Investigating the influence of two different flow routing algorithms on soil-water-vegetation interactions using the dynamic ecosystem model LPJ-GUESS

This paper compares two flow routing algorithms' influences on ecohydrological estimations in a northern peatland catchment, within the framework of an arctic-enabled version of the dynamic ecosystem model LPJ-GUESS. Accurate hydrological estimations are needed to fully capture vegetation dynamics and carbon fluxes in the subarctic peatland enviroment. A previously proposed distributed hydrological method based on the single flow (SF) algorithm extracted topographic indices has shown to improve runoff estimations in LPJ-GUESS. This paper investigates model performance differences caused by two flow routing algorithms, and importantly both permafrost processes and peatland hydrology are included in the model. The newly developed triangular form-based multiple flow (TFM) is selected due to its improved consideration of flow continuity and more realistic flow estimation over flat surfaces. A variety of measured data is included to assess both hydrological and ecological accuracy, and the results demonstrate that the choice of flow algorithm does matter for mesoscale ecohydrology applications. The allowance of flow convergence and consideration of flow partition differences from different terrain forms in the TFM algorithm yield better correspondence with the observed hydrological processes and also carbon fluxes. By directing flow to only one downslope cell together with its poorer depiction of flow over flat areas, the SF algorithm can result in too high runoff estimations for low-flat regions and overestimate carbon uptake and release in the peatland. The results of this study also highlight the need for care when selecting flow routing algorithms for biogeochemical estimations, especially within hydrologically and climatically sensitive environments. Copyright (c) 2014 John Wiley & Sons, Ltd.