Hurricane Rita and Gridlock on I-45 in Houston, TX (Wikimedia)

Last weekend Hurricane Irene caused an unprecedented shutdown of Metro New York’s transit (including all air travel, MTA, NJT, LIRR, Metro-North, and Amtrak). Given the assurance of wind damage to catenary wire, the probability of toppled trains on exposed elevated alignments, and the assurance of flooding – it takes all-day pumps just to keep New York’s subways from flooding every day – a shutdown was the only prudent course of action. New York was fortunate both that Irene was weaker than expected and that a large-scale evacuation was mostly unnecessary. Though there were very vulnerable spaces, such as Lower Manhattan and the Rockaways, relative safety could be had simply by traveling to the wild uplands of, say, midtown Manhattan. The New York Times deserves much credit for featuring transit news and advice prominently throughout the storm, but most other communities spoke only to drivers’ needs.

Much more after the jump…

Where New York was fortunate, the New Orleans of Hurricane Katrina was infamously unfortunate. Its 100,000 transit-dependent people (27% of New Orleans households!) were ordered to evacuate of their own accord, with no provision or direction on how to do so. Famous aerial shots depicted hundreds of flooded and ruined buses, while maxed-out contraflow highways made ‘evacuation’ on I-10 a sub-5 mph crawl. Those stuck in New Orleans in the first few days after Katrina endured conditions rivaled only by the poorest nations on earth, and they starkly revealed the fundamentally unjust result of a society built around personal cars. When I lived in Britain, more than once my colleagues marveled that a society could order an evacuation with such hand-waving as, “Y’all better get out!”

Ruined Schoolbuses in New Orleans – Wikimedia

I’d like this post and the comments to be a discussion of how transit can best contribute to disaster planning, both here and elsewhere. Though Katrina showed us the worst of both individualism (traffic, looting, etc..) and government ineptitude (FEMA, etc…), it is clear to me that societies owe each other the means to collectively evacuate themselves irrespective of economic class, and that those choosing to be carless should be provided a means of evacuation.

There are significant barriers to transit-based evacuations. Trains, though they have the highest capacity, are generally more vulnerable than highway infrastructure, as their rights-of-way were usually carved out along lowlands and waterways in 19th century. Bus evacuations are much more likely but have their own problems, such as storage of personal effects (people generally evacuate with lots of stuff) and the role of paid labor vs. volunteer driving and all the attendant liability issues there.
But perhaps the most difficult problems are trust and chaos. In disasters survivalism instinctively appears. Social contracts can quickly dissolve, leading to self-defeating self interest, textbook examples of how game theory explains disaster sociology. The urge toward self-reliance is very alluring; there is no other time I would want to be less beholden to government than when its failure to deliver could mean my death.

Hurricanes give lots of warning, a luxury our quake and volcano-prone region wouldn’t have. A major Rainier eruption could bury our rails and make the valleys impassable.  A major quake on the Seattle Fault would damage or destroy up to 80 of our bridges and up to 1,000 of our old masonry buildings. In a major disaster, then, our region should plan less for evacuation and more towards preemptive preparedness and adaptation. But the structural question – our collective responsibility in disasters – is one we need to rehearse over and over until we know exactly what we owe each other.

(For further reading, check out the 2008 report by the Transportation Research Board (TRB), or for those of you who don’t like reading 300+ pages, here’s a simple but well-written article by Michael Schwartz, an academic who has made disaster transit a central research theme of his.)

35 Replies to “The Role of Transit in Natural Disasters”

    1. Everybody has a disaster plan. Usually assembled after some hand-wringing exercise and then it is promptly shelved, rarely updated, and never tested.

    2. I helped in the early selection of one of their control centers a few years ago (though my input was limited to HVAC systems, so I was more of an observer than an active participant). From what I saw they really thought through the details. I assume they run “fire drills” at least every year or two to make sure things still work.

  1. By the photo above, the mode of choice for most people trying to get out of Dodge is still the much aligned private automobile. Katrina and Irene has proved how vulnerable our airports, rails, and buses are for accommodating the need, even though on paper they have the greatest capacity to fulfill the role. That seems like a major disconnect.
    Locally, we are subject to earthquakes – with bridges to cross the ship canal along it’s course in few places. I tried to find the seismic venerability rating of both I-5 and SR99 bridges one time, and gave up.
    This seems to be a pretty important thing to know, and I’m sure some engineers in Olympia or Bellevue have the answer (magnitude, acceleration, etc), but where does the public go to find out how precarious our two threads of mobility are?

