Debris-flow fans are ubiquitous features of mountainous areas worldwide. The flows that build these fans are a major hazard, while the fans themselves form a potentially-useful archive of past debris-flow activity.

To assess both the hazard and the depositional debris-flow record, however, we must understand the ways in which debris-flow fans evolve as they are fed by repeated flows.

Fan evolution depends upon (1) the sequence of flows of different volumes that are supplied to the fan head, (2) the changes in flow volume as flows entrain or deposit material across the fan surface, and (3) the frequency and locations of channel avulsion between different parts of the fan surface.Field observations show that flows can both erode and deposit as they traverse the fan surface, but the probability of erosion appears to scale with flow thickness and thus with basal shear stress.

Debris-flow avulsion can occur anywhere on the fan surface, leading to avulsions at different scales. Many, but not all, avulsions are triggered by deposition of a ‘plug’ of material in the channel during one or more relatively small flows, so that subsequent larger flows are diverted into a new sector of the fan.

Numerical and analogue fan evolution experiments show that avulsion frequency depends both on the flow volume distribution and on scaling parameters that represent differences in flow composition and rheology.

The results provide a framework by which we can start to read the topographic and stratigraphic record of debris-flow fans for clues to past flow behaviour, and to anticipate the potential effects of future flows.