A glacier runoff extension to the Precipitation Runoff Modeling System. Issue 11 (4th November 2016)

• Record Type: Journal Article
• Authors: Van Beusekom, A. E.; Viger, R. J.
• Is Part Of: Journal of geophysical research. Volume 121:Issue 11(2016)
• Page Start: 2001
• Page End: 2021
• Publication Date: 2016-11-04
• Subjects: glacier runoff -- Precipitation Runoff Modeling System -- Alaska; Geomorphology -- Periodicals; 551.3
• DOI: 10.1002/2015JF003789
• Languages: English
• ISSNs: 2169-9003

Abstract

Abstract: A module to simulate glacier runoff, PRMSglacier, was added to PRMS (Precipitation Runoff Modeling System), a distributed‐parameter, physical‐process hydrological simulation code. The extension does not require extensive on‐glacier measurements or computational expense but still relies on physical principles over empirical relations as much as is feasible while maintaining model usability. PRMSglacier is validated on two basins in Alaska, Wolverine, and Gulkana Glacier basin, which have been studied since 1966 and have a substantial amount of data with which to test model performance over a long period of time covering a wide range of climatic and hydrologic conditions. When error in field measurements is considered, the Nash‐Sutcliffe efficiencies of streamflow are 0.87 and 0.86, the absolute bias fractions of the winter mass balance simulations are 0.10 and 0.08, and the absolute bias fractions of the summer mass balances are 0.01 and 0.03, all computed over 42 years for the Wolverine and Gulkana Glacier basins, respectively. Without taking into account measurement error, the values are still within the range achieved by the more computationally expensive codes tested over shorter time periods. Key Points: Details of glacier runoff module addition to existing hydrological simulation code (Precipitation Runoff Modeling System (PRMS)) Module designed to work in remote areas with works with limited or no on‐glacier measurements Module tested on two well‐studied glaciers and showed comparable results to other models with more data and computation demands