North America

Gaga for Gondolas: The Surprising Relationship Between Semi-Autonomous Cars, Cable-Cars, and Good-Ole-Fashioned Railways

The really great thing about cable-cars is, of course, that they can fly.


Banff, Alberta

Obviously, this has caused cable-cars to be used mainly in mountainous areas: in ski towns like Banff or Whistler or in high-altitude cities like Medellin or La Paz. Where they are used outside of mountainous areas, as in NYC, it is usually to cross over rivers or sea channels.

The limitations that cable-car gondolas have faced thus far are for the most part the same as those of their riverboat counterparts: they travel relatively slowly, they can’t be used in most areas of any given city, and they can’t usually be used to transport cars. These three limitations have caused urban gondolas (or Trams) to be relatively few in number, and to be considered useful mainly for recreational rather than functional purposes.

Gondolas .png

“Gondolas” in Venice and Rio de Janeiro

Intermodal Transportation

This relative scarcity of urban gondolas may be coming to an end. With new transit technologies emerging — ride-sharing, car-sharing, bicycle-sharing, apps that find parking spots, cars that can parallel park themselves, semi-autonomous cars that can go find parking spots, even wholly self-driving cars — cities could be entering a new era of “intermodal” transportation. In this era, commuters may generally have the option of transferring seamlessly from one vehicle, or one mode of transportation, to another.

With people no longer tied to using a single vehicle or mode of transportation, cable-cars might end up with a significant role to play within city transportation networks, ferrying passengers over topographic barriers like superhighways, valleys, rivers, or hills.

cable car .jpg


Vertical Parking Towers and (Semi-)Autonomous Cars 

Autonomous cars might work well with gondolas; a car could drop off a passenger at a gondola entrance, and another car pick up the passenger at the exit. However, even if fully autonomous cars do not become commonplace for many years yet, gondolas might still work extremely well in tandem with semi-autonomous cars:

Either for political or technological reasons, it might be that autonomous cars first become tolerated only in certain places (e.g. designated areas within parking lots), certain times (e.g. from 4am-5am, the “witching hour”), or at certain speeds. Even just allowing autonomous vehicles from 4am-5am, at no faster than a measly 10 km per hour, would benefit car-sharing services enormously, allowing them to deliver cars to customers for the following morning or have electric cars drive themselves to battery-charging stations. This, along with autonomous parking lots, might spur intermodal transit even more than fully-autonomous cars would.

The biggest impact of semi-autonomous cars, both for intermodal transit in general as well as for gondolas, might come from vertical parking lots. Today most parking lots are no more than a few stories tall at most, because drivers don’t want to wind their way up and down many stories to get in and out of the lot. (Using car elevators instead of ramps might be too slow unless the elevators were very large, in which case they would take up too much room). If, however, humans could get out of their cars and then have them park themselves, then winding up or down many stories in a vertical lot would no longer be a problem.

Gondola Car Tower Pic

Okay, I’m not exactly a graphic designer, but you get the idea…

Vertical lots would also be able to hold more cars in them without humans, as they could have shorter ceilings per story, ramps that take up less space by winding more tightly, less space between each car on the ramps, and narrower parking spaces.  In some cases, if the vertical lots were to hold car-sharing cars rather than personally-owned cars, they could be able to hold even more cars, and distribute them to passengers more quickly; they would function more like large take-a-penny/leave-a-penny jars for cars rather than traditional parking lots.

As a result, vertical parking lots could also serve as ideal urban gondola stations.

Urban Archipelagos 


Highway cloverleafs might be one good place to put car-sharing parking lots, ride-sharing transfer areas*, and/or gondola stations. These cloverleaf loops are often huge and unused;   the intersection of the 401 superhighway and Don Valley Parkway in Toronto, for example, has enough unused, highway-surrounded land to contain within it two entire Skydomes. Gondolas could in many cases be useful in linking these otherwise unusable lands to one another, while also helping to alleviate traffic jams on the highway itself and — especially in cases where highways run along the floors of valleys — connecting the highway to other nearby transit.

*A “ride-sharing transfer area” is a place where people could disembark their ride-sharing or car-sharing vehicles in order to switch to either a different vehicle or public transit. For example, a commuter from the suburbs could share a minibus to the city with people from his or her neighbourhood, then switch to a different carpool in order to get from the city to his or her office downtown. A transfer area is, in effect, a marketplace for transportation.  As with other markets, they require “liquidity” to be effective, which is why putting them near major highway intersections and connecting them to public transit and parking lots is useful.

