Hybrid DTMs derivided by LiDAR and colour bathymetry for assessing fluvial geomorphic changes after flood events in gravel bed rivers (Tagliamento, Piave and Brenta Rivers, Italy)

Risk management and flood protection are frequently assessed through geo-morphometric evaluations resulting by fl oods events. If we aim at elevation models with high resolutions and covering large areas, airborne laser imaging detection and ranging (LiDAR) surveys can represent a good compromise among costs, time and uncertainty. The major limitation of the non-bathymetric LiDAR surveys consists in the detection of wet areas. Indeed, accounting for more than 20 cm of water depth, LiDAR signal increases its error exponentially. In this article we present a comparison of the results concerning the application of a colour bathymetry methodology for the production of hybrid digital terrain models. These elevation models were derived by merging LiDAR data for the dry areas and colour bathymetry for the wet areas. The methodological approach consists in a statistical regression between water depth and RGB band intensity values from contemporary aerial images. This methodology includes the use of fi lters to reduce possible errors due to the application of the model and to estimate precise ‘in-channel’ points. The study areas are three different human-impacted gravel-bed rivers of the north-eastern Italy. This methodology has been applied in three sub-reaches of the Brenta River, two of the Piave River and two of the Tagliamento River before and after relevant fl ood events with return intervals of ≥10 years. Potentials and limitations of the applied bathymetric method, the comparison of its use in different fl uvial contexts and its possibility of employment for geo-morphometric evaluations, were then tested. DGPS control points (1841, 2638 and10 473 for the Brenta, Piave and Tagliamento Rivers, respectively) were fi nally used to evaluate the accuracy of the wet areas. The results showed that, in each model, the wet areas’ vertical errors were comparable with those featured by LiDAR data for the dry areas.