September 02, 2005

Katrina: some thoughts

I believe the Katrina disaster shows us that mass disasters basically scale in two dimensions: both the size of area affected, and the order of immediate property damage and economic dislocation, and the resulting effect that has on the social environment.

In the model I'm stipulating, first-order disasters, for the majority of those affected, are basically about the disruption of essential services: food, water, power. Examples would be the Canadian ice storm of 1999, or the major power outage of two years ago. These rarely seem to result, in Western countries, in extensive vandalism or the loss of social order.

Second-order disasters, as we're seeing in New Orleans, see the outright and rapid destruction of vast swathes of personal property and capital, with the resulting side effects of wage loss, personal mobility, etc: the main natural causes of these today are hurricanes, seismic/volcanic activity, or riverine flooding. If this kind of disaster affects the greater part of an urban area, it seems very difficult for even a fully Western society to keep order for long with local resources.

In the Globe today, Doug Saunders seemed to be leaning toward congratulating the ice storm victims I worked with in a military capacity in 1999 for not rioting in the streets back then. But the effects of that kind of first-order disaster are at a quantum level lower: while property and personal prospects in the affected areas had the potential to degrade over time if a person stayed in place, total sudden loss of everything one owned happened to very few if any of the victims, and the possibility of near-100% restoration at some point in future remained obviously high. As my experience showed, all that was needed to keep everything under control in the case of the ice storm was an effective governmental response focussed on rapid service restoration, plus (particularly in the rural areas) a dash of good old North American individualism. (Some of my colleagues reported trudging through snowdrifts to check in on backwoods house-owners, to see if they were getting through the frigid, power-free nights okay, only to be told, "what power loss?")

In a second-order disaster, that's insufficient. The experience of New Orleans leads one to conclude that the only real hope for a major Western urban centre affected by such a strike to keep any kind of order at all is to have prepositioned massive aid resources as close as was safe, and move immediately into the response mode as soon as the warning sirens stop. This clearly was not done in the case of New Orleans.

The difference between these kinds of disasters and the other kinds is so stark, and the likely areas where second-order disasters could occur so well-defined, one could imagine a map of North America with the likelihood of second-order effects displayed as a percentage gradient, with the Gulf region and earthquake-prone Pacific coast red, and much of the American Northeast an almost entirely safe yellow-white.

My city, for instance, Toronto, has an almost-zero chance of being affected by a second-order type disaster as I'm defining it. Short a nuclear bomb or an asteroid impact, the chance of a second-order strike to Toronto is zero (although we have suffered from occasional localized flash flooding of some city streams in the past, the effects were highly localized, and there has never been a significant earthquake or hurricane in this area). The same is not universally true across Canada... there is a non-zero risk of a tsunami in Halifax, significant earthquake/tsunami risks on the West Coast, and most of the prairie cities carry at least some small risk of major riverine flooding.

(Caveats: Biological/radiological disasters, like a flu epidemic or the Chernobyl disaster, follow somewhat different rules, but they're still closer to first-order disasters than second-order ones in this model... an epidemic left unchecked will gradually overwhelm the local health system and then other essential services, but by comparison to natural disasters, its effects on property salvageability and value tend to be slow-moving, and they are unlikely by themselves to impede flight or destroy most private and public property over a wide area in a matter of only a few hours or minutes. Other kinds of disaster can cause extreme destruction, but only over very localized areas, leaving most infrastructure and civilian response mechanisms in a large urban area unscathed, if overloaded... terrorist attacks like 9/11 or tornadoes in the midwest occupy this category. The one other kind of second-order mass disaster humans have experienced that I can think of, city-wide fires, has not been a risk in Western urban centres for decades. One other, extremely rare possibility for a second-order strike would be a major airborne chemical event, whether natural (Lake Nyos) or man-made (Bhopal).)

Obviously a lot of that success in prepositioning is going to hinge on the locations of military basing within a country, because that's the skeleton on which all mass disaster preparation inevitably rests. On that gradient map pictured above, you want major military bases equipped for relief support somewhere in the orange band... neither in an area likely to be directly affected, nor so far as to be of little use. A military base in Toronto is near useless if Vancouver slides into the sea... but so is one in Vancouver. The argument has been made that the nearest army base to Vancouver, at Wainwright, Alberta, is too far to be effective in this regard, and there's some point to that.

The ideal basing structure for military support in a civil crisis would seem to be large bases on the perimeter of earthquake/hurricane prone regions, and cities generally, and small installations like reservist armouries in the urban cores (to provide rapid staging and emergency shelter capability).

Posted by BruceR at 03:11 PM

McGill to take in Tulane students

From McGill University's press office:

"McGill University is joining other members of the American Association of Universities (AAU) in relief efforts to accommodate students from Tulane University in New Orleans who have been affected by the devastation of Hurricane Katrina."

Posted by BruceR at 02:05 PM

Canadian rescuers in Louisiana: update

The Canadian civilian SAR team in Kenner, Louisiana has been forced to cease operations due to rising disorder. Team leader Brian Inglis:

"It's far too dangerous for even the state troopers and police to wander out. It's absolutely crazy, the devastation is unreal -- the gunfire, the shooting, the looting is like something you see in a movie."

The B.C search-and-rescue team moved rapidly in response to a state-to-province request from the governor of Louisiana. Federal aid, including Canadian military support, is awaiting a similar request from the American government. If only for the irony alone, I'd love to blame shoe-shopping for the delay, but I can't: apparently all foreign offers of support are being personally handled by the Secretary of Defense. Mm hmm.

Anyway, Paul Wells is right: by any objective measure so far our government has moved faster on this to help its friends than after either 9/11 or the tsunami.

UPDATE, Friday: I guess they got that request: Operation Union was announced by the defence minister this afternoon, with 1,000 Canadian personnel apparently leaving for Louisiana by ship on Tuesday.

Posted by BruceR at 09:26 AM