Highways and Railways

Putting parking lots and ride-sharing transfer areas alongside highways makes sense not only because highways are where automobiles tend to congregate most heavily, but also because you ideally would not want to put too many parking lots or transfer areas in other places in the city, where you could be building better things like housing or green spaces instead.

Cars, however, even if they do become autonomous or electric, are simply not as efficient as railways are in most cases, or even as efficient as Bus Rapid Transit (especially if the busses become autonomous) is*. The cheapest way of building new railways or Bus Rapid Transit lines in a city is by doing so on a highway. The problem, however — the reason this has not happened too often thus far — is lack of accessibility. You cannot easily use the outer lanes of a highway for a rail or rapid-bus line, as it would block the highway’s exit/entrance lanes, thus requiring either expensive construction or traffic lights to overcome.

Moreover, the rail/bus line would still only be accessible to those on the side of the highway you choose. The people on the other side of the highway would have to cross a bridge or tunnel — a long bridge or tunnel, given that highways can be very wide (the 401 is usually about 100 metres wide). And all this is in addition to the fact that building residential condos next to highways is not ideal, so people must first get from their home to the highway; something that car-sharing parking lots and (semi-)autonomous cars and parking spot apps and ride-sharing may make far easier, but which has not been too easy up until now.

train in highway.jpg

It would be better, as in the graphic above, to put the railway/busway in the middle of the highway, where it would not block the highway’s entrance/exit lanes. This, of course, only makes the accessibility problem more severe. A gondola could help address this problem, linking parking lots and ride-sharing areas alongside the highway (and/or in highway cloverleafs) to the rail/bus stations in the middle of the highway. This would, of course, be especially useful for highways that run along the floors of valleys, like Toronto’s Don Valley Expressway.

*(While Bus Rapid Transit is not generally as efficient as railways are, it can in certain cases have advantages over rail, especially in the shorter-term. Not only does Bus Rapid Transit require far less construction, but it also, unlike railways, offers the possibility of building a one-lane system: a Rapid Bus lane could be used by busses going downtown in the morning, then switch directions and travel uptown after noon in order to keep with the main flow of traffic).

 Where to Put Gondolas in Toronto

1. Don Valley-Danforth 

DVP and the Danforth .png

A 1.2 km line from Parliament to Broadview, with stations at St James Town, Castle Frank, Bayview, and Broadview. The St James Town and Broadview stations would include take-a-car/leave-a-car vertical parking lots; Bayview’s would include both a ride-sharing transfer area and a new “Relief” train station on the Richmond Hill GO line.

richmond hill go line.png

Richmond Hill GO line

St James Town development

This is a digital rendering of the enormous towers that condo developer Tridel is planning to build in North St James Town, by the intersection of Parliament and Bloor. St James Town is  already perhaps Canada’s most densely populated neighbourhood

Broadview Subway Station

Castle Frank Subway Station

2. Allen Expressway-Highway 401 

401-Allen 1.75 km Gondola .png

A 1.75 km line, linking two shopping facilities and parking lots (at Yorkdale, just south of the 401, and next to Costco, just north of the 401 ), connecting to two subway stations (though you might have to walk through Yorkdale or through Yorkdale’s parking lot to get to Yorkdale’s subway), and next to two separated areas of Downsview “Park”. It could be built in two phases: its southern half from Yorkdale to Wilson first, and its northern half in the longer-term, if and when the Park is finally built. The parking lots by Yorkdale and/or Costco could be an ideal place to build a gondola station that is also a vertical car-sharing parking lot.

…to paraphrase Trump, let’s build a gondola and get Nordstrom to pay for it.

3. DVP-Thorncliffe Park

thornc. park.png

A 1.1 km line, descending in and out of the Don Valley at Don Mills Rd., which is at an elevation 25-30 metres below where the stations at Thorncliffe Park and at Coxwell Ave would be.

Thorncliffe Park .png

Thorncliffe Park, as seen from the DVP (just west of the DVP’s intersection with Don Mills Rd.)

