Wednesday, November 10, 2010

Is Seattle Ready for Snow?

During the past few days there has been a great deal of media attention regarding Seattle's updated plans for dealing with lowland snow in the city. At a recent press conference, a comprehensive city response plan was discussed and the Mayor made it clear that the city had taken wide-ranging steps to be prepared.


This whole discussion is very timely considering the latest weather forecasts for later next week--the atmosphere is moving towards a La Nina-like pattern with ridging in the eastern Pacific and cold air plummeting southward. (Don't get too excited yet---most of the time cold and snow forecasts more than a week out don't materialize for reasons I can discuss in another blog). But I am sure Jim Forman of King-5 is waterproofing his famous ski parka.

Last month I was invited to talk to the Mayor's staff and some of the department heads (as well as Metro staff) about the potential for snow this winter. Let me tell you, these folks are serious about dealing with snow and preventing a repeat of the disaster of December 2008. And their newly released snowstorm plans are impressive in many ways and many of the mistakes of 2008 will not be repeated.

First, they will aggressively pre-treat roadways BEFORE the snow hits so it will not bond to the pavement. They will use salt on roadways after snow falls to encourage melting. They will plow the snow to the sides instead of into the center of streets (snow in the middle of the streets creates major barriers preventing turns, the snow melts and then refreezes as ice in the traffic lanes, and many other bad things). No more packing the snow down and sprinkling some sand on it. They are going to coordinate with METRO to get the buses off streets on which they will only get stuck. They are going to triage the roadways, as noted in their snow map below. They are making it clear that only main routes will get treated and plowed. This makes sense--with limited resources you have to use what you have to open the city rapidly to buses and commercial traffic. Then, with main roads clear and the main bus routes running you take on the side streets.


Clearly, Seattle must triage since they have only 30 snowplows, which is not nearly enough to clear a city this size--so SDOT has to work smart. (FYI--Denver has 68 plows but they get nearly 60 inches of snow a year, roughly 6 times more than we do, but we have more of an ice problem).

So we are in a much better place than two years ago and Mayor McGinn and City of Seattle staff are highly motivated and working hard to be prepared.

But let me provide some suggestions considering the meteorology of Seattle snow events.

Seattle snow generally falls on relatively warm ground that initially melts the snow into slush before it refreezes. That is why we frequently have an ice problem after snow events. You have to plow it off rapidly before it freezes, and if you can't do that you need to pretreat to stop it from bonding to the concrete so when it does freezes you can scrape it off. And you need to get salt on top of any frozen roadways to encourage melting.


So rapid response is key. And so is getting the vulnerable buses off the roads...particularly the trolley buses and the double buses, both of which are disasters in snow.

Now the city has 30 plows. Lets say that 25 are available at any one time (which would be very good I would think). And lets assume that they plow at 20 mph. So they could cover at best roughly 500 miles of roadway an hour. How many miles of roadway lanes are included in the routes outlined in the map...my rough estimate is 1200 lane miles. So it would take 2.5 hours to cover the city--and that is optimistic. If the snowfall is relatively light (.5-1 inch an hour) this might be ok, but if the snow is heavy (several inches per hour) there is going to be trouble. My suggestion: have a smaller core set of major streets that get priority in heavy snow. Get those clear, keep them clear, and steer the buses to those streets.

If snow is forecast and some people absolutely need their cars, they might be encouraged to move them to major parking lots near major streets and bus routes...like those at Magnuson Park, Northgate Mall, the UW, etc.

Now a major worry is Metro bus service. As noted above, the trolley buses need to be pulled at the first sign of snow (they skid, move off the overhead lines, and are like dead whales) as should the articulated (double) buses that jackknife on snowy hills. Let SDOT clear the roads to bare pavement and then those buses can be used. Core, protected snow routes, need to be established (I think they are going to do this). And service needs to be increased above normal frequency on those routes. People might have to walk a bit, but they will be able to get home.

Now a sore point for me is Metro's bustracker software. This should be a critical element for snow situations, allowing people to see where the buses are (and aren't). In December 2008 the Metro bus tracking site completely failed under the load of potential bus riders. They have made some moderate upgrades (roughly three times the capacity), but those will probably not be enough, particularly now that many people have internet on mobile devices. Getting a properly size server would not be that expensive and should have been more of a priority.

The other problem is the ancient bus tracker technology Metro is using. Instead of using GPS, which is modern and now inexpensive, they are using an old fashioned radio system that is not only undependable, but which is guaranteed to give poor results when the buses go off their normal routes. Metro knew they had this deficiency two years ago and haven't dealt with it. These days, accurate positioning of buses is essential and is desperately needed in snow situations.

