Divergent sensitivity of surface water and energy variables to precipitation product uncertainty in the Tibetan Plateau

JOURNAL OF HYDROLOGY

卷:581

Article

Qi, Wei¹  Liu, Junguo¹  Xia, Jun²  Chen, Deliang³ 

[1] Southern Univ Sci & Technol, Sch Environm Sci & Engn, Shenzhen 518055, Guangdong, Peoples R China

[2] Wuhan Univ, State Key Lab Water Resources & Hydropower Engn S, Wuhan 430072, Hubei, Peoples R China

[3] Univ Gothenburg, Dept Earth Sci, Reg Climate Grp, Box 460, S-40530 Gothenburg, Sweden

关键词: Precipitation uncertainty;    Runoff;    Sensitivity;    Tibetan Plateau;    Water Balance;    Distributed hydrological modeling;   

DOI  :  10.1016/j.jhydrol.2019.124338

来源: Elsevier

【 摘 要 】

Precipitation is a major driving factor for land surface water and energy balances. Uncertainty in global precipitation products over observation sparse regions such as the Tibetan Plateau (TP) is generally large. Sensitivity of surface water and energy variables to precipitation uncertainty can provide clues for confidence that can be assigned to simulated water and energy variables in such regions. In this study, the sensitivities of surface water and energy variables to global precipitation product uncertainty over four large river basins in the TP are quantified and inter-compared based on a newly developed sensitivity analysis approach. A water and energy budget-based distributed hydrological model including biosphere is utilized after calibration and validation against observed runoff and Land Surface Temperatures (LSTs) from Moderate Resolution Imaging Spectroradiometer (MODIS). Eight global precipitation products are used to represent the precipitation uncertainty. Results show that Canopy interception loss (CIE) and runoff are highly sensitive to the uncertainty in general, whereas LSTs are not sensitive. Therefore, confidence in simulated CIE and runoff can be considered relatively low when using global precipitation products in the four basins. These results imply that other simulated variables may have large uncertainty even when LSTs simulation performs well, and accurate simulations of CIE and runoff require high accuracy in precipitation. Because CIE has profound influence on local hydrological cycle, the results also imply that utilizing the most accurate precipitation product is critical for local scale hydrological cycle research.

【 授权许可】

Free