Many distributed hydrological models are computationally demanding, limiting their use in calibration, sensitivity analysis, or scenario analysis. Based on the theory that many catchments behave essentially as simple dynamical systems, Joost Buitink recently developed a distributed model that does not have this limitation and can be run on a normal computer if needed.
Hydrological processes are coupled in space and time, meaning that the spatial resolution of a hydrological model implies that an appropriate temporal resolution is needed in order to solve certain processes (like storm runoff and saturation excess overland flow). With the resolution of distributed (global) hydrological models decreasing towards what is referred to as hyperresolution (scale 100 m to 1 km, see Bierkens et al. 2015), there is a need to reduce the temporal resolution of models from daily to hourly timescales. This is illustrated in the figure below.
temporal and spatial resolution at which models typically have been applied in the past (A) and presently (B). The dashed arrow pointing downwards shows the ambitions of spatial hyper-resolution modelling, whereas the dashed arrow pointing towards (C) shows the temporal and spatial resolution of hyper-resolution modelling if it follows the direction of characteristic velocity of hydrometeorological processes.
The problem here is that, while spatial resolution can be changed relatively easy in most models, many models operate at a fixed daily timestep, and those with hourly timesteps are computationally demanding. Interestingly, the complexity present in these models is not needed to achieve high model performance at high spatio-temporal resolution. Kichner (2009), for instance showed that headwater catchments at Plynlimon behaved as simple dynamical systems, and that forward modeling with this approach could result in Nash-Sutcliffe efficiencies as high as 0.95. Good results were also obtained in order basins, such as the Rietholzbach in Switzerland (Teuling et al., 2010). So far, the approach has mainly been applied to smaller basins in a lumped fashion.
The good performance of the simple dynamical systems approach at resolutions needed for simulation of hydrological extremes was a motivation to use it as a basis for an efficient distributed model that would be easy to parameterize (since parameters can be derived from recession analysis) or calibrated. The distributed model has been published recently in Geoscientific Model Development.
An initial application of the distributed model to the Swiss Thur basin found a consistent model performance (Kling-Gupta efficiency) of 0.83-0.87 for hourly discharge across subbasins spanning almost three orders of magnitude in size (Buitink et al. 2020a). In a follow-up study for the Rhine basin, we found an extended version of the model to perform similarly well for the whole Rhine basin, with KGE values of 0.84 and 0.87 for different time periods. In particular, the model seems to deal well with periods of low flow, which is the focus of ongoing research.
Further reading
Bierkens, M. F. P., et al. (2015) Hyper‐resolution global hydrological modelling: what is next?, Hydrol. Process., 29, 310– 320, doi: 10.1002/hyp.10391.
Buitink, J.; L. A. Melsen; J. W. Kirchner & A. J. Teuling (2020), A distributed simple dynamical systems approach (dS2 v1.0) for computationally efficient hydrological modelling at high spatiotemporal resolution. Geoscientific Model Development, 13, 6093–6110, doi:10.5194/gmd-13-6093-2020.
Buitink, J. et al. (2020b), Seasonal discharge response to temperature-driven changes in evaporation and snow processes. https://doi.org/10.5194/esd-2020-73.
Kirchner, J. W.: Catchments as Simple Dynamical Systems: Catchment Characterization, Rainfall-Runoff Modeling, and Doing Hydrology Backward, Water Resour. Res., 45, W02429, https://doi.org/10.1029/2008WR006912, 2009.
Melsen, L. A.; A. J. Teuling; P. J. J. F. Torfs; R. Uijlenhoet; N. Mizukami & M. P. Clark (2016), HESS Opinions: The need for process-based evaluation of large domain hyper-resolution models. Hydrology and Earth System Sciences, 20, 1069–1079, doi:10.5194/hess-20-1069-2016.
Teuling, A. J.; I. Lehner; J. W. Kirchner & S. I. Seneviratne (2010), Catchments as simple dynamical
systems: Experience from a Swiss prealpine catchment. Water Resources Research, 46, W10502, doi.org/10.1029/2009WR008777.