10.4122/1.1000001553 Schaarup-Jensen, K. Johansen, C. Thorndahl, S. Uncertainties related to extreme event statistics of sewer system surcharge and overflow DTU Library, Technical University of Denmark (DTU) 2005 Hydrological reduction factor initial loss urban runoff modelling MOUSE LTS extreme event statistics combined sewer overflow flooding Aalborg University, Department of Civil Engineering NIRAS, Consulting Engineers and Planners en Conference presentation 10.4122/1.1000001554 text/xml 1 Today it is common practice - in the major part of Europe - to base design of sewer systems in urban areas on recommended minimum values of flooding frequencies related to either pipe top level, basement level in buildings or level of road surfaces. Thus storm water runoff in sewer systems is only proceeding in an acceptable manner, if flooding of these levels is having an average return period bigger than a predefined value. This practice is also often used in functional analysis of existing sewer systems. If a sewer system can fulfil recommended flooding frequencies or not, can only be verified by performing long term simulations - using a sewer flow simulation model - and draw up extreme event statistics from the model simulations. In this context it is important to realize that uncertainties related to the input parameters of rainfall runoff models will give rise to uncertainties related to the corresponding extreme event statistics. This paper illustrates this problem in a case study with two different values of one input parameter - the hydrological reduction factor - in two otherwise identical operations of the MOUSE LTS model. The use of a long historical rainfall time series makes it possible to draw up extreme event statistics covering return periods of as much as 33 years. By comparing these two different extreme event statistics it is evident that these to a great extent depend on the uncertainties related to the input parameters of the rainfall runoff model.