Finally, there is an important element of personal responsibility. Moving your car down the hill is one. Having a snow shovel and some salt/deicer is another. Some neighborhoods (like mine) are planning to have a large supply of deicer to take on the worse street sections in our area. Sort of like the snow version of the block watch program.

Anyway, a great American city was unnecessarily crippled during the December 2008 due to poor planning and operations. To be fair, that event was unusually prolonged...but such events do happen here every decade or so (e.g., Nov 1985). The city seems far better positioned today to deal with snow.

And parenthetically snow forecasting has never been better. But keep in mind that forecasting snow is the most difficult local prediction problem. Why? Because we have to get BOTH temperature and precipitation right? Because we have to get the AMOUNT of precipitation right. And because our area is often right on the edge between rain and snow. The Olympics of forecasting!


27 comments:

snapdragon said...

November 1985- I remember that well. Took 10 hours to drive from WWU in Bellingham to Portland. It's usually a 4.5 hour drive. From Kelso south, it was basically a sheet of ice. That's when I learned it was not MANDATORY to go home for Thanksgiving!

I'm hoping for snow in the Portland/Vancouver area this winter, and I plan to not drive in it at all!

Todd said...

I've got my generator, AWD car, and years of snow-driving skills. I'm set. Bring it on!

Upupaepops said...

It certainly sounds like they are better prepared. To me it sounds like they have thought out the plan. If I am not mistaken they did a dry run for the plows

But citizens need to prep too.

They need to know their emergency bus route BEFORE the issue strikes. Metro need to advise, and people need to listen and think.

They need to stage their car so they can potentially drive without getting up a beastly hill from their house. I lived on Magnolia for years and would park on Thorndyke and walk to my house. I never had to deal with the super steep hill. I easily got from their to Capitol Hill for work

wanderchow said...

Aw, no more "dead whales" dangling out over Interstate 5? No more sandy berms to scale in the centers of roads? What a BORE. Kidding...now let's hire about 30 more private snow-plowers to get out there and supplement.

signalius said...

Metro also needs to close bus stops which are on hills. The buses can make it up hills just fine-- as long as they don't stop to pick people up! A simple sign: Walk to the nearest stop on level ground --->

and now for a more radical idea: pump seawater out of Elliott Bay into tanker trucks and spray it on the roads. Free salt! And maybe a little plankton too.

windlover said...

I remember 1985! I almost ended up in the ditch on my way home from work with my 2 small children (ages 2 & 3) in the back seat. It was such a treacherous drive home I called in and said I was not coming to work the next morning...I wasn't willing to take the chance. It had barely begun to snow...they were going to fire me. Needless to say, later that afternoon, I received a call from my manager...He wanted to apologize for his reaction that morning. By the time he called we had over a foot of snow on the ground! By the time it was done snowing we had 2 feet!

d33ann said...

I can't tell from the map, but I'm hoping that First Hill gets special attention in the case of a heavy snow. I snow-shoed to work at one of the hospitals, but most of my coworkers were not so lucky. Due to the Code Delta (Disaster) status, they were required to spend several days working >12 hour shifts and to stay the night on cots in auditoriums and un-used back hallways in order to maintain safe staffing levels.

AtlantaWxGuy said...

Since these events are relatively rare, education of the public is key. First and foremost, a simple (color) scale needs to be used to define the different levels of serverity. For example:
Code Yellow - Caution. Prepare.
Code Orange - Take Action. Minor changes to city operation.
Code Red - Emergency. Drastic changes to city operation.
This will keep everyone on the same page and avoid comparing apples to oranges (ice vs snow).

drexnefex said...

bummer....there was some epic sledding opportunities in Seattle last winter.

Katherine said...

The articulated coaches manage the snow quite well if their center dual tires are chained as well. I have driven them in snowy conditions. Metro's biggest mistake two years ago was sending many buses out without chains even with snow already on the ground. Metro still neglects to train their operators how to drive in the snow as well as inform passengers how to ride in the snow (don't expect curb service, don't wait on a hill.) Metro has fixed some of the mistakes made previously, but they still have several more they need to rethink.

bLarryB said...

I grew up in New York City, where snowplows are simply affixed to the front of garbage trucks and turned loose to plow. I've always wondered why Seattle doesn't do something like that to supplement the number of available plows. Is it that hard to co-ordinate between SDOT and Seattle Public Utilities?

steven said...

