In the Caribbean, when public policy goes wrong, it is as a rule the poor, typically black descendants of slaves who pay the price: in further impoverishment, in loss of livelihood, in dislocation and distress, and -- far too often: in blood.
So, given our region's sad and unfinished, painful history: slavery, colonialism, genocide of native peoples and general oppression, incompetence and injustice, our decision-making and decision-influencing classes have special burdens and face special challenges. That holds for the rising class of educated Caribbean leaders; it holds for the international, expatriate experts and advisers who do so many technical consultancies here; and it -- especially -- holds for the remaining colonial powers in those territories that (for one reason or another) have not obtained independence.
So, with three-fold force, it holds for Montserrat in the face of a lingering volcano crisis that has now played out over eleven years.
In that light, it pains me to have to summarise and comment below on the latest events over the past few days and weeks, as the growth rates in the dome again surged and led to a much escalated threat to our largest current township, Salem and its environs. But, integrity demands that we must face unwelcome but credible facts, seek the truth, and set out to act with that prudence that stems from wisdom:
1] By early 2005, the public discussion on the state of the eruption since 1995, was (in light of reported quiescence since the July 2003 major dome collapse) to set the criteria for declaring the eruption officially over. Public policy was shifting gears from emergency management and recovery to rebuilding and redevelopment. An award-winning video, Volcano Generation, was produced by a journalist at the local access cable TV channel, on how a new generation was going to "take back" the island from the volcano.
2] However, in a press conference held in the July 2005 10th anniversary volcanology conference, Professor Glen Mateoli [sp?] of the USA dropped a bombshell: his monitoring of the swelling of the mountain edifice over the period since the 2003 collapse indicated that the eruption was anything but over or fading away. Indeed, he predicted that within a short time, the mountain would return to active eruptive behaviour. Although this was initially received with public skepticism and even dismissals by other volcanologists at the conference and by public officials and public relations officers, within weeks, he was justified by events.
3] After this, the hosts of the Let's Talk local talk show [disclosure: myself included] highlighted Dr Mateoli's remarks, brought to the public attention the Okada tetrahedral model for transparent management of potential volcano crises presented at the 10th anniversary conference, and called for managing the crisis robustly in light of a full fan of credible scenarios, from optimistic to "most likely" to pessimistic. We pointed to the longstanding pattern of an optimistic bias in public pronouncements and management, and its consequences. For this, we were publicly rebuked by the Government media figure who hosts the weekly volcano interview with the Observatory staff, for "misconstruing" the situation. A journalistic, openly and acknowledgedly on the record interview with Professor Okada which was played, unedited, on the Let's Talk programme as the hook for our discussion, was also publicly misrepresented by the then head of the MVO as a violation of the privacy of a conversation with a scientist.
4] This unfortunate -- and never apologised for or retracted -- incident highlights the precise point made by Prof Okada: scientists, officials and the independent media should jointly collaborate to provide a three-legged, transparent platform supporting the public's need and right to know about the risks and hazards they face in living with a potentially active volcano.
5] Also, when I subsequently discovered a copy of the 1987 Wadge-Isaacs report, to my amazement, I discovered that in points 9 - 10 of its two-page executive summary -- right in the front of the document, it highlighted the need for transparently reckoning with just the sort of fan of scenarios we had discussed on Let's Talk:
. . . 9. [Scenario-based approach:] We suggest that emergency planning should allow for three very different types of eruption:
1. [Relatively Optimistic case:] A small eruption within English’s Crater. The only community that would be directly threatened by pyroclastic flows would be Long Ground which should be evacuated as soon as the eruption starts. There would also be a hazard from mudflows, specifically in Fort Ghaut in Plymouth.
2. [Intermediate case:] A moderate to large eruption. Most of southern Montserrat should be evacuated. The only two communities that would be relatively safe are Richmond and Cork Hill. The sequential hazard zone map gives a guide to the priority of evacuation in the event of such an eruption.
3. [Pessimistic case -- more or less what has now happened several times] A collapsing dome/lateral blast eruption. This is a very remote but dangerous possibility. The conditions leading up to it may be detectable in advance and once diagnosed immediate evacuation of the relevant 180 degree sector of the volcano would be required.
