Extreme flood impact on riparian vegetation dynamics in the Ahr catchment, Germany

<p>On 14^th and 15^th July 2021 heavy rainfall in western Germany, Belgium and the Netherlands caused severe floodings. The most affected area in Germany was the 86 km long Ahr river valley, which suffered from severe damage to buildings and infrastructure and where more than 130 people died. The Ahr flood exceeded a return period of at minimum 500 years. The river Ahr drains around 900 km² of the Rhenish Massif with a dendritic catchment from west to east causing differences in slope properties and covering different land uses. The flood water carried large woody debris that caused clogging in bridges of the main valley and some tributaries, some of which collapsed. This extreme event thus offers the opportunity to explore the spatial impact and characteristics of large wood on channel dynamics. This study aims to find thresholds for the initiation of large wood recruitment, dependent on catchment size, valley slopes, water quantity and land use.</p> <p>The study focuses on the whole catchment area of the Ahr river. Using general vegetation data obtained from the German national forest inventory, we quantified the type and amount of flood-affected vegetation. We adopted an NDVI (normalized differential vegetation index) based change detection approach using Landsat/Sentinel satellite data (Google Earth Engine based Hazmapper) to identify recruited live vegetation and deadwood transport during the flood. We validated this remotely obtained data with field surveys along selected valley sections.</p> <p>Large wood was predominantly recruited from the fluvial corridor in the main Ahr valley and not from the tributaries, even if those experienced heavy precipitation and surface runoff (up to few meters high discharge on the flood plains). Although we have observed transported tree trunks in those tributaries, there was no large pattern. We aspire to identify deposition areas using ortho photos to investigate a wood balance.</p> <p>Including large wood in flood modelling would improve flood hazard assessments. Remote sensing analyses offer an interim solution in this regard by helping to identify potential large wood recruitment areas and inform designing flood hazard prevention measures.</p>