Multi-basin Hydrological Modeling at a Pan-European Scale Using a Combination of Catchment Classification and Regional Parameter Estimation

Friday, 19 December 2014: 2:55 PM
Yeshewatesfa Hundecha Hirpa, Anna Kuentz, Ilias Pechlivanidis, Jafet Andersson and Berit Arheimer, Swedish Meteorological and Hydrological Institute, Norrköping, Sweden
Numerical hydrological models are essential tools for the retrieval of water information in many societal sectors. Although such models are usually applied at a catchment scale, there is an increasing need to model large spatial domains to be able to manage resources in a coordinated manner. However, modeling a large spatial domain brings with it a number of challenges. One of the issues that needs to be addressed is estimation of the model parameters under a situation where the physical and climate characteristics of the catchments making up the model domain are highly variable. A semi-distributed continuous hydrological model HYPE is applied to model daily stream flows over the entire pan-European region. The model includes simplified components to simulate human impacts, such as irrigation, reservoir regulation, and water abstraction. The entire region was subdivided into more than 35,000 subcatchments and a catchment classification scheme was implemented to group catchments based on similarity of their physiographic and climate characteristics. A range of catchment physiographic and climate attributes were used and principal component analysis was employed to reduce the dimensionality. These attributes were iteratively selected so that catchments of the same group have one or more of their hydrological signatures similar. The model parameters were estimated using a regionalization approach. Soil and landuse parameters were fixed based on soil and landuse classes respectively. Lake and irrigation parameters were estimated separately. Model parameters that are defined at a catchment scale were estimated for each group as functions of the catchment physical and climate characteristics. Simultaneous calibration was performed on a number of catchments to optimize the overall model performance and the functional relationships between the parameters and the catchment properties. ERA-interim reanalysis forcings were used as climate drivers and MODIS products were used to constrain the evapotranspiration parameters. Around 3,000 gauging stations were used to calibrate and validate the regionalization scheme. The regionalized parameters were evaluated by testing the performance on catchments that were not used in the derivation of the relationship and good results were obtained.