It seems like we are way to concerned with our ability to drive places. When it snows in Seattle its usually an episode that lasts a day or two. Why so much concern about it being easy to drive on those days? I think we whine about this issue a bit excessively. Our lives are pretty easy if snow only makes us worry about driving conditions and not things like, staying warm, will the water run and do we have enough food for this winter? It really seems like there are more important issues for people to get all excited about.

Josh said...

I don't think there was any changing of the guard or lynching of the king after the 1985 storm. But these days when things aren't up to technological-connivence-speed ( like restless waiting for the iphone app to show us the way home) people tend to tire quickly and grab their pitch forks. Kind of like the recent election. Pandora's box has been opened and pretty soon we will want a snow plow map on our ipads to see when they will be by our place.

steven said...

Why do we make such a big deal out of snow? Snow in Seattle usually means a day or two of inconvenient commuting and months of complaining about how bad the traffic was. Is it really such a big deal or do we simply get a kick out of whining. If we had concerns like; will our water run? Will the population be able to stay warm? If stores close will we have enough food to last two days? What are the implications for the homeless and those in poor housing? Nope. Instead we are concerned about being able to drive places quickly and conveniently. Not that big of a deal.

C.P.O. said...

Always love the Forman references! December 1990 was another instance as well.

rainier beacher said...

One thing not mentioned in any of the city's plans that I've seen is how to deal with the myriad of abandoned cars that prevented the plows and buses from doing their work.

rainier beacher said...

The only fly in this ointment is the large number of abandoned cars in the middle of all those lanes.

MK said...

Re: bLarryB's comment - I wonder how much time was spent looking at other cities' snow plans. We are far from the only city with occasional but infrequent snowfall. This new plan sounds great, but it would be even better if we weren't trying to reinvent the wheel.

John Franklin said...

Number of days (annually) Seattle might lose to snow: typically well below 5

Number of days (annually) people in Seattle can spend talking about snow removal: well over 30

Entertainment benefits of living in a city where people spend more energy talking than acting: Priceless.

stacey said...

Some people work jobs that do not include PTO, and if you don't work, you don't get paid. I'm lucky to have a reasonable, understanding employer. Others are not so lucky.

Hair Bob said...

drexnefex -

Looking at the SDOT map, they have already surrendered the Queen Anne Counterbalance before the first flake ever falls.

Courier & Ives on crack - the hill is soon salted with the blood of the locals taking on the SUVs going around the "street closed" signs.

Glorious... ;^}

Gerhard said...

Bustracker is also a major sore point with me. I am tired of the unusable applications. I have been looking for the protocol documentation to fix this problem myself, but it seems that info is not available. Please encourage whoever owns bustracker to publish that information.

Soupman said...

Cliff,
This question is sort of off-topic but it does indirectly correlate with potential snow events here in the PNW. In the NWS discussions, there is talk of a ridge of high pressure retrograding to the west. With the predominant air flow from W to E in the Northern Latidudes, just curious how a High can move west? I would appreciate a short explanation if that is possible.

Wx Enthusiast said...

Scott Sistek from KOMO wrote a rather good explanation of why many times cold and snow forecasts in the models more than a week in advance do not pan out.

http://www.komonews.com/weather/blog/107298353.html

Basically, the GFS (specifically) beyond 192 hours (8 days) is run on coarser resolution, which includes much less accurate depiction of the mountains to our east (Cascades, even the Canadian Rockies) that tend to block a lot of the cold air from moving into western Washington. Thus, more than a week in advance, the GFS pretty much assumes that all that air can just move into our area with no problem. But the conditions have to be right for it to move in through the Fraser River Gap (and to the south, the Columbia River Gorge contributes as well). One of the runs of the GFS earlier this week brought low temperatures of -11 F to Bellingham and +3 F to Seattle on Thanksgiving morning. Yea, maybe, if those tall things called mountains were not in the way.

Tim said...

@snapdragon: I was on I-5 that day as well. Couldn't leave until Les Schwab got my snow tires on at noon after a four-hour wait, and then it took me eight hours to get to Kalama.

I did a 270-degree spinout at the Martin Road exit in Lacey, but no one else was behind us, and I remember asking my passenger if she was all right and her responding in the affirmative before we'd even stopped sliding.

My memory, though, is that things got easier after that, but then I didn't go as far south from Kelso as you did.

wayne said...

