Evaluating fluvial lateral connectivity from topo-bathymetric LiDAR data and Gaussian mixture model.

The concept of connectivity is a hot-topic in many scientific fields for 20 years ago. This concept refers to the capacity of a system to let flow fluxes of matters, energy (Harvey, 2012) and organisms (Heckmann et al., 2018). In fluvial environments, this connectivity is often studied through the prism of hydrologic connectivity but recent studies focus on sediment dynamics, as it is fundamental factor for a stable and diversified fluvial mosaic (Brierley et al., 2006; Wohl et al., 2019). However, connectivity is easily negatively affected by disturbances such as bed-degradation. Many European rivers experiment a decrease in their connectivity between main and lateral channels (Grill et al., 2019).

Studying the connectivity of riverine ecosystems can be challenging and numerous indices has been developed (Heckman and Vericat, 2018). Many of them focus on the structural aspect of the connectivity or, in other words, they describe the morphological configuration of a river reach. The functional aspect of the connectivity, which refers to the modality and the frequency of connection, is rarely explored.

We propose a method to estimate the connectivity between lateral channels and main channel of a large river using topobathymetric LiDAR data. To achieve this goal, we mobilize a Gaussian Mixture Model (GMM) to identify elevation planes that best fit to the main and lateral channels (Andréault et al., 2024). Then, the gradient between both units, representing the structural connectivity of the reach, is calculated. Functionality of the site is approximated by comparing the waterlevel of diverse hydrological events with the surfaces associated to the median elevation of lateral channel. Compilation of the different comparisons approximate the frequency of inundation or in other words the functionality of the site. However, to better estimates both connectivities, results are analysed according to the entrance of the channel, which is the main limit to bedload transport.

Results of the study show a variability of the structural connectivity along the Loire river according to its main incision sectors. It varies from highly connected to highly disconnected. In-depth analysis of the geomorphological situation highlights the presence of morphological units at the entrance of lateral channels that might be responsible in the case of high structural disconnectivity. The analysis of waterlevels compared to mean plane associated to lateral channels revealed that systems connectivity were not much affected in case of flood but that situation change in case of low flow and average discharge values.

This work highlights a variety of situation and a sensitivity of the methods to characterize objectively the connectivity of the river reaches. It is an innovative approach able to process dense data. The relative ease of the method would also allow river management entities to hierarchize operations of management (re-opening of side channels). In that sense, this work could be of interest in a global change context.