10. [Recommended robust policy response:] We suggest that some consideration be given to strategies for mitigating the damage done to the island by the loss during the eruption of the centralized facilities at Plymouth. Land use planning should take specific account of the mudflow hazards in the ghauts and valleys and the high risk area east of English’s Crater. [Bracketed, explanatory comments added]
6] From early August 2005 a new dome emerged, at first growing slowly, but then accelerating sharply in February 2006. On May 20, without warning (and the sirens did not sound, similar to the failure in the fatal Central Corridor in 1997!) the dome, then about a hundred million cubic metres [100 mn cu m] collapsed, mostly falling to the east. However, some very heavy ash falls occurred to the west. A hot ash surge cloud rushed out to sea on the eastern [open -- no crater wall] side of the volcano, and inexplicably bounced back inland, panicking those who had gone to the Jack Boy Hill lookout site to watch the events. A surprise bubble of irritant gases also flowed to the west, discomfiting , many residents of Salem, which is on the flanks of the Belham Valley [a major radial valley that runs to the West from the volcano edifice]. (Across the cycle of the eruption, scientists, officials and the public alike have repeatedly been caught by surprise by events.)
7] Within days, a new dome was rapidly growing, and in December 2006, it was about 150 mn cu m. For much of this period, the growth rate of this dome was said to be about 9 cu m/s, or about 3/4 mn cu m/day. So as Christmas approached, we faced a large dome, growing at an unprecedentedly rapid average rate for the overall eruption. The "good news" was that, even though growth shifted to the NW side of the dome closest to the Belham valley, Salem and environs, there was still a gap between the dome and the crater wall; which provided a barrier to collapses and hot ash and debris flows.
8] These flows are called pyroclastic flows, aka nues ardentes. On May 8, 1902, such a flow down a river valley in Martinique led to a spreading hot ash cloud that wiped out St Pierre, perhaps 2 miles to the South of the valley; killing 28,000 people in maybe three minutes. It is worth pausing to cite a brief summary of how that happened:
St. Pierre was destroyed by nues ardentes (glowing avalanches) which detached from the growing lava dome and crashed under gravity down the flanks of the volcano. For the most part, these avalanches were topographically controlled by preexisting valleys, such as that of the Rivière Blanche . . . but some were big enough to engulf St. Pierre on their margins.
9] Indeed, it is worth noting here, that in the original, exemplary 1986-7 Wadge-Isaacs study and report cited above on the risks posed by the Soufriere Hills Volcano in Montserrat, a thought-provoking figure superimposes a map of the deadly Martinique flows on May 8, 1902 on the map of the Belham valley, obviously engulfing Salem and Environs. So, while indeed, computer modelling of topographically-controlled pyroclastic flows and surge clouds [e.g. the PYROFLOW model currently in use at the MVO] are an important tool to guide hazard and risk assessment, we should recognise that -- as was acknowledged in the public presentation of the results, such models are subject to error, surprises on the ground and uncertainties so should be used with caution. (For, citing and expanding my public warning of Oct 30, 1995: Volcanology at its root is an observational science, as the very name of the Montserrat Volcano Observatory should remind us.)
10] Then, early in the new year, rumours began to circulate that the Vue Pointe Hotel at the mouth of the Belham was being evacuated. Shortly thereafter, the sirens sounded and those who could tuned to the local radio station, ZJB. We heard plainly worried officials informing us that the dome had surged in its growth, and was now estimated as "over 200 mn cu m." [NB: 50 mn cu m growth in 3 weeks or so corresponds to an average growth rate of 25+ cu m/s; or, ~ 2.4+ mn cu m/day.]
11] Worse, the crater rim on that side had now been overtopped by the surging dome, and pyroclastic flows were running down the Gages Valley and also Tyre's Ghaut. Already, runouts had reached a kilometre, about 3/5 of a mile. Thus, both Plymouth and the flanks of the Belham were at risk. The daytime entry zone was immediately cancelled, and an evacuation of the residents on "the flanks" of the Belham was announced. However, within hours this last was being dialled back to getting prepared to evacuate, and the next afternoon we were told by the [UK-appointed] Governor that a proclamation was being prepared for issue "within 48 hours."
