetan Plateau (TP) and Northeast China (NC). To achieve these goals, observational data from 1982 to 2010 and a water and energy budget-based distributed hydrological model including biosphere after calibration and validation were employed. We find that the GLDAS2.0 data (except for wind speed) can reasonably replicate observed seasonal variations. However, its specific humidity and wind speed have large uncertainty, and precipitation has large uncertainty in summer. In NC, the trends of its precipitation, air temperature, downward longwave radiation and wind speed are consistent with the observations. In the Yarlung Tsangpo, Lancang and Nu rivers, the trends of all GLDAS2.0 data reproduce the observation very well, i.e., wetting, warming and dimming trends. Validations show that the corrections are effective and the corrected forcing data can be successfully used in hydrological simulation with improved performance than the raw GLDAS2.0 data, which demonstrates the usefulness of the methodology and corrected forcing data to hydro-meteorological studies in transboundary rivers in China, as well as in other nearby regions/countries.
geosciences, multidisciplinary,astronomy & astrophysics
What problem does this paper attempt to address?