Predicting channel dimensions and bed materials in intermittent Mediterranean rivers.

Floods cause severe natural disasters over the world generating property and infrastructures damages, poverty and loss of human life, among others. Mediterranean watersheds are especially sensible to floods due to their typical drainage basin features (steep slopes, short concentration times, complex orography, etc.) and the high rainfall intensity typical of convective systems. Knowing channel dimensions and other fluvial morphological features is key to (i) understanding the morphological evolution of the fluvial system in response to changes in land use and climate, and (ii) as an input for hydrological and erosion modelling. The objective of this study is to develop a simple method to obtain reliable estimates of channel dimensions and granulometry of bed material, based on statistical relations with catchment characteristics (e.g. topography, land use, soil properties, lithology, precipitation, connectivity indicators).

First, channel dimensions were estimated based on GIS analysis using a high resolution digital elevation model (2x2m) and ortophotos (50 cm resolution) for the Upper Segura River catchment, a Mediterranean catchment of 2,592 km² located in the southeast of Spain. These estimates were validated with field measurements of depth and width of bankfull channels in the catchment headwaters. This comparison revealed that there was generally good agreement between channel dimensions obtained with the GIS method and those observed in the field for all evaluated channels (depth: RMSE=0.06; R²=0.93; width: RMSE=0.59, R²=0.45), although with better results for dry channels than for channels with continuous water flow. At each field observation, we also took sediment samples and characterised the granulometry of the channel bed material in the laboratory through dry sieving. Preliminary results show that bed material is composed mainly by gravels (67.8%), followed by sands (31.1%) and clays and silts (1.1%).    

Next, channel dimensions, obtained using GIS analysis across the entire catchment, and granulometry of bed material were used as dependent variables in advanced statistical analyses (such as machine learning algorithms) at sub-basin scale, with catchment characteristics as independent variables. The outcome of this analysis is now being used to make spatially continuous predictions of channel dimensions and granulometry of bed material at the catchment scale. This information will then ultimately serve as input for a coupled model that simulates channel hydraulics and morphodynamics at the catchment level.

Keywords: geomorphology; sediment yield; depth and width of bankfull channels; Mediterranean environment; GIS based tools; southeast of Spain.

We acknowledge funding from the Spanish Ministry of Science and Innovation and ‘Agencia Estatal de Investigación’ (PID2019-109381RB-I00/AEI/10.13039/501100011033).