Hydraulic modeling using image-derived bathymetry

River batheymetry is crucial for the numerical modeling of floods and other river phenomena. Airborne scanning with LiDAR provides accurate river bathymetry. However, data collection using such an instrument is costly and there is a need for appropriate depth retrieval methods from a good quality areal imagery. In the present study, a regionalized linear model (introduced by Sundt et al. 2021) was used to retrieve depth for Lærdal river and then compare the resulting depth with the airborne scanning LiDAR measurements. Moreover, the retrieved and observed bathymetry were used to model flood inundation for minimum, normal, and high flood flows. Our results show that the Regional method provides a good estimation of river depth but with a slight overestimation. A comparison of flood inundation simulations between the regional method and the airborne scanning LiDAR shows the regional method provides a good estimate for the considered flood flows in this thesis