BARDOU, E., FOURNIER, F., SARTORI, M. Palaeoflood reconstruction at Illgraben torrent (Switzerland): a current need for event frequency estimation. 2003. In: V.R.Thorndycraft, G. Benito, M. Barriendos and M.C. Llasat (Eds). Palaeofloods, Historical Floods and Climatic Variability: Applications in Flood Risk Assessment, Proceedings of the PHEFRA Workshop, Barcelona, 16-19th October, 2002, p. 53-59.

Major changes occurred at the location analysed in between the middle of the warm phase corresponding to a new and intense exploitation of the territory and the middle of the Little Ice Age. Transitions between periods do not correspond with possible upsurge of flooding on the fan. It is difficult to separate effect due only to the climate or to the impact of anthropologic changes.

Comparing well documented events of the last decades (from 1932 to 2000) to the rainfall in Sierre (a town located 4 km away) no trend in the rainfall-magnitude relation appears. Furthermore the largest recorded event (in 1961) occurs with very low rainfall.

During the XXth century more than 3 events of over 100'000 m3 were reported. During years without high magnitude events it is assumed that 100,000 to 200,000 m3 of sediments are eroded from the catchments. It induces productivity rates from 25,000 to 50,000 m3/km²/a.

Prediction of “magnitude-frequency” characteristics of solid transport events in such torrents is impossible if geological or anthropological events are not taken into account (rock avalanche, fan formation, deforestation, dams, etc.). It cannot be directly related to climate change. Further, even if it is not totally sediment-constrained, this torrent shows poor correlation with rainfall events.

Almost every year a few flash floods with high suspended sediment content have been observed in the studied area. The nature of these phenomenon ranges from hyperconcentrated flows to debris flows. The total catchment's area is 10 km² but only 3 km² are intensely involved in sediment supply.

Available datas only cover the last millennium. Four main sources were used to cover the wide range of time to be investigated: 
• Dendrochronology to date flooding terraces in the current overbank channel and on the inactive part of the alluvial fan;
• Stratigraphy analysis to date palaeosoils in eroded banks;
• Radiocarbon (C14) dating (palaeosoils in eroded banks);
• Historical records (additional information on the dynamics of the system).

This study shows how a careful analysis of the system’s dynamic helps to explain changes of the “magnitude-frequency” relationship through the time. However it only gives a qualitative trend and more work is needed to get to a quantitative analysis. Analysis of all debris flow events could be helpful for “magnitude-frequency” reconstruction of geological events too. The public authorities needed a magnitude-frequency analysis to be carried out. Particularities of these phenomena (transitory, non-newtonian, 3D deposits) did not allow us to use classical hydraulic and hydrological approaches to identify their magnitude and frequency.