  2. I found this article about disasters (published after the Touhoku earthquake and tsunami) interesting:

    Remember the gangs “marauding” through New Orleans, raping and even cannibalizing people in the Super-Dome after Hurricane Katrina? It turns out they didn’t exist. Years of journalistic investigations showed them to be racist fantasies. They didn’t happen. Yes, there was some “looting” — which consisted of starving people breaking into closed and abandoned shops for food.

    It depends on the nature of the disaster of course, but I would not be surprised to see huge tent cities in major parks all over Seattle, as happened after the 1906 San Fransisco earthquake. Transit can certainly help with that assuming the roads are passable and bases functional, which actually seems like a similar problem to a big snow.

  3. Ordering evacuations sooner, rather than later, would allow more complete utilization of transit and road resources. You’d still have to have a plan in place before ordering the evacuation (buses go here, trains make these trips, etc), but, in theory, you’d be able to utilize your transportation networks before any damage is done.

    That said, I realize that the political reality of ordering evacuations makes this a bit of a pipe dream. It depends on the science of forecast models, which will never be perfect (Katrina was a tough one, for example). And, for some reason, it seems to be political suicide to order an evacuation for a storm that never actually does much damage. Bizarrely the public seems more forgiving in the opposite case (I’m picturing canceling schools in advance of snow storms that turn out not so bad around here).

  4. Here’s a detailed description of how a mega-earthquake might affect our region. And here’s one about the lahars that would result from Mt. Raineer erupting. Those are our two largest possible events, short of nuclear war or a terrorist attack.

    My general recommendations:
    1. Very local, not high wind (floods, tsunami, or anything where you have to get to “high ground” or from one place to another up to a few miles away): Gondola. You can build the towers to withstand anything, and it’s high above the ground. Generally there’s a diesel motor for use when electricity goes out.

    2. Very local, high winds: This one’s more tricky. Some sort of tunnel to protect you from wind? Could be as simple as a pedestrian tunnel, and best as a subway with well planned drainage. Useless if water gets too high, but I suppose at that point it’s too late to evacuate anyway.

    3. Long distance: Elevated. Monorail, train, cars, whatever. Get people above water (overbuild columns and supports near possible flooding) and move a large number quickly.

    4. Nuclear/terrorist. Probably roads. It’s hard to really plan for this – setting up some seperate system doesn’t make sense. Cars and buses.

    As far as our local area, we don’t need people to go very far. Most of our emergencies involve getting to high ground quickly.

    1. So if the ground shakes, the waterline starts receding and/or Rainier starts erupting- GET TO HIGHER GROUND- there could be a seiche, a lahar or even magma headed your way.

      But is public transit really equipped/trained to get people to high ground? Is it better to have people just walk? Can a bus driver simply leave his route and drive up into Cap Hill or towards High Point if there is an Earthquake? Now that he’s off the route, aren’t the passengers going to have something to say as they are driven away from their original destination?

      If we had an earthquake and the bridges/overpasses to high ground were damaged but standing, wouldn’t it be better to have people walk across versus riding in a heavy bus over it?

      In a serious disaster like that we should be very cautious about everything and I’m not sure any advice other than “Get to high ground if you can do so safely” makes sense; because we can’t know the conditions the survivors will face.

      It’s a planning nightmare, to be sure.

  5. Having independently guided vehicles to take optimal evac routes rather than single points of failure that can be incapacitated with a single felled tree…seems like a better plan.

    1. As someone who grew up on the Gulf Coast and has witnessed more than a few evacs:

      LOL! Did you miss the photo at the top? Even when they make 65 North only it’s still gridlock.

      1. Well yeah but what about all the people ahead of those who were already getting out.

        And how many buses/trains would you have needed to evac millions?

        And are you going to send them back into the hurricane to get more people?

        And where do you drop the people off?

      2. So your advice is “be the first ones out”? Ok, we’ll just make sure everyone does that.

        Seems like the sensible option is to use your entire fleet to make several one way convoys out of town. Pick the nearest city outside the evacuation zone, and use their largest transit hub as your evacuation destination. It seems like you could send buses out to run regular routes, having them continue to the evacuation point at the end of the route. You’d want a publicly announced, coordinated time that the “last buses” will be leaving the start of their routes, and then send out your fleet in waves up to that last time. Preferably you could group them into convoys, once loaded, and give the convoys a supervisor or police escort. For longer distance evacs, they’ll need fuel on the road, so a supervisor or someone with purchase power to buy fuel for the fleet is necessary in every convoy.