I’m the IT manager for the King County Department of Transportation, and would like to offer up some more info. Since the winter of 2008-2009, we have upgraded the Tracker servers (production and backup) and have done preliminary stress testing on the Tracker application to determine that we can handle at least 20 times the normal daily peak load on the site and about 10 times more peak load requests than occurred during the snow storms of 2008. We will do further stress testing after we obtain additional testing tools.

Prof. Mass, like many others, hope that transit riders can rely on our Tracker system even when service become highly disrupted due to snow route operation and significant traffic delays. Unfortunately, server hardware capacity is not the primary problem. The current legacy Automatic Vehicle Location (AVL) system cannot provide data to the Tracker application if a bus deviates from its normal route path. As soon as buses go to snow routes, it renders the Tracker application nearly useless. This also happens to private applications - like One Bus Away - that also use the same Metro data. If this happens, we will disable our Tracker application and post notices to that effect on our website. Quite a bit of Metro data is available at http://www.datakc.org/ and also at http://www.kingcounty.gov/transportation/kcdot/MetroTransit/Developers/AppCenter.aspx for anyone who would like to develop an application.

Metro has been working on a project to equip all buses with GPS for some time. GPS enables “off-route” buses to be tracked, although predictive arrival times at stops would still not be accurate (because they are based on historic travel times). We have already begun installing the new system and all buses in the fleet should have it by 2012. This will provide much better real-time vehicle location information to the public. Outfitting 1,400 coaches with GPS emitters is a difficult task in itself – but that is almost trivial compared to the data transmission infrastructure and back office systems that are needed to process that large amount of information to match to service routes, do predictive arrival times, and present the information to the public in a meaningful way. Consumer solutions using GPS-enabled devices don’t have the back office data management to handle 1,400 buses. A complete radio system is also required to transmit the location data cost effectively. The cost of a commercial cell carrier just to transmit the data would be about $1 million a year.

Others who have posted on this blog are correct - the best thing you can do to travel safely during this winter is prepare ahead of time. Sign up for Transit Alerts at www.kingcounty.gov/metro/alerts. Know the snow routing for the bus routes you use most often - all are included in the online and print timetables. And, check Metro's service status page at www.kingcounty.gov/metro/snow before you leave.

Wayne Watanabe
IT Service Delivery Manager
King County Dept or Transportation

wayne said...

I’m the IT manager for the King County Department of Transportation, and would like to offer up some more info. Since the winter of 2008-2009, we have upgraded the Tracker servers (production and backup) and have done preliminary stress testing on the Tracker application to determine that we can handle at least 20 times the normal daily peak load on the site and about 10 times more peak load requests than occurred during the snow storms of 2008. We will do further stress testing after we obtain additional testing tools.

Prof. Mass, like many others, hope that transit riders can rely on our Tracker system even when service become highly disrupted due to snow route operation and significant traffic delays. Unfortunately, server hardware capacity is not the primary problem. The current legacy Automatic Vehicle Location (AVL) system cannot provide data to the Tracker application if a bus deviates from its normal route path. As soon as buses go to snow routes, it renders the Tracker application nearly useless. This also happens to private applications - like One Bus Away - that also use the same Metro data. If this happens, we will disable our Tracker application and post notices to that effect on our website. Quite a bit of Metro data is available at http://www.datakc.org/ and also at Metro Online (click on apps icon) http://www.kingcounty.gov/transportation/kcdot/MetroTransit/Developers/AppCenter.aspx for anyone who would like to develop an application.

Metro has been working on a project to equip all buses with GPS for some time. GPS enables “off-route” buses to be tracked, although predictive arrival times at stops would still not be accurate (because they are based on historic travel times). We have already begun installing the new system and all buses in the fleet should have it by 2012. This will provide much better real-time vehicle location information to the public. Outfitting 1,400 coaches with GPS emitters is a difficult task in itself – but that is almost trivial compared to the data transmission infrastructure and back office systems that are needed to process that large amount of information to match to service routes, do predictive arrival times, and present the information to the public in a meaningful way. Consumer solutions using GPS-enabled devices don’t have the back office data management to handle 1,400 buses. A complete radio system is also required to transmit the location data cost effectively. The cost of a commercial cell carrier just to transmit the data would be about $1 million a year.

Others who have posted on this blog are correct - the best thing you can do to travel safely during this winter is prepare ahead of time. Sign up for Transit Alerts at www.kingcounty.gov/metro/alerts. Know the snow routing for the bus routes you use most often - all are included in the online and print timetables. And, check Metro's service status page at www.kingcounty.gov/metro/snow before you leave.

Wayne Watanabe
IT Service Delivery Manager
King County Dept of Transportation