Parameter sensitivity and optimization of a catchment-scale hydrologic model across Sweden
Independent thesis Advanced level (degree of Master (One Year)), 40 credits / 60 HE creditsStudent thesis
Understanding of the timing and spatial patterns of the fluxes of water is a vital part of any hydrologic-nutrient transport model. To investigate hydrological and nutrient transport interactions, the Baltic Nest Institute (BNI) has implemented the catchment simulation model (CSIM) in the Baltic Sea drainage basin (BSDB). This study focuses on the application of the CSIM model across Sweden as a part of the Baltic Sea drainage basin with specific focus on parameter sensitivity and optimization/calibration. To this end, the spatiotemporal hydrology parameter sensitivity of the CSIM model is explored. In addition, potential improvement over the existing base model parameter calibration is considered using a Genetic Algorithm (GA) optimization method.
The CSIM model parameters show remarkable spatial and temporal (seasonal scale) variation in Sweden. Several regions and watersheds are rather insensitive to the parameters of the CSIM model. Further, the spatial parameter sensitivity emphasizes the relevance of site-specific calibration of the model. The temporal sensitivity analysis clearly shows distinct sensitivity variations in the northern-boreal basins relative to the southern-temperate basins. The annual and seasonal optimization results show that the GA could improve model performance particularly when considering season-specific values for the most sensitive parameters in the CSIM model.
Moreover, the optimization outcomes highlighted that the CSIM model, like any hydrological model, faces parameter uncertainties and the concept of equifinality should be considered in the optimization process. This consideration can help to assess the uncertainties that the CSIM model derives from either formulation (structure) or data (calibration).
Place, publisher, year, edition, pages
2012. , 69 p.
CSIM, parameter sensitivity, spatial, temporal, optimization, Genetic algorithm (GA), Baltic Sea drainage basin (BSDB), Sweden, uncertainty, equifinality
Oceanography, Hydrology, Water Resources
IdentifiersURN: urn:nbn:se:su:diva-87200OAI: oai:DiVA.org:su-87200DiVA: diva2:601377
2012-12-14, 14:15 (English)
UppsokLife Earth Science