12] By Friday evening, we learned from the Volcano Observatory during its weekly briefing, that runouts on the Gages front had reached 4 km [~ 2 1/2 miles], and on the Tyre's Ghaut front, 1.5 km [~ 0.9 miles]. Consultations with Dr Willy Aspinall, a leading expert on hazards of volcanoes and who had been present for the June 25, 1997 event that killed perhaps 19 people, led to a warning that the fact that current flows were small should not lull us into ruling out the same sort of sudden large scale, lethal flow that had happened that fatal day on his watch. Nevertheless, we were informed by officials, that only fifty households on the flanks of the Belham were to be evacuated at present, and that a high-powered official delegation [including the Governor] was visiting the houses one by one to explain the situation.
13] By this time, the Vue Pointe Hotel at the mouth of the Belham [and about a mile downstream from the communities on the flanks of the valley] had been evacuated two full days previously. This gap raises some obvious questions:
 If the risk to Vue Pointe was enough to trigger an immediate evacuation (on common reports, even before alerting the general public), given the known and publicly admitted uncertainties in volcano hazard modelling why was it not serious enough to call for a similar implemented evacuation of the communities located on the banks of the Belham?
 If on the other hand -- notwithstanding the known facts that (a) pyroclastic flows can spread out on the sides of constraining valleys, that (b) large ones can happen with little warning, and that (c) these flows easily run up to 60+ miles per hour so cannot be outrun on foot or in a car -- the degree of assessed risk for residents on the flanks of the Belham was low enough for them to remain in their homes pending an official order that takes several days to prepare, why was the Hotel -- a mile or so further away from the source of danger -- immediately evacuated?
 Or, more directly: why is it that ALL of "the 180 degree sector" facing the W side of the dome, up to at least the line of the Nantes River identified as potentially at serious risk was not "immediate[ly]" evacuated, as per the 1987 recommendations?
 Given the sad history of what has already happened with this crisis in June to August1997, and the underlying socio-cultural and historical issues raised at the head of this post, how can this apparent contradiction be credibly explained?
But, the real challenge is not fault finding but to identify a way forward that gives us some hope for a sustainable and desirable future.
a] Salem: I suggest, with all due respect, that we face the obvious: Salem and environs are at serious risk, and in a context that the mountain remains persistently eruptive, should be excluded from future development plans until the situation clarifies itself. We should not put in any further new infrastructure and public institutions into that township, and should move major institutions such as the High School, Community College and UWI Extramural Centre. [Given the public statement by Dr Aspinall et al in a SAC presentation in March 2006, on being asked to define the "safe" North, long term investments should only be put in North of the Lawyers Ghaut, just north of the Woodlands beach.]
b] Evacuations: These tend to cause impoverishment as people have to abandon residential property and business places, with their contents. The inhumane shelter conditions of 1995 - 7 led people to drift back into plainly unsafe areas to farm and to get a chance to rest in comfort, desensitising them to the hazards, which contributed to deaths. Thereafter, it contributed to depopulation. Consequently, it is suggested that:
[i] We should immediately purchase some of the unused shipping containers in the Port, and use them to create a warehousing facility for residents and small businesses in the Salem area, so that people may store their movable property. [As soon as possible after that, prefabricated warehousing -- perhaps ex UK military? -- should be brought in to upgrade the facilities.]
[ii] Following the example by Mr John Ryan who recently built a U-shaped walk-in plaza with eleven  business units on a half-acre site in Brades, we should use modular construction technologies to rapidly create small business plazas that can house the small businesses of Salem and Environs. A similar plaza could accommodate one or more evacuated schools.
[iii] A similar approach could create small cottage estates, which could serve as housing with dignity for evacuees. (If one were built right away, it could house a significant fraction of the currently expected evacuees, and would show a commitment that would discourage evacuees from simply abandoning the island. A further major flight from the island would have devastating social and economic impacts, as would a forced exodus of the many Caricom Nationals who have joined our workforce in recent years.]
[iv] There is enough land available in the North that targetted farming areas could be developed for the evacuated farmers of Salem. This would avert one of the acknowledged contributory factors to the deaths in 1997.
c] Public Policy and Public Relations: We should immediately and decisively move to a franker, more transparent, more accountable approach to managing volcanic hazards. The use of robust, scenario-based planning, and the Okada model for transparency with the public should be major planks of this change of approach.
Of course, many other ideas can be put into the mix. The above are just a few starters for a frank, serious discussion on how we as a community can work together to build a sustainable future as we continue to live with an active, andesitic (thus, explosive) island-arc volcano.
So, let us now face the truth together, so that we can build soundly and sustainably on it. END