The intersection of Coxwell and O’Connor, or of Don Mills and O’Connor, might, maybe, be the site of a Downtown Relief Line subway station eventually

4. Don Valley Parkway-Eglinton

eg dvp.png

A 1.3 km line, similar to the Thorncliffe Park line in that it would descend into and ascend out of the Don Valley. In this case, however, its valley-floor station would be at a rail line (the Richmond Hill GO Line) and ravine trail, rather than at the DVP. It would, instead, meet the DVP where the DVP intersects with Eglinton, next to the Aga Khan Park & Museum stop on the Eglinton Crosstown.


East of Laird, with the brief exception of Science Centre Station at Leslie, the Eglinton Crosstown LRT stops being a subway and travels overland

5. 401-DVP-Sheppard!

DVP-401 Intersection.png

The 401-DVP intersection is where Toronto’s main north-south and east-west expressways meet. It also happens to be less than a kilometre (as the gondola flies) away from Don Mills Station, the easternmost station on the Sheppard Subway line.

It is also enormous:

Downtown put Uptown .png

Part of downtown Toronto, cut-and-pasted onto the 401-DVP intersection

There are a few ways that you might be able to build useful gondola/vertical car-sharing lots/ride-sharing transfer areas at this intersection. The simplest is to build a 2.2 km gondola linking the 401-DVP intersection to Don Mills subway station, stopping along the way at the highway cloverleaf at the intersection of Sheppard and the DVP (aka Highway 404):

401-DVP Gondola.png

(Just west of Don Mills Station, a new 35-floor residential high-rise put in a development application at the end of 2016)

An alternative to this could be having the gondola run only 1.4 km, stopping at the DVP-Sheppard intersection. In this plan, the Sheppard Subway would be extended 650 metres eastward, to a new subway station/gondola station/vertical parking at the DVP-Sheppard intersection.

A different, more intrusive (but shorter and cheaper) alternative would be a 1.6 km gondola line, taking a more direct route to Don Mills station:

Don Mills line.png

Given its intrusiveness, this alternative plan would not be worthwhile unless it were to be  accompanied by a an ambitious project to turn this area between Don Mills Station and the  401-DVP intersection into a gigantic new transit/transportation hub for the northern part of Toronto.

This new hub would have to include some or all the following additional features:

— A 1 km line linking Leslie Station (on the Sheppard Subway line) to North York General Hospital, the 401 Highway, and Oriole Station on the Richmond Hill GO Line:

Hospital Line

The 401-Leslie intersection is 1.5 km west of the 401-DVP intersection. Leslie Station is the station immediately west of Don Mills Station

— A transit line reaching eastward from the new hub. This could be either a Sheppard East LRT, or a Rapid Bus route along either Sheppard or the 401, or a rail line on the 401 east of the DVP, and/or a gondola linking Leslie in the west to Warden in the east:

DVP-401 Gondola

7.4 km gondola network at the site of a major transportation marketplace

DVP- 401 Aerial view.png

The gondola’s eastern terminus would be Warden or just west of Warden, where the north-south Warden Hydro Corridor and the northwest-southeast diagonal hydro corridors meet (both of which, in turn, link to the wider Gatineau Hdyro Corridor bicycle trail, which will itself link up with the Eglinton Crosstown). The Warden Hydro Corridor could similarly be future site of a green-space bike route

If we can even way, way more ridiculous here for a minute, I would like to propose not only turning the 401-DVP-Sheppard area into a transportation hub, but also into a housing and sports hub. It would feature mid-rise residential developments in the neighbourhoods surrounding the intersection, and would feature the world’s first transit station-sports stadium hybrid.

The stadium — Rob Ford Place — would be home to a new NFL team called the Toronto Housing, as well as a new NHL team (with affordable ticket prices!) called the Toronto Transit. The Transit would only play half their home games in the football stadium – the other half could be handled by the ACC — which would mean the stadium would only be used for sports about 30-40 days a year. The rest of the time, it would convert into a large transit station.

It’s the Mary Poppins approach: hockey and football is the sugar, transit and housing the medicine.


…told you I can’t do graphic design

It would also be great for tourism: utterly unique, weirdly highly efficient, accessible from the airport, kicked-off with a Superbowl featuring a trio of Drake, the Weekend, and Ryan Gosling at the halftime show, and leveraging Canada’s primary brand: affordably priced ice hockey.

