A new dunetracking tool to support input parameter selection and uncertainty analyses using a Monte Carlo approach

Abstract. Precise and reliable information about bedforms, regarding geometry and dynamics, is relevant for many applications – like ensuring safe conditions for navigation along the waterways, parameterizing the roughness of the riverbed in numerical models, or improving bedload measurement and monitoring techniques. There are many so-called dunetracking tools to extract this information from bathymetrical data. However, most of these tools require the setting of various input parameters, which in turn influence the resulting bedform characteristics. How to set the values for these parameters and what influence they have on the calculations has not yet been comprehensively investigated. This is why we introduce a new dunetracking tool, which is able to quantify the influence of varying input parameter settings by performing a Monte Carlo Simulation. The core of the tool is a combination of the two existing applications Bedforms-ATM (Guitierrez, 2018) and RhenoBT (Frings et al., 2012), which have been extended by adding additional features. A wavelet analysis has been adapted from Bedforms-ATM while a zerocrossing procedure and a cross correlation analysis have been implemented based on RhenoBT. The combination of both tools enables a more accurate and sound procedure, as the results of the first step are required input parameters in the second step. By performing a Monte Carlo Simulation, comprehensive sensitivity analyses can be carried out and the possible range of results is revealed. At the same time, the high degree of automation allows the processing of large amounts of data. By applying the tool to a test dataset, it was found that bedform parameters react with different sensitivity to varying input parameter settings. Bedform lengths appeared to be more sensitive (uncertainties up to 50 % were identified) than bedform heights. The setting of a window size in the zerocrossing procedure (especially for the upper layer of bedforms in case secondary bedforms are present) was identified to be the most decisive input parameter. Here, however, the wavelet analysis offers orientation by providing a range of plausible input window sizes and thus allows a reduction of uncertainties. By choosing values outside this range, divergence behavior could be observed for several resulting bedform parameters. Concurrently, the time interval between two successive measurements has proven to have a significant influence on the determination of bedform dynamics. For the test dataset, the faster migrating secondary bedforms were no longer traceable for intervals longer than two hours. At the same time, they contributed to up to 90 % of the total bedload transport, highlighting the need for measurements in high temporal resolution in order to avoid a severe underestimation.