Human-induced climate warming, “peak water” and “peak sediment” in deglaciating Alpine catchments

It was first hypothesised by Church and Ryder (1972) that when a glaciated basin starts to witness declining glacier cover there would be an initial increase in sediment yield associated with enhanced glacier melt and sediment transport and also a landscape response as newly exposed sediment is reworked. This would reach a peak and then decline as glacial erosion and sediment transport capacity decline and the landscape slowly stabilises. The implicit assumption is that this is a transient response from an ice-erosion dominated landscape to a rainfall-erosion dominated landscape and is manifest as a period of “peak sediment” yield. In theory, due to rapid climate warming, we are now in that phase of transience but there are no datasets that describe it, we have little idea at whether peak sediment has been reached or passed, nor how this may vary between river basins with different altitudes, lithologies, geomorphic settings etc. There are no reliable multi-decadal measurements of bedload export from mountain basins let alone ones that are close enough to glaciers to be able to make links to glacier recession.

This paper uses a unique dataset based upon the high frequency (multiple times per week/year) flushing of small hydropower installations to reconstruct decadal-scale bedload export from 20 Alpine glacierized catchments in western Switzerland over the past 50 years. These catchments are heterogeneous in terms of geomorphology, geographic characteristics and glacierized area. Built to divert water from river channels for hydroelectric power generation, the hydropower installations are equipped with sediment traps designed to separate sediments from the water before it is routed to turbines or stored in artificial reservoirs. Once a threshold volume is reached, these traps must be emptied from sediments (an operation known as flushing). By combining information on the annual frequency of flushing events with analysis of the flushing operation and the volumes of sediments released, it is possible to reconstruct the evolution of bedload over recent decades.

Results show an upward trend in sediment transport since the late 1980s for most of the catchments analysed, coinciding with the onset of rapid Alpine warming in the mid 1980s. Bedload transport slowed in the 1990s, seemingly associated with a series of years with reduced up-glacier snowline recession, before accelerating again in the early 2000s. There is some evidence of peak sediment export in the 2010s. An observed relationship exists between glacierized area and sediment export: bedload export in catchments with a glacier cover lower than about 1 km² invariably appear to have gone through a shift from ice-erosion driven to summer-rainfall driven. However, some glaciers show anomalous behaviour, including emerging evidence of the direct effects and legacy of glacial overdeepenings. This can lead to site-specific, geomorphologically-influenced responses of bedload transport on top of the underlying regional-scale trend of climate warming.