<p>The highland cirques mostly created by nivation and glacial exaration take large areas in mountains and have a significant role in the sediment transit of the basins. The approximate view on the connection of cirques and low levels in the sediment flow could be given with the sediment connectivity index analysis. We study the spatial distribution of the index for typical ice cirque – the Koiyavgan cirque near the join of the Main Caucasus Range and its offshoot (the Gumachy range). This area is located in the tops of the Adyl-Su valley (left side of the Baksan river basin). In August 2020, we got a high-resolution orthophoto image (13+ cm) and digital elevation model (27+ cm) from aerial photography. The territory located in the elevation range from 3230 to 4022 m. Geological conditions: gneiss, metamorphic shale and basic dark coloured igneous rocks. There is no developed vegetation cover. Typical post-glacial cirques topography includes (top-down): mountain tops, very steep bedrock slopes, colluvial footslopes and fans, cirques bottom (moraine ridges with dividing valleys, craters from melting of the in-moraine covered ice etc.) with fluvial, avalanche and creep post-shaping, and bottom surface break as analogue of riegels in glacial trough valleys. The connectivity index (CI) after Cavalli et al. [2013] is very dependent on initial DEM resolution, from the method for filling mistaken depressions, from window size for computing intermediate geomorphometric variables (e.g. roughness index), from choice in flow impedance variable, from area coverage and terrain diversity and others. We compute connectivity index with the parameters: 1) DEM resolution – 27 cm; 2) impedance variable – terrain roughness index (standard deviation of elevation) with window 7*7 cells; 3) standard filling method used in the ArcMap (filling local depression without any limitations on maximum depth); 4) range of impedance values before normalization (partially related to area coverage) is from 0 to 72 m. In the some buffers from the channel network the connectivity index generally grows in the top-down direction. Greatest spurt of the CI values relates to the cirques low border - the riegel (3300 m asl). There are two levels characterised with low values of the CI: 3550 m and 3750 m. The first one is backside of cirques bottom with relatively low flow accumulation area and low-moderate slopes (0-25°), the second one is mountain tops with high steep slopes, but with lowest flow accumulation. For different geomorphodynamical zones the threshold of IC where sediment transit turns into sediment accumulation has differ values: for example, -2.3 for colluvial fans and -2.5 for alluvial fans (p-value for differences significance « 0.01). Maximum values of CI (quantile: the top-95%) for accumulative positions again are -1.27 and -0.72. Its means, those accumulative processes areas with different mechanics of the deposition may be delineated with using non-constant CI values only. The potential of sediment flow connectivity modelling for high mountain isn’t exhausted, but its application needs wide discussion and calibration.<br>The study was supported by the Russian Science Foundation (project No. 19-17-00181).</p>

The volumes of loose-clastic material moved by modern earth surface processes were quantified for a small mid-mountain catchment area ( F = 1,86 km2), located in the Baksan River basin using a number of independent methods (modeling of erosion and geomorphologic mapping). The total volume of removed material was estimated through the assessment of the thickness of bottom sediments accumulated in a reservoir located in the outlet section of the catchment area. The rates of linear and rill erosion have been found using the results of repeated UAV surveys and the pin method. The rates of rockfall processes are estimated on the basis of repeated ground-based laser scanning. Additionally, published data on the rates of earth surface processes in the mountains were used. Based on the results of field mapping and interpretation of space images, a map of the spatial pattern of leading earth surface processes was compiled. A digital terrain model of the catchment was constructed on the basis of detailed UAV survey. The volume of loose clastic material delivered to the reservoir was estimated using data on the average annual rates of leading earth surface processes, a map of their distribution over the catchment area, and the connectivity index. It has been established that the average annual denudation rate within the catchment area is 0,16 mm/year. At the same time, only about 3-10% of the products of denudation are delivered to the reservoir. The sediment yield, calculated on the basis of the volume of sediments in the reservoir, is 514 t/km2 per year. It is comparable to the sediment yield of small rivers draining the mid-mountain belt of the Alps with similar bedrock lithology, which was calculated using the results of long-term observations of water and sediment discharge at gauging stations.

Soil erosion and sediment export from hillslopes are significant problems associated with agriculture, especially in parts of the world where society is already living in extreme environments. In particular, mountainous environments remain severely understudied, with only a few runoff and sediment transport measurements available. It is necessary, therefore, to develop and validate independent methods that do not rely on long-term observations at gauging stations. Here we used three independent methods to predict soil erosion and associated sediment yield (SY) from a 1.84 km² basin in the North Caucasus. The first part concerns assessing the sedimentation rate, which was made using in-situ measurements of volumetric sediment deposition rates. Secondly, we look at the connectivity of sediment sources and the lake. A combination of remote sensing data and field surveys was used to estimate sediment connectivity and erosion mapping. The third part regards the computation of soil erosion using the Revised Universal Soil Loss Equation (RUSLE). There are three major findings in this study that help us understand sediment redistribution patterns in mountainous areas. First, based on the lake sedimentation rate, we found that the mean annual area-specific sediment yield is 514 (95% CI, 249–839) t km− 2 yr− 1. Similar results were obtained from the erosion mapping (i.e., a map of erosion processes) — 428 (95% CI, 322–546) t km− 2 yr− 1. Secondly, the spatial distribution and rates of the erosion processes suggest that sheet and rill erosion are responsible for ca. 40% of total sediment export, slides and rockfalls — 18%, while the rest is removed by soil creep. Additionally, the RUSLE-based modelling of sheet wash and rill erosion has highlighted the areas most prone to soil erosion. The corresponding mean annual soil erosion rate of 1.59 mm yr− 1 was very close to the results obtained from the literature review.

No abstract