Abstract. - Sensitivity of the HEC-HMS runoff model for near-surface soil moisture contents on the example of a rapid-response catchment in SW Hungary. Due to the global climate change, flash floods are one of the most significant natural hazards of today. To prevent, or at least mitigate flash flood triggered losses, numeric model based flood forecasting models are ideal tools to predict stream water levels. Model accuracy, nonetheless is profoundly influenced by input data quality. To obtain input data for the HEC-HMS distributed rainfall- runoff model, widely used for runoff forecasting, in present study we have regularly monitored ground precipitation, discharge and soil moisture in the Pósa Valley watershed (1.7 km2) in SW Hungary and data was extrapolated and upscaled to the broader area of the Bükkösd Watershed (99 km2). To test model applicability for flow time series reconstruction, the peak flow event of May 15 to 18, 2010 on the Bükkösd Stream was reproduced with the HEC-HMS. Model sensitivity was tested for various antecedent soil moisture values estimated from 2009, 2011 and 2012 in situ measured data. The output of the current research could be utilized for increasing the accuracy of rainfall-runoff model based flash flood warning systems for forested rapid response catchments that are representative for low-mountain environments under humid continental climates.
Keywords: Flash flood, soil moisture, runoff, monitoring system, HEC- HMS, model sensitivity