For more about the 401-DVP intersection, see A Bazaar Alternative to the Scarborough Subway 

6. Exhibition Station and the Gardiner Expressway 

Exhibition Station is, today, one of just two GO Train stations in the City of Toronto south of Bloor/Danforth, outside of Etobicoke. The Exhibition has a vast parking lot, the only lot of any significant size next to one of the downtown Gardiner Expressway’s on/off-ramps. A new gondola could be built at Exhibition in something like one (or both) of the following two ways:

Exhibition Stn. gondola.png

A 2.3 km from the marina at Ontario Place Island,  to Ontario Place’s parking lots (and the Martin Goodman lakeshore bike trail) next to Lakeshore Avenue, to BMO Field and Exhibition Station, then east over the GO rail line to Garrison Crossing next to Fort York and, finally, to Front Street’s (and the proposed Rail Deck Park’s) western terminus at Bathurst Street.


1.75 km Exhibition to Queen line .png

A 1.75 km line to the southern foot of Ossington, on Queen Street, passing over both GO rail lines along the way (and perhaps stopping not only at Exhibition Station, but also at the new planned Liberty Village GO and SmartTrack  station). This could be preferable if there is a Queen Street subway built eventually. But, it would have to cross partly over a neighbourhood, which residents of the area might not want.

7. 401-400-409 


A 5.1 km line, linking highways to the Etobicoke North GO Train station (on the Kitchener-Waterloo GO line).

Or, much shorter alternatives might be possible instead, such as just linking the 409 cloverleafs to Etobicoke North station.


8. Mt. Dennis-Black Creek-Eglinton 


Mt Dennis.png

A 1.3 km line, linking to the Mt. Dennis station on the Eglinton Crosstown, which is also expected to be the site of a new GO train/SmartTrack station and the starting point for the Crosstown’s extension westward to Pearson Airport.

Here’s a zoomed-in view:

Mt Dennis Gondola.png

9. Pearson Airport 

John Lorinc of Spacing Magazine writes:

I have every confidence that not a single person involved in the advocacy or planning debate over this [Pearson Airport transit] hub will ever schlep by bus or foot through the sidewalk-less industrial wilderness north of Pearson, asking themselves whether the last-mile problem isn’t really a last-ten miles problem…No compelling evidence exists to support an airport multi-modal hub which purports to serve the super low-density employment zones nearby, or shuttle executives between Pearson and the office parks in Mississauga or Waterloo. The densities simply aren’t there, and won’t be for a century or more.”

Transitway Mississauga.png

The Eglinton Crosstown west of Mt. Dennis has not yet begun construction, and, even when it is done, it is not clear that it is a good thing to have, as is planned, westernmost final stops at Convair or Silver Dart. These stops would not serve many people, and they make the trip to the airport slower than it would otherwise be (and it will already be fairly slow, since  will not be an underground line west of Mt. Dennis). It might be better to build the line without stops at Convair or Silver Dart, either having the LRT run direct from Eglinton the airport or else having it not leave Eglinton at all.

The airport could instead be made accessible from Eglinton via the Mississauga Transitway’s rapid bus, and/or via the Kipling Subway Station bus (the 192 “Airport Rocket”, which uses Highway 427 to get from Bloor to the Airport, making only one stop, at Dundas St., in between). A car-sharing lot at Eglinton and Highway 427’s intersection could also be a way to get from Eglinton and other roads to the Airport.

Pearson Gondola.png

A 5.5 km line linking Eglinton Crosstown to the Airport, to multiple highways, and to the Mississauga Transitway “MiWay” Rapid Bus corridor.

Whatever does happen with the LRT, a gondola from Eglinton to the Airport could be a way of linking the two together, and also linking them to a cloverleaf car-sharing parking lot and/or ride-sharing area. The gondola could, perhaps, also provide access to Convair and Silver Dart, so that a future LRT would not have to make stops there.  Finally,  a gondola built from Eglinton to the Airport would provide interesting views of the Airport’s southern runways.

10. The Don River Portlands 

A 1.8 km line starting at Portlands’ marina in the south, which would cross a 110 metres wide channel of water, then a 50 metres wide channel of water, then the Gardiner Expressway, and finally a 100 metres wide GO rail tracks, ending at the Corktown Common.
Portlands Gondola

Portlands Gondola Zoom in

Valley Network Toronto .png

The Don Valley Network

The Next Big Thing 

Mont Blanc

A cable-car station near the peak of Mont Blanc, the highest mountain in the Alps

Even if urban gondolas do experience another renaissance, the natural environment of cable-cars will of course still remain the world’s mountainous areas. Today, in the United States, mountain living is not very common. The only medium-elevation cities of any significant size are Denver, El Paso, and Albuquerque; the largest city higher than those is Colorado Springs, which has only around 400,000 inhabitants–and is still more than 400 metres lower than Mexico City. The highest US “city”, Leadville (pop. 2,000 ), is around a kilometre and half lower than El Alto (pop. 650,000) and barely higher than Bogota (pop. 8 million).

