Global‐scale land surface hydrologic modeling with the representation of water table dynamics. Issue 1 (14th January 2014)

• Record Type: Journal Article
• Authors: Koirala, Sujan; Yeh, Pat J.‐F.; Hirabayashi, Yukiko; Kanae, Shinjiro; Oki, Taikan
• Is Part Of: Journal of geophysical research. Volume 119:Issue 1(2014:Jan.)
• Page Start: 75
• Page End: 89
• Publication Date: 2014-01-14
• Subjects: Atmospheric physics -- Periodicals; Geophysics -- Periodicals; 551.5
• DOI: 10.1002/2013JD020398
• Languages: English
• ISSNs: 2169-897X

Abstract

<abstract abstract-type="main"> <title>Abstract</title> <p>[1] Water table dynamics, a basic hydrologic process, was traditionally not considered in global‐scale land surface models (LSMs). In this study, a representation of water table dynamics is integrated into a global LSM to address the shortcomings of existing global‐scale modeling studies and the appropriateness of specifying certain parameters as globally constant. Evaluation of model simulation using globally varying parameters against river discharge observations in selected large rivers shows improvements when the water table dynamics is included in the model. The mechanisms by which the water table dynamics affects land surface hydrologic simulation are then investigated by analyzing the sensitivity simulation of groundwater (GW) capillary flux. The result indicates that global mean evapotranspiration (ET) increases by ~9% when GW capillary flux is considered. The semiarid regions with marked dry season have the largest increase (~25%), while the humid and high‐latitude regions with sufficient moisture but limited radiation energy have the smallest increase. Increase in ET is more pronounced in dry season when GW recharge becomes negative (upward moisture supply from the aquifer), but its magnitude depends on the water table depth (WTD). On the other hand, a deeper WTD caused by the GW capillary flux is found to decrease runoff throughout the year in regions with a large increase in ET in dry season only. Based on our modeling result, about 50% of global land area (especially in humid and high‐latitude regions) is simulated to have the mean WTD shallower than 5 m, which emphasizes the significance of representing water table dynamics in global‐scale LSMs.</p> </abstract>