Premium
Evaluations and Improvements of GLDAS2.0 and GLDAS2.1 Forcing Data's Applicability for Basin Scale Hydrological Simulations in the Tibetan Plateau
Author(s) -
Qi Wei,
Liu Junguo,
Chen Deliang
Publication year - 2018
Publication title -
journal of geophysical research: atmospheres
Language(s) - English
Resource type - Journals
eISSN - 2169-8996
pISSN - 2169-897X
DOI - 10.1029/2018jd029116
Subject(s) - environmental science , data assimilation , climatology , plateau (mathematics) , precipitation , water cycle , rain gauge , forcing (mathematics) , surface runoff , structural basin , climate change , scale (ratio) , meteorology , biosphere model , biosphere , geography , geology , mathematics , mathematical analysis , ecology , paleontology , oceanography , cartography , biology
Hydroclimatic data are of importance to understand the water cycle and therefore for water resource assessment. Such data are of paramount importance for the Tibetan Plateau (TP), which is the source region of several large rivers in Asia. The Global Land Data Assimilation System (GLDAS) 2.0 and 2.1 provide abundant fine resolution hydroclimatic data. However, evaluations on their applicability have not been carried out for the TP. This study aims to evaluate and improve their applicability in basin‐scale hydrological applications in the TP. Gauge‐based data, a hydrological model including biosphere and seven state‐of‐the‐art global precipitation products are utilized to carry out the study in four large basins in the TP. We find that GLDAS2.1 shows significant warming trends from 2001 to 2010, whereas GLDAS2.0 shows cooling trends, although only significant in the Upper Yellow River basin. The contrasting trends imply that caution should be taken when using them to analyze climate change impacts. On a monthly scale, GLDAS2.1 precipitation on average is closer to the gauge‐based data than GLDAS2.0, but both of them have high uncertainty. Therefore, further quality improvements in precipitation are of importance. We also find CMORPH‐BLD has better performance than other products in terms of Nash‐Sutcliffe Efficiency (NSE), Relative Bias (RB), and root‐mean‐square error. Combining CMORPH‐BLD with GLDAS2.0 forcing data generates more realistic runoff simulation than GLDAS2.1, with NSE and RB being 0.85 and 16% on average. The results provide unique insights into the studied data and are beneficial for water resource assessment in the TP.