        There may not be vehicle capacity for the full evac, depending on the size of the transit system. That’s simply a limitation of having a small transit system. I wouldn’t try to send buses back into the evac zone to pick up people left behind, as they’d probably end up stuck in gridlock, and never make it back out. An awful lot of people weathered through Rita stuck in their car on I-45 – imagine how much worse that situation would be on a crowded bus even with a police escort (although the bus is less likely to get blown or washed away).

        Odds are several of the buses will break down on the road and need mechanical work – I think Metro’s average for DE60LF’s is something like 1 breakdown per 2500 miles, and that’s considered good for a bus in city transit service. That means in an 100 mile evac, 1 out of every 25 buses will likely need on-the-road work. So you can either have mobile maintenance crews following, or simply mix an appropriate number of empty buses in the convoys and abandon the ones that don’t make it.

    2. It’s comments like this that make me very sure we need a good mass transit system. People who believe we don’t rely on *magic* in transportation planning.

  6. The problem with disasters is that no two are the same. An earthquake could paralyze the region as a whole making both road and rail impassable very quickly. Besides that, where do you take all these people? To yakima? The one benefit of private cars is that everyone heads to grandpa and grandmas far away, but with public transportation you need somewhere for these people to be able to go until you can either get them home or sent on further. One of the sad things in this country is that we have dismantled so much of our public transportation infrastructure that we would be hard pressed to even make such a scheme work. Transit agencies, greyhound and amtrak dont have large caches of equipment and operators in reserve and no one can talk to each other on the radio either if you could muster and enough operators and had a valid plan to run everything on the road as well.

    1. So your proposal is that we evacuate everyone to the freeways, at the same time? What are we fleeing from? Will the freeways even be passable?

      Luckily the Red Cross has evacuation centers identified all over. We can distribute people if need be.

      1. One of the bigger problems is how do you keep people from doing just that anyway. We cant close the freeways off, theres no gates or other intellegent emergency provisions built into the design of the interchanges, which would come in handy at other times (accidents, maintenace, etc) which could be activated at the flip of a switch.

        One of the things that can be done if the region is serious about this, is plan out ahead of time evaucation routes, boarding locations (schools, community centers), places that need special attention (Nursing homes/retirement communities) and have the routes planned ahead of time for a general evacuation (i.e. within the region, x miles away, another part of the state). As part of that you can modifiy the facilitys that will serve as your boarding locations to handle a large number of buses with easy ingress/egreess to the facility plus staging. Also with suitable amenties for the passengers (generator ran power and lighting, covered walkway so people can stand under some form of cover, etc) This can have other uses such as regular parking lot or whatnot during non emergency situations, but it would be ready incase it’s needed. Also as part of this planning process, make contracts and sign up people who are willing to work in such a situation, have contracts in place with the local transit, charter, and schoolbus operators to provide the equipment to the willing drivers, and have the equipment equipped for use in the disaster. Communications is a big deal, but it could be easily handled through cell phones that have been given priorty on the network, and CB radios for use at the staging areas. Both are redily available and cheap to obtain. No fancy solutions required.

  7. Different disasters require different responses.

    Are we trying to get everyone out of Seattle as fast as possible?

    Are we trying to get everyone home after a power outage?

    Are we trying to get everyone home after an earthquake has blocked roads with fallen brick?

    My first reaction to the “independently guided vehicles” comment is that you should know how to walk home. Tokyo produces bilingual maps for example to guide people home. Here’s a start,

  8. I would imagine the best bet for weathering out the kind of disaster that would hit Seattle would be to be prepared to stay in place.

    Water, Food, Gasoline, CASH (as power and communications usually come on later than shops open), generator, etc.

    1. Everyone who has a garage or shed to store it in should have one of those little cheap 2-stroke 800-1000w generators with a full tank of (stabilized) gas. They’re only around $150 at outlets like Harbor Freight, lightweight, and still put out enough power to run a refrigerator, charge a laptop/cell phone, run a computer, power an electric chainsaw, or even operate a small microwave oven.

      If your emergency plan involves staying put and waiting it out, you’re gonna need some power for refrigeration and communication (radio, cell, whatever). Not everyone can or should get a full 5000w generator to power their whole house, but with how inexpensive small equipment has become, there’s not much excuse to not have one of these and a couple extension cords.

      Honestly, the wait-it-out strategy is probably the most feasable. My other plan – trying to evacuate to Cle Elum – doesn’t really seem like it would work if everyone else in the region is trying to get out, too.

      Water supply, though, is a big issue. I suppose I could just go out to one of Seattle’s few remaining open reservoirs with a bucket if water supply was interrupted.

