Eruption of Mount Sinabung, in North Sumatra, Indonesia

Photo from CNN.Com by Einsar Bakkara/AP
in the cited article in text
(if I read the credit correctly)

Mount Sinabung, an Indonesian volcano dormant since 1600 came to life in 2010 and, on Saturday, spewed forth pyroclastic flows that killed at least 14 people. Tragically, it appears that these people had been evacuated last summer and only the day before this eruption, had been allowed to return to their villages. The Wiki site for Mount_Sinabung appears to be updated in a timely way, so I won't go into details here.
     It is difficult to tell what the source of the erupted material is in detail, but from photos of the volcano (a classic beautifully conical stratovolcano) and the lack of any indication of lateral bulges on the flank, a good assumption is that the flows are originating in a summit crater. A question/assumption, is whether they are being driven by volatiles (presumably H3) from magma or whether or not groundwater is involved. According to the Wiki article, in late December, a lava dome had formed on the summit.
     The eruption gas/ash material from lava domes results in eruptions known as "Pelean" or "Merapi"-type pyroclastic flows. Two processes contribute to the high-velocities observed from such eruptions: gravitational collapse (supplemented by heating and expansion of entrained air), and sudden expansion of pressurized gases from inside the domes. If gravity controls the energy transfer, then areas affected can be predicted on the basis of topography. If gas expansion adds a significant contribution, which is likely in the proximal region around a dome, then velocities beyond those acquired by acceleration in a gravitational field, exist, and these imply that much larger areas are at risk than might be predicted from the gravitational forces alone.
     In 1993, Jonathan Fink and I published a paper "Estimate of pyroclastic flow velocities resulting from explosive decompression of lava domes," Nature, v. 363, pp. 612-615, 1993.  In this paper we examined the two processes above, and concluded that the decompression process produces velocities comparable to those acquired by gravitational accelerations. In snapshots, such as that in the photo in this post, my guess is that the flow is clearly already some distance down the slopes of the volcano where it has assume the classic profile of a dense gravitational flow with air entrainment. More proximal regions have already been hit, and are, apparently, where the casualties have occurred. With the complicated sequence of recurring explosions/eruptions from the summit, it may never be possible to reconstruct the dynamics of the flows in the proximal region.