The interaction of ash flows with ridges

Using both laboratory experiments and theoretical models, we examine the different flow regimes that may develop when an ash flow encounters a ridge. For cery small ridges, all the flow may pass over the ridge. For intermediate-size ridges, the flow may be partially blocked, with a fraction of the flow reflected upstream as a travelling bore. In this case, the remainder of the flow, which does pass over the ridge, is hydraulically controlled at the ridge crest. Finally, if the ridge is sufficiently high, then the flow will be totally blocked. New laboratory experiments show that the sedimentation patterns associated with these flow regimes may be very different. Most importantly, flows that involve partial blocking and the formation of upstream propagating bores display enhanced sedimentation upstream of the ridge, analogous to valley-ponded and caldera-fill deposits. In contrast, under some circumstances, if the flow is able to scale a reidge, the deposit may be relatively unaffected by the presence of the ridge. The minimum ridge height that leads to total blocking of the flow increases with mass eruption ratge and has a complex variation with distance from the sourve. In a one-dimensional channel, the minimum ridge height that causes blocking increases with distance downstream. This is because the flow becomes less dense through sedimentation of particles and entrainment of air and so requires less energy to scale a ridge of a particular height. In axisymemtric flow, the minimum ridge height initially decreases with distance downstream as the flow spreads radially, but susseqnetly increses as the flow becomes les dense through sedimentation and entrainment. A new quantitative model of dilute ash flows propagating over ridges indicates that flows with mass fluxes in excess of 10⁸-10⁹ kg/s can partially scale barriers as high as 1000m at distances of tens of kilometres from the source, whereas smaller flows are likely to be totally blocked by such an obstacle. Our results shed new insight on the possible long-range transport mechanism of several large flows including the Ata, Fisher and Aniakchak pyroclastic flows.