Parameter's Controls of Distributed Catchment Models—How Much Information is in Conventional Catchment Descriptors?

Abstract

One major challenge in large scale modeling is the estimation of spatially consistent distributed parameters, that are parameters with a clear functional relationship to climate and landscape characteristics. We present a newly developed PArameter Set Shuffling (PASS) approach, which is able to provide such regionally consistent parameter sets. The PASS method does not require any a priori assumption on the relationship between model parameters and catchment descriptors. It instead derives these relationships from observed patterns of calibrated parameters and available catchment descriptors. We tested the PASS approach to derive parameters of a conceptual hydrological model applied to 263 German catchments. The resulting median model efficiencies for training and test catchments are, respectively, 0.74 and 0.72, similar to those obtained by other modeling approaches, which use regional calibration. In this study, a combination of catchment descriptors that clearly controls model parameters is not found. In fact, we show that various regional functional relationships between catchment descriptors and model parameters result in similarly good model performances. Moreover, catchment descriptors used for parameter prediction can be replaced in the parameter prediction, without any decrease in model performance. Our results suggest that by using conventional catchment descriptors based on averages, only the amount of information that is also retained in the existing correlations among climatic and catchment indicators is exploited. Development of a new generation of hydrologically meaningful catchment and climate descriptors is required to further improve our capability of forecasting hydrological dynamics of interest by means of large scale models and regionalization approaches.

Flexible vector-based spatial configurations in land models

“…Techniques such as multiscale parameter regionalization (MPR, Samaniego et al, 2010) can be used to scale parameter values for different model configurations. However, applying these techniques, such as in this case that has significant parameter and process uncertainty and significance accuracy-performance tradeoff, should be put through rigorous tests (Merz et al, 2020, Liu et al, 2016.…”

A Process‐Based Framework to Characterize and Classify Runoff Events: The Event Typology of Germany

“…In both cases only the interpolation method differed while the amount of data was not altered. Data resolution : the original spatial resolution of observed and simulated data used for event characterization was upscaled by a factor of 2 (i.e., we analyzed the effect of a coarser spatial resolution). Parameter set : we used snowmelt and soil moisture data simulated by 100 different representative parameter sets (i.e., 100 equifinal realizations of parameters in terms of streamflow efficiency for major German river catchments) of the mHM model (Kumar et al, 2013; Samaniego et al, 2010; Zink et al, 2017). Model structure and calibration technique: we used simulated snowmelt and soil moisture of two regionally calibrated distributed conceptual hydrological models, namely, the mHM and SALTO (Merz et al, 2020) models. The comparison was performed at 8‐km resolution due to the coarser resolution of the latter model outputs.…”

“…Model structure and calibration technique: we used simulated snowmelt and soil moisture of two regionally calibrated distributed conceptual hydrological models, namely, the mHM and SALTO (Merz et al, 2020) models. The comparison was performed at 8‐km resolution due to the coarser resolution of the latter model outputs.…”

“…We used daily gridded rainfall data (1-km grid) from the REGNIE (Regionalisierte Niederschlagshöhen) data set (Rauthe et al, 2013) provided by the German Weather Service, and snowmelt and soil moisture time series (on a 4-km grid) simulated by the mesoscale Hydrologic Model (mHM; Samaniego et al, 2010;Kumar et al, 2013) along with air temperature interpolated using external drift kriging (Zink et al, 2017). For uncertainty analysis and robustness check, we additionally used a precipitation data set interpolated by means of external drift kriging (as reported in Zink et al [2017]), a temperature data set interpolated by using the inverse distance method, 100 equifinal realizations (i.e., 100 different simulations resulted from using 100 different representative parameter sets) of the mHM (Zink et al, 2017), as well as simulations of the alternative distributed regionally calibrated hydrological model SALTO (Merz et al, 2020), which has a structure similar to the conceptual Hydrologiska Byråns Vattenbalansavdelning (HBV) model (Bergström, 1995).…”

On doing large-scale hydrology with Lions: Realising the value of perceptual models and knowledge accumulation

“…Troy et al, 2008;Kuentz et al, 2017;Addor et al, 2018). Subsurface characteristics on the other hand are much harder to observe and characterize (Beven and Cloke, 2012;Merz et al, 2020), and therefore might have to be equally based on our expectations as they are on directly observable properties. Few attempts to integrate (expected) system conceptualizations and data have been made thus far (Boorman et al, 1995).…”