Mountain living is difficult: the reason that places like South America have so much more of it is that high-elevations provide a cooler-climate alternative to the tropical lowlands or rain-forests they overlook. Will Americans take to the hills more often in the coming years? Certainly people do like to go on mountain vacations, if they can afford it. So, if new technologies, like autonomous cars, trucks, snowplows, and aircraft, or like tele-commuting, allow living and transporting bulk goods in mountains to become cheaper, then it may indeed be that more Americans will choose to live in a more frequent state of apres ski.  


In the nearer term, though, gondolas could be one of the surprise stories of city life. With intermodal transportation — easy parking, easy transferring, and easy public transit — it makes a whole lot of sense to provide more connections between superhighways, valley expressways, subways, railways, parking lots, and even ravines and waterways. In many cases, the cheapest, most efficient, and least intrusive way of doing this is with a flying car. A cable-car!

North America

Ontario: The Borderland Economy

Source: RBC, predictions from March 2015

Source: Royal Bank of Canada, predictions of provincial economic growth for 2015, published March 2015

With the economy of Western Canada hit hard by the fall in oil and other commodity prices that began last year, Canada’s most populous province, Ontario, has begun to account for quite a large share of the country’s economic growth. Many Canadian economists – most of whom live in Ontario, as I do – assume this economic resilience is the result of Ontario’s economic diversity and size. Ontario’s population is much larger than that of any other Canadian province (see graph below), and its economy is mixed between services (in Toronto),  government (Ottawa), industry (southwestern Ontario), and commodities (northern Ontario). Ontario’s economy is also more oriented toward the auto sector than other provinces are, and so may be benefiting more than others from the fall in oil prices.

the provincials

Ontario accounts for around 38 percent of Canada’s population, compared to 23 percent for Quebec and 13 percent for British Columbia. Most other countries do not have provinces/states that are as large as this. California, for example, is the largest state in the US but has just 12 percent of the US population; source: Future Economics

Still, this may be missing the point to a certain extent. What really sets Ontario aside from other Canadian provinces is the proximity of large population centres in Ontario to large population centres in the United States. This is unique among Canadian provinces (see graph below), particularly if you ignore Quebec (which is separated from US populations by a language barrier as well as a political one) and British Columbia (which, perhaps not incidentally, is the other major province that has decent economic growth right now, in spite of the fact that it is a significant commodity exporter and has close ties to oil-rich Alberta). Ontario is the only province to have a handful of cities which straddle the US-Canada border. These include Detroit-Windsor, Buffalo-Fort Erie, Niagara Falls, Sault St Marie, and Sarnia-Fort Huron.

US-Canada 15


Since the US economy has remained relatively strong in recent years, unlike those of Europe, East Asia, or much of the developing world, Ontario’s ties to the US may be what is driving Ontario’s economic growth. This should make Ontario concerned; the US economy has not had a recession for almost eight years now, so, in a certain sense at least, it is due for one soon.

Below, I have tried to show some of the ways in which Ontario’s proximity to the US is unique. I’ve gathered all the data myself using Google Earth and recent Canadian and American censuses, so if you think you’ve found any errors in the following graphs please let me know.

us-can 50.pngcan-us 50 land:water.pngus-can 100.pngus-can 200 .png

us-can neighboursus-can 1st and 2nd degreelength us-can

US-Can real gdp.png

us-can gdp growth

[New Brunswick had zero in this category because Maine’s economy has been neither growing nor shrinking in the past year or so. Alberta and BC are high in this category because of the growth of Washington state and Montana, respectively. Saskatchewan and Manitoba were doing great in this category before the oil crash caused their shared neighbour North Dakota to go from the fastest-growing US state into a serious recession. Ontario and Quebec are roughly equal in this category because of the huge size of New York state, which they both border. However if you were to ignore New York state, then Ontario’s border states, namely Michigan and Minnesota, are growing much faster than Quebec’s border states, namely Maine, Vermont, and New Hampshire. Minnesota’s growth, meanwhile, is also why Manitoba is relatively high in this category in spite of North Dakota’s recession — as Minnesota’s  GDP is nearly seven times higher than North Dakota’s]

Ontario’s ties to the US have also meant that it is less dependent on inter-provincial trade of goods than other parts of Canada are: in recent years Ontario has conducted 2.5 times more trade with other countries (led by the United States, of course) than it has with other provinces. This is compared to just 1-1.5 times more for Quebec, Alberta, and British Colombia.

