Evaluation of Streamflow Predictions in an Ungauged Swedish Catchment : A Study of Håga River

Sammanfattning: The Håga river, located west of the Swedish city Uppsala, is currently without a proper gauging station. Knowing the streamflow is nonetheless important to, for example, be able to calculate the nutrient transport in the river. This project aimed to evaluate different indirect methods of streamflow estimation to investigate how they perform, in particular in relation to SMHI’s S-HYPE model. Two of the methods used were based on transferring streamflow of nearby catchments to Håga, either by using relationships between the mean and standard deviation of the streamflow time series (MOVE), or by simply scaling relative to catchment size (DAR). Furthermore, a hydrological model, HBV, was calibrated for Håga using different amounts and types of calibration data. All the methods were then compared to streamflow data from a previously active gauging station in Håga.  It was found that the overall best method to estimate the streamflow in Håga was using the MOVE method with one particular donor catchment. However, the performance of the simpler MOVE and DAR methods varied a lot from catchment to catchment. HBV was found to be able to produce better performing simulations than S-HYPE, despite being a simpler model. Even HBV-calibrations using alternative or limited data could perform rather well, although rarely at the level of a calibration utilising all available streamflow data. A big uncertainty of the study was the fact that the most recent available validation data for the Håga catchment was from two decades ago, when the old gauging station was decommissioned. Most likely the methods that worked well during the 90s would work well today as well, but this is a matter that could be studied further.