TY - JOUR
T1 - On characterizing the temporal dominance patterns of model parameters and processes
AU - Guse, Björn
AU - Pfannerstill, Matthias
AU - Strauch, Michael
AU - Reusser, Dominik E.
AU - Lüdtke, Stefan
AU - Volk, Martin
AU - Gupta, Hoshin
AU - Fohrer, Nicola
N1 - Funding Information:
We thank Schleswig-Holstein Agency for Coastal Defence, National Park and Marine Conservation of Schleswig-Holstein (LKN-SH) and the Wasserwirtschaftsamt Hof in Bavaria for the discharge data and the land survey office of Schleswig-Holstein for providing the digital elevation model as well as the Landesamt fuer Natur und Umwelt Schleswig-Holstein (LANU) for the river network for Treene (Digitales Anlagenverzeichnis, DAV). We thank the German Environmental Agency (UBA) for the river network for the Saale. Furthermore, we thank the German Weather Service (DWD) and the Potsdam Institute for Climate Impact Research (PIK) for the support with climate data. This project has been carried out with financial support of the Deutsche Forschungsgemeinschaft (DFG) via the project GU 1466/1-1 (hydrological consistency in modelling). The work of the second author was supported by financial support of a scholarship by the German Environmental Foundation (DBU). This work was partly funded by the Helmholtz Programme – Terrestrial Environmental Research. Dominik Reusser was supported by the German Federal Ministry of Education and Research (BMBF) via its initiative Potsdam Research Cluster for Georisk Analysis, Environmental Change and Sustainability (PROGRESS – grant: 03IS2191B). We would like to thank the community of the open source software R, which was used for these analyses. Finally, we thank the editor and two anonymous reviewers for their very constructive comments, which improved the manuscript.
Publisher Copyright:
Copyright © 2015 John Wiley & Sons, Ltd.
PY - 2016/6/30
Y1 - 2016/6/30
N2 - Diagnostic analyses of hydrological models intend to improve the understanding of how processes and their dynamics are represented in models. Temporal patterns of parameter dominance could be precisely characterized with a temporally resolved parameter sensitivity analysis. In this way, the discharge conditions are characterized, that lead to a parameter dominance in the model. To achieve this, the analysis of temporal dynamics in parameter sensitivity is enhanced by including additional information in a three-tiered framework on different aggregation levels. Firstly, temporal dynamics of parameter sensitivity provide daily time series of their sensitivities to detect variations in the dominance of model parameters. Secondly, the daily sensitivities are related to the flow duration curve (FDC) to emphasize high sensitivities of model parameters in relation to specific discharge magnitudes. Thirdly, parameter sensitivities are monthly averaged separately for five segments of the FDC to detect typical patterns of parameter dominances for different discharge magnitudes. The three methodical steps are applied on two contrasting catchments (upland and lowland catchment) to demonstrate how the temporal patterns of parameter dynamics represent different hydrological regimes. The discharge dynamic in the lowland catchment is controlled by groundwater parameters for all discharge magnitudes. In contrast, different processes are relevant in the upland catchment, because the dominances of parameters from fast and slow runoff components in the upland catchment are changing over the year for the different discharge magnitudes. The joined interpretation of these three diagnostic steps provides deeper insights of how model parameters represent hydrological dynamics in models for different discharge magnitudes. Thus, this diagnostic framework leads to a better characterization of model parameters and their temporal dynamics and helps to understand the process behaviour in hydrological models.
AB - Diagnostic analyses of hydrological models intend to improve the understanding of how processes and their dynamics are represented in models. Temporal patterns of parameter dominance could be precisely characterized with a temporally resolved parameter sensitivity analysis. In this way, the discharge conditions are characterized, that lead to a parameter dominance in the model. To achieve this, the analysis of temporal dynamics in parameter sensitivity is enhanced by including additional information in a three-tiered framework on different aggregation levels. Firstly, temporal dynamics of parameter sensitivity provide daily time series of their sensitivities to detect variations in the dominance of model parameters. Secondly, the daily sensitivities are related to the flow duration curve (FDC) to emphasize high sensitivities of model parameters in relation to specific discharge magnitudes. Thirdly, parameter sensitivities are monthly averaged separately for five segments of the FDC to detect typical patterns of parameter dominances for different discharge magnitudes. The three methodical steps are applied on two contrasting catchments (upland and lowland catchment) to demonstrate how the temporal patterns of parameter dynamics represent different hydrological regimes. The discharge dynamic in the lowland catchment is controlled by groundwater parameters for all discharge magnitudes. In contrast, different processes are relevant in the upland catchment, because the dominances of parameters from fast and slow runoff components in the upland catchment are changing over the year for the different discharge magnitudes. The joined interpretation of these three diagnostic steps provides deeper insights of how model parameters represent hydrological dynamics in models for different discharge magnitudes. Thus, this diagnostic framework leads to a better characterization of model parameters and their temporal dynamics and helps to understand the process behaviour in hydrological models.
KW - catchment modelling
KW - diagnostic model analysis
KW - model understanding
KW - parameter characterization
KW - temporal parameter sensitivity analysis
UR - http://www.scopus.com/inward/record.url?scp=84959432211&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84959432211&partnerID=8YFLogxK
U2 - 10.1002/hyp.10764
DO - 10.1002/hyp.10764
M3 - Article
AN - SCOPUS:84959432211
VL - 30
SP - 2255
EP - 2270
JO - Hydrological Processes
JF - Hydrological Processes
SN - 0885-6087
IS - 13
ER -