Economists and financial journalists in Ontario need to be more careful than they have been in the recent past. During the 2007-2009 economic crises they ascribed the relative success of Canada’s financial system (which is centred in Toronto, Ontario) to the fact that Canadian bankers and regulators were more prudent and conscientious than their peers in other countries, rather than to the fact that Canada was flush with capital at the time as a result of the sky-high commodity prices that existed just before and just after the financial crisis, and as a result of the fact that Canadian Baby Boomers  were then in the prime of their financial lives (as Canada, unlike the younger US or older Japan, is dominated by the Baby Boomer generation).

But instead of acknowledge these facts, much of the Canadian media decided instead to help create a cult of personality around Canadian bankers and Bank of Canada leader Mark Carney — a cult of personality they have since exported to Britain, where Carney has become a figure of great importance (especially since Brexit and the resignation of Prime Minister David Cameron) and the first non-Briton to ever become the central banker over the British financial system, a system that is far larger, far more worldly, and far less dependent on  commodity sectors than the Canadian financial system is. Similarly, Ontario’s economic resilience is now being described (by some people) as if it was basically an inherent condition of the Ontario economy, rather than a result, at least in part, of Ontario’s unique ties to the growing US economy.

Don’t get me wrong: I am not saying that Ontario is not a resilient place or that bankers and regulators in Toronto and Ottawa are not prudent and wise. And certainly I would like Ontario’s economy to continue growing, since it is my home. However, believing either one of these stereotypes about Ontario too much could be a dangerous mistake for investors or governments to continue to make.

North America

Fasten Your Snowbelts – Technology and the Great Lakes

Outside of the Rocky Mountains, most of the snowfall in the United State falls within the Great Lakes “Snowbelts”. So too does a significant portion of the snow that falls within Canadian cities. These Snowbelts are located, almost entirely, in Michigan, upstate New York, or Ontario:


The Great Lake Snowbelts

US Snow Map .png

Source:; Portacup

In the map above, which shows average yearly snowfall in the more than 3000 counties of the United States, there are just 29 counties (according to my count) that receive 120+ inches of snow, 30 counties that receive 100-120 inches of snow, 50 counties that receive 80-100 inches of snow, and around 100 counties that receive 60-80 inches of snow.

13 of the 29 with 120+ inches of snow are in the Great Lakes Snowbelts (8 of the remaining 16 counties with 120+ inches of snow are in Colorado). 15 of the 30 counties with 100-120 inches of snow are in the Great Lakes Snowbelts (compared to 5 in Colorado and 5 in Vermont or Maine). 24 of the 50 counties with 80-100 inches are in the Great Lakes Snowbelts (compared to 7 in Colorado, 4 in Alaska, and 9 in Vermont, New Hampshire, or Maine). And about half of the 100 or so counties with 60-80 inches are in states which border the Great Lakes.

As you can see in the maps below, this has had a big impact on urban development within the Great Lakes basin. The largest cities, namely Chicago, Toronto, Detroit, and most of Cleveland-Akron, are located outside of the region’s snowbelts. Chicago, for example, gets only a third of the snow on average that Rochester, NY gets in any given year, and a sixth of what cities like Oswego, NY get.

great lake pop


average snow.png

With the exception of Sault St Marie, each of these cities has a population of at least 100,000. Sault St Marie is the quintessential Great Lakes city, however; it is located around the place where Lake Superior, Lake Huron, and Lake Michigan, the three largest great lakes, converge. It has a population of 75,000 on the Canadian side of the city and 14,000 on the US side. New York City, meanwhile, is obviously not on the Great Lakes, but I included it anyway to show as a comparison    …also, please forgive our misspelling of “Erie” on the graph above

There are, of course, some notable cities within the Great Lake Snowbelts. Buffalo, for example, which serves as New York state’s outlet on Lake Erie, was the 8th most populous city in the United States in 1900, and the 4th most populous city in the US that did not have an ocean port. (A year later, in 1901, President Mckinley was assassinated in Buffalo at the Pan-American Exposition). This was before the construction of the US’s road and rail networks stripped the Erie Canal, and thus Buffalo, of most of its economic significance. Today Buffalo is estimated to be just the 76th most populous city and the 46th most populous “urban area” in the country. It has recently had one of America’s fastest shrinking populations.


