Towards a non-intrusive method employing digital twin models for the assessment of complex large wood accumulations in fluvial environments

Quantification and assessment of large wood (LW) accumulations in fluvial systems is still considered difficult due to the complex nature of wooden deposits. Although knowledge about volumetric measures and porosity parameters of LW accumulations is crucial for the prediction of hydraulic and geomorphic effects, it has not yet been possible to obtain accurate measurements. These limitations are mainly based on a lack of applicable sensing technologies available in the past. In the present study, a close-range aerial surveying technique (Structure from Motion (SfM) photogrammetry) is applied for generating 3D replicates (digital twin models) of wooden deposits, enabling their volumetric assessment. In addition, manually conducted volumetric measurements of corresponding prototype LW accumulations help to improve and calibrate the SfM-derived estimates. For the first time, precise porosity parameters for LW accumulations, ranging from 52.5 to 83.2%, are provided. In addition, a novel parameter – the packing arrangement – is used, which describes the structural alignment of individual elements in the LW accumulation and benefits porosity estimates based on the applied 2.5D and 3D photogrammetric approach. Accordingly, randomly and loosely organised LW accumulations allow for a high penetration depth of the 3D approach, resulting in a more accurate estimate of the actual porosity, as the 3D volumetric estimate approaches the solid wood volume of the corresponding LW accumulation. An empirical approach has been developed for future approximation of LW accumulation porosity, without the need of knowing the solid wood volume. With the present work a significant improvement of our understanding in employing a non-intrusive sensing technique is provided, linked with manually conducted field measurements of the solid wood volume of LW accumulations. Our study contributes to an improved data acquisition and processing plan, which represents a further important step towards a systematic assessment framework that is urgently needed by river managers and engineers to better evaluate and manage LW in fluvial systems.