Hydrological Modeling for Determining Flooded Land from Unmanned Aerial Vehicle Images—Case Study at the Dniester River

Abstract

In recent decades, in the Pre-Carpathian region of Ukraine during the summer period, floods and flood events became more frequent. They were accompanied by significant economic and environmental loss. Especially powerful were the floods of 2008 and 2020, but the floods in 2014 and 2016 also had destructive consequences. Therefore, the study of river channel processes, river stability and assessment of flooded land areas due to floods is an urgent problem. The aim of the study is to propose a methodology for hydrological modeling of sections of riverbeds with complex morphometric and hydrological characteristics. The construction of a digital elevation model (DEM) and the selection of the distance between the cross-sections, as well as the determination of the Manning coefficients, have the greatest impact on the accuracy of the modeling, so these factors should be given maximum weight when calibrating the model. The object of the study was the section of the Dniester River in Ukraine in the place of transition from the foothill part of the channel to the hilly–marshy part with complex meandering and significant shifts of the river. The methodology of hydrological modeling includes three principal components: construction of the DEM, determination of the type of underlying surface and determination of the level of water rise in the riverbed. The research was carried out on the basis of imaging from unmanned aerial vehicles (UAVs). In 2017, the imaging of a section of the Dniester riverbed was carried out in the summer during a period of significant vegetation growth, which affected the accuracy of determining the heights of the model points. According to the results of this imaging, the residual mean square (RMS) for determining the heights of the points exceeded the permissible value of the RMS (0.25–0.3 m) by two times. In 2021, imaging was performed in the autumn period when there was no leaf cover. The RMS of the DEM for 2021 imaging was 0.26 m. According to the results of the survey in 2017 and 2021, orthophotoplans were created, which were used to determine the planned displacements of the river bed and clarify the Manning coefficients, which characterize the roughness of the underlying surface. The value of the water level rise was obtained on the basis of the graph on the date of the maximum rise of the water level on 24 June 2020 according to the hydrometeorological station located near the selected area. The result of the research is hydrological modeling using the HEC-RAS module for a site with complex hydrological and morphometric characteristics on the date of the maximum water rise. It was established that in order to achieve the required accuracy of the DEM, imaging should be carried out in the leafless period of the year, since the accuracy of constructing the DEM decreases by half during the growing season. On the basis of the obtained orthophoto plans, a methodology for determining refined Manning coefficients was developed, which allows taking into account changes in the underlying surface of the channel area. The area of the flooded area was calculated based on the level of water rise during the 2020 flood.