Erie Canal

Erie, Pennsylvania, meanwhile, was the country’s 69th most populous city in the US in 1930. Erie once served as the meeting place for three separate American railway networks, which used different gauges as one another, before these networks were standardized during the middle of the 19th century. Today Erie is not even in the top 300 most populous cities in the US and is just its 183rd most populous “urban area“. It too has a fast-shrinking population.


Rochester and Syracuse were the 22nd and 40th most populous cities in the US in 1930, respectively, but are now just the 103rd and 182nd most populous cities and the 60th and 90th most populous “urban areas”. Rochester and Syracuse serve as New York state’s outlets on Lake Ontario, just as Buffalo does on Lake Erie. Rochester is located 11 km inland from Lake Ontario and Syracuse 53 km inland, however, unlike some of the non-snowbelt cities on Lake Ontario like Toronto, Hamilton, and Kingston, which are situated directly on the lake.

Grand Rapids is listed as the 123rd most populous city in the US and the 70th most populous “urban area”. It is Michigan’s second largest city and serves as the state’s chief outlet on Lake Michigan. Like Syracuse, it too is located inland: it is 50 km upriver from the Lake. As you can see in the map below, Grand Rapids is the only place along Lake Michigan’s coast where the lake’s coastal lowland (the green areas on the map) extends relatively far inland.

great lakes

Note, by the way, how Lake Superior presents challenges to urban development: it is further north than the other Great Lakes, has snowbelts on both its eastern and its southern shores, and it has very narrow coastal plains overlooked by escarpments. The main Lake Superior port city of Duluth, Minnesota gets more snow (86 inches) than any other city in the state –  a lot more snow in most cases. (Minneapolis gets 54 inches). Thunder Bay, which is Canada’s primary Lake Superior outlet, gets 64 inches of snow.

average snow.png

I forgot to add Syracuse, which should be second on this list: it gets 123 inches of snow on average, according to

Cleveland’s numbers on the graph above are somewhat misleading. They are skewed upward because some of Cleveland’s suburbs, like Broadview Heights, usually get much more snow than Cleveland proper does. Akron, meanwhile, gets even less than Cleveland: just 47.4 inches of snow a year. Akron gets less than half of what nearby Erie in Pennsylvania gets. Pittsburgh, by the way, gets 41.9 inches of snow a year, on average.

eerie snowbelt

Outside of the big and medium-sized Great Lakes cities, there are places in the Snowbelts that get even more snow. Lake Ontario coastal towns in the area from Oswego (population: 18, 142) to Watertown (population 27,823), get well over 100 inches of snow a year. Marquette in Michigan and Owen Sound in Ontario do too. And even smaller places like Boonville, New York (population 2056), in the foothills of the Adirondacks, and Hancock in Michigan’s Upper Peninsula (population 4596) get more than 200 inches on average. According to this source, Hancock is the snowiest city in the United States among cities or towns with at least 1000 residents, with the exception of Valdez, Alaska or Crested Butte, Colorado.

watertown to oswego

…assuming you don’t get caught in a whiteout


Why does any of this matter? 

The Great Lake Snowbelts have posed challenges for urban development thus far. Buffalo is by far the most populous snowbelt urban area, and even Buffalo is not a big city. And apart from Erie, Rochester, and Syracuse in the US or London, Barrie, and Sudbury in Ontario, there are no other Great Lake cities that get more than 70 inches of snow a year on average and have populations of at least 100,000. It simply has not made sense to grow a city in a place with so much snow.

Technology, however, could be a game-changer when it comes to dealing with snow and with snowstorms, which could in turn could give a boost to the economies – or real estate values – of this region. This is particularly relevant given that this is a region that otherwise has great assets, such as the Great Lakes and physical proximity to Manhattan, Toronto, Chicago, and other major North American cities. Indeed, even the snow itself can be an asset once its limitations can be overcome. Snowfall is not only beautiful, but also provides recreation (skiing, cross-country skiing, taboganning, etc.) and can help to prevent forest fires.

