IN13B-1834
mRM - multiscale Routing Model for Land Surface and Hydrologic Models

Monday, 14 December 2015
Poster Hall (Moscone South)
Stephan Thober1, Juliane Mai1, Luis E Samaniego1, David J Gochis2, Rohini Kumar1 and Matthias Cuntz1, (1)Helmholtz Centre for Environmental Research UFZ Leipzig, Leipzig, Germany, (2)National Center for Atmospheric Research, Boulder, CO, United States
Abstract:
Routing streamflow through a river network is a basic step within any distributed hydrologic model. It integrates the generated runoff and allows comparison with observed discharge at the outlet of a catchment.

The Muskingum routing is a textbook river routing scheme that has been implemented in Earth System Models (e.g., WRF-HYDRO), stand-alone routing schemes (e.g., RAPID), and hydrologic models (e.g., the mesoscale Hydrologic Model). Most implementations suffer from a high computational demand because the spatial routing resolution is fixed to that of the elevation model irrespective of the hydrologic modeling resolution. This is because the model parameters are scale-dependent and cannot be used at other resolutions without re-estimation.

Here, we present the multiscale Routing Model (mRM) that allows for a flexible choice of the routing resolution. mRM exploits the Multiscale Parameter Regionalization (MPR) included in the open-source mesoscale Hydrologic Model (mHM, www.ufz.de/mhm) that relates model parameters to physiographic properties and allows to estimate scale-independent model parameters. mRM is currently coupled to mHM and is presented here as stand-alone Free and Open Source Software (FOSS). The mRM source code is highly modular and provides a subroutine for internal re-use in any land surface scheme.

mRM is coupled in this work to the state-of-the-art land surface model Noah-MP. Simulation results using mRM are compared with those available in WRF-HYDRO for the Red River during the period 1990-2000. mRM allows to increase the routing resolution from 100m to more than 10km without deteriorating the model performance. Therefore, it speeds up model calculation by reducing the contribution of routing to total runtime from over 80\% to less than 5\% in the case of WRF-HYDRO. mRM thus makes discharge data available to land surface modeling with only little extra calculations.