Impact of vegetation dynamics on hydrological processes in a semi-arid basin by using a land surface-hydrology coupled model

JOURNAL OF HYDROLOGY

卷:551

Article

Jiao, Yang¹  Lei, Huimin¹  Yang, Dawen¹  Huang, Maoyi²  Liu, Dengfeng³  Yuan, Xing⁴ 

[1] Tsinghua Univ, State Key Lab Hydrosci & Engn, Dept Hydraul Engn, Beijing 100084, Peoples R China

[2] Pacific Northwest Natl Lab, Atmospher Sci & Global Change Div, Richland, WA USA

[3] Xian Univ Technol, State Key Lab Base Ecohydraul Engn Arid Area, Xian 710048, Shaanxi, Peoples R China

[4] Chinese Acad Sci, RCE TEA, Inst Atmospher Phys, Beijing 100029, Peoples R China

关键词: Eco-hydrological modeling;    Model coupling;    Vegetation dynamics;    Regional hydrological change;   

DOI  :  10.1016/j.jhydrol.2017.05.060

来源: Elsevier

【 摘 要 】

Land surface models (LSMs) are widely used to understand the interactions between hydrological processes and vegetation dynamics, which is important for the attribution and prediction of regional hydrological variations. However, most LSMs have large uncertainties in their representations of eco-hydrological processes due to deficiencies in hydrological parameterizations. In this study, the Community Land Model version 4 (CLM4) LSM was modified with an advanced runoff generation and flow routing scheme, resulting in a new land surface-hydrology,coupled model, CLM-GBHM. Both models were implemented in the Wudinghe River Basin (WRB), which is a semi-arid basin located in the middle reaches of the Yellow River, China. Compared with CLM, CLM-GBHM increased the Nash Sutcliffe efficiency for daily river discharge simulation (1965-1969) from-0.03 to 0.23 and reduced the relative bias in water table depth simulations (2010-2012) from 32.4% to 13.4%. The CLM-GBHM simulations with static, remotely sensed and model-predicted vegetation conditions showed that the vegetation in the WRB began to recover in the 2000s due to the Grain for Green Program but had not reached the same level of vegetation cover as regions in natural eco-hydrological equilibrium. Compared with a simulation using remotely sensed vegetation cover, the simulation with a dynamic vegetation model that considers only climate-induced change showed a 10.3% increase in evapotranspiration, a 47.8% decrease in runoff, and a 62.7% and 71.3% deceleration in changing trend of the outlet river discharge before and after the year 2000, respectively. This result suggests that both natural and anthrojJogenic factors should be incorporated in dynamic vegetation models to better simulate the eco-hydrological cycle. (C) 2017 Elsevier B.V. All rights reserved.

【 授权许可】

Free