“…Generally, the uncertainty in flood inundation modeling can be categorized into seven major types: (1) topographic data, (2) hydrologic data, (3) data for preliminary estimation of roughness, (4) method applied for final roughness calibration, (5) method applied for estimation of maximum flows, (6) structure of the applied model and (7) transition of the model results to the maps (e.g., Refsgaard and Storm 1990;Cook and Merwade 2009;Calenda et al 2009;Liu and Merwade 2018). Some of the above-listed elements become less uncertain due to the development of the technology, e.g., increasing accuracy and resolution of LIDAR data for DEM elaboration (e.g., Gilles et al 2012;Sampson et al 2012;Walczak et al 2013;Laks et al 2017), involvement of satellite and remote sensing data (e.g., Jung et al 2014;Arseni et al 2017), elaboration of more detailed databases on flood events (e.g., Kundzewicz et al 2017) and broader availability of more accurate hydraulic models (e.g., Szydłowski et al 2013;Gąsiorowski 2013;Brunner 2016a;Kolerski 2018). A specific problem is the uncertainty related to the channel and floodplain roughness (e.g., Dimitriadis et al 2016;Pappenberger et al 2008;Engeland et al 2016;Liu and Merwade 2018).…”