| Summary: | River restoration is something that is often used to help restore watercourses that were historically used for timber floating. In these restorations, instream wood (IW) plays a big role in increasing biodiversity and habitat heterogeneity in watercourses. Uncrewed Aerial Vehicles (UAVs) have proven to be useful when monitoring changes following river restoration but there is a need for less time-consuming methods to detect IW. This study’s objective was to investigate IW following a river restoration in an alluvial reach in Vargån, Northern Sweden and how well UAVs can be used to do this. Manual digitizing of IW was done in GIS from orthomosaics of three different flight occasions: pre-restoration, directly after and one year after restoration. An object- and color-based automated image thresholding was done to investigate if it could be used to automatically detect IW. The results showed that there were differences in number of IW, volume, width, and length between the different flight occasions. There was also indication that there had been movement of IW as well as changes in cluster composition, with more clusters with a higher number of wood pieces in the latest flight occasion. The automated image thresholding was able to accurately detect IW with an accuracy of 47,4 %, but it had limitations due to natural conditions. However, it showed the possibility of using automated methods to detect IW and with improvements it could become a faster and more accessible way of detecting IW in river monitoring.
|
|---|