      Cell phones are not as disaster-resilient as CB or HAM radios, but still invaluable. Local cell towers may be down, but when not constrained by geographical obstacles (buildings/hills/trees), you can make a usable GSM cell connection to a tower up to 75 miles away (I don’t know about CDMA phones, but I assume they have a similar range). Get up on top of a nice tall hill, and you’ll usually get a signal. It might be one from Bremerton or Fall City, and the tower’ll probably be so overloaded it’ll take a half hour to place a call, but you’ll eventually be able to do it.

      1. Well for water you have a couple of options. First, go to a home improvement store and pick up some of those blue kerosene gas cans and use a paint marker to clearly mark it for water. Also rain barrels are a good idea. You can boil the water to make it potable, or just add iodine. Iodine water may taste like #*&$ and not be… refreshing, but it will keep you alive.

        For a generator, I think a good option is the Honeywell 2000i (like the Honda 2000i but not as quiet). You can get a factory refurbished one for ~$300 online. It has an econ mode so you only put out as much power as you use and the inverter means your electronics are safe to plug in.

        As for Comms, a few years back the FCC introduced a new amateur radio class, Technician Class, that is pretty easy to get (35 question multiple choice test). Get little Yaesu VX-7R and with your cell phone you’ve got some pretty serious Comms options.

      2. This all seems like overkill for almost any emergency. Keep enough canned food to last a few days, and enough fresh food to last a few days – that’s a week worth of food, which most people have on hand. Electricity is a luxury, not a necessity, and I don’t really see the investment in a generator as worth it. What could be useful is water. Your toilet cistern will last a family a few days, so you’ll need more than that. A few 2-gallon jugs should do it.

        Actually, the odds of water supplies failing across the city are quite low. An earthquake can break some lines, so a neighborhood or two might lose water, but then you walk to another neighborhood for a refill or to your local store that’ll probably be stocking water like crazy 2 or 3 days in.

  9. Duh. The last resort is of course feet – assuming you’re reasonably healthy you can get away from most things by walking. What’s the best option before feet? Bikes!

    Alan at The Daily Score linked to this story about DC’s Capital Bikeshare tripling in rides right after the earthquake. What would that freeway above look like if everyone was biking away? It would be moving a whole lot faster.

    1. A friend of mine biked across Tokyo to rescue his wife when the trains there shut down after the big earthquake. He said that other than being exhausting (particularly on the return trip with her riding side-saddle on the cargo rack) it worked pretty well.

      1. Awesome. Disaster safety device: bicycle. Should be in every home. We can also use this argument to get a bike sharing system in Seattle (once we repeal our helmet law). Best way to high ground when the roads are gridlocked and a tsunami or flood is on it’s way.

  10. “and that those choosing to be carless should be provided a means of evacuation.”

    I would prefer “those that must be carless” as some might complain that we’re subsidizing their choices, whereas those that can’t afford a car still deserve to have their lives protected (which would still help the carless-by-choice).

    That came off as more heartless than intended…

  11. I have a problem with calling the disruption “unprecedented”. I seem to remember 9/11 rewrote the book on unprecedented transportation cancellations and delays.

  12. What the best response for disaster planning is depends on what the nature of the disaster is and what the hazards a particular area faces in a particular scenario. The amount of advance warning is also a factor as is the likelihood of an event.

    As the PNW doesn’t get hurricanes and doesn’t have millions of people in low-lying areas along the coast we don’t need to do large-scale evacuations like you see on the Gulf and Atlantic coasts.

    Floods, tsunamis, and volcanoes require evacuations, but only of limited areas. With floods and volcanoes you tend to have a fair bit of warning before the situation becomes critical as well. Though to be fair a steam vent under a glacier can create a lahar with little warning.

  13. No one has mentioned what singular transportation mode shone forth in Seattle’s last transportation “emergency”. That being rail.

    1. … rail with its own ROW, grade-separated, and sturdy enough walls to keep cars from flying out of the sky (or off a steep hill) and landing on the tracks.

  14. One part of the response to the blizzard that could have easily gone more smoothely would have been to close the I-5 express lanes to private vehicles, so that the hundreds of buses carrying tens of thousands of passengers could get through first. If people know buses will get priority on the express lanes in a blizzard or other natural disaster, they’ll be more likely to take the bus and leave their car in their worksite’s garage.

    All the commuter buses could have gotten to their destination by 9 pm, allowing cars back onto the express lane so that everyone gets home, rather than having tens of thousands of cars and buses still stuck in a freeway parking lot, in the middle of a blizzard, until the next morning.

    From my perusal so far, there are no emergency shelters for non-homeless people who simply can’t get home.

    1. Love the bus-only idea. Tough to implement though, since nobody would know when it’s in effect. Maybe use those electronic information signs.

Comments are closed.