Implementation of GARTO as an infiltration routine in a full hydrological model
MetadataShow full item record
Climate change is expected to give more intense rainfall events, while urbanisation leads to more impervious surfaces. This combined with growing cities, increase the stress on the existing storm water infrastructure, and may result in more frequent flooding in urban areas.The land use in urban catchments affects the stormwater runoff patterns. More impervious surfaces lead to runoff hydrographs reaching the flood peaks faster, and having higher peaks. Having green spaces and water retention on the other hand, results in hydrographs with a shape more similar to natural conditions: smaller peaks reaching maxima later. To do urban hydrology assessments, e.g. for understanding the effects of land use changes, modelling of the hydrological processes in an urban area might be beneficial. Hydrological models can simulate the runoff patters from a catchment based on input parameters such as soil properties and rainfall patterns. Spatially distributed hydrological models are hydrological models that divides a catchment into smaller cells, each with its own properties and input parameters. Spatially distributed hydrological models used in Norway today do not consider the mechanisms of infiltration and are therefore not applicable for urban hydrology assessments. This thesis therefore suggests including infiltration by implementing an infiltration routine in a hydrological model. The infiltration model Green-Ampt with redistribution Tablot and Ogden (GARTO) was chosen to create the base of the routine. Further, the widely used hydrological model Hydrologiska Byråns Vattenbalansavdelning (HBV) model was chosen for implementation of the infiltration routine. The HBV model already exists in open source database gihub, under Statkraft Hydrological Forecasting Tools (SHyFT). The routine was coded separately in C++ and the results were compared to results from Lai et al. (2015). The routine was also tested against infiltration and soil moisture data of a green roof in Trondheim. The code was then implemented in SHyFT.The comparison to the results from Lai et al. (2015) gave satisfactory results. The routine also managed to match parts of the results from a green roof. A sensitivity analysis on the soil properties shows that saturated conductivity is the most sensitive soil moisture constant. Further the routine match good with expected infiltration responses.