One technology that could help the Snowbelt is self-driving snowplows. It makes sense that self-driving snowplows should come into operation even before self-driving cars or trucks do, as snowplows are often in demand overnight, when few other drivers are on the road and labour costs are high. Snowplows drive and work slowly, meaning that plowing roads normally tends to be labour-intensive and that self-driving plows working overnight would probably not be as much as a safety hazard as faster self-driving cars might initially be.

Similarly, snow-clearing robots can help clear parking lots, sidewalks, and, most helpfully, rooftops. Rooftop snow can be especially damaging to buildings, and is often difficult, expensive, and time-consuming to clear.

Another technological change is e-commuting and e-commerce. If you do not want to commute to work following a heavy snowfall, you may now work from home instead, or from an office or co-working space near your house. And you may order your groceries directly to your house.

If you do not want to drive while it is snowing or has recently snowed, you may also soon be able to use an app like Uber or Uberpool to get around in a vehicle that can better handle harsh conditions.

And if you are driving, you can use tools like GPS and smarter cars to better handle snow. Today, driving in a road that does not have street lighting while it is snowing can be hugely irritating and dangerous; when you turn your headlights on the light ends up scattering off the snowflakes, making it nearly impossible to see. There is also often snow covering the surface of the road, making it difficult to see the barrier between your lane and the lane for oncoming traffic. You often have to drive extremely slowly, and even then can easily get your car stuck in a snowbank or suffer a car accident.

GPS can help you can quicker emergency response or roadside assistance. And technologists are working on tools to help cars and trucks navigate through heavy snowfall and help drivers to avoid getting into accidents when in rough, wintry conditions. Self-driving trucks travelling overnight could also help get trucks off the road during the daytime, and to save truckers the trouble and danger of having to drive through a snowstorm.

Now, a little bit more on the Snowbelt:

Nearly all the Snowbelt are in Ontario, Michigan, or upstate New York. The snowbelts in upstate New York and Eerie, Pennsylvania are strategically located at the “backdoor” to the Boston-to-Washington megacity (which you can see in the map below), and are similarly adjacent to the Toronto-to-Detroit (or, more broadly, the Montreal-to-Chicago) region.

pop dens .png

population density/major urban areas in the US

According to Accuweather, of the 10 snowiest “major” cities in the world, three are in New York state (Buffalo, Rochester, and Syracuse), three are in Canada (Saguenay in Quebec, St John’s in Newfoundland, and Quebec City), and four are in northern Japan (Sapporo, Aomori, Toyama, and Akita). Aomori gets the most by far: 312 inches, compared to second-placed Sapporo which gets “just” 191 inches.

In Canada, unlike the United States, the Great Lakes Snowbelts don’t dominate in the snowfall category, since places further north where the weather is colder often get more. French-speaking cities like Quebec City, Saguenay, and Sherbrooke, for example, get more snow than places throughout much of the Great Lake Snowbelts, mainly because they have very cold climates.

average snow in canada

St John’s, Moncton, and Cape Breton, on the other hand, which are in Canada’s Atlantic Maritime provinces, are actually relatively warm, yet still receive enormous amounts of snow. St. John’s,  which is the largest and snowiest of these three, is actually one of the warmest cities in Canada outside of British Columbia during the winter; its coldest month is February, when it averages highs of -1 degrees Celsius and lows of -9 degrees Celsius. But in return for this “balmy” winter weather St John’s also gets cooler summers than other Canadian cities: its average high in August is just 20 degrees Celsius (68 degrees Fahrenheit).


Finally, one last thing on Snowbelt snow patterns, from They can be highly erratic and different to forecast in advance:

“Lake-effect snow has been a forecaster’s nightmare from when maps were drawn by hand to the current days of computer predicted models. But no one computer can accurately predict the magnitude or severity of a lake- effect event with the same success as a synoptic event. To describe lake effect snow as temperamental would be a gross understatement. Often arranging itself in rogue bands of heavy snow, lake- effect can stop and start on a dime, and it can dump a foot of snow on one neighborhood and leave the residents of another wondering why the idiot meteorologist keeps breaking into Oprah about some kind of Lake Effect Snow Warning.

And in its unpredictable nature comes its beauty. One of natures precious wintertime treats, just the prospect of lake- effect snow strikes both fear and awe in the hearts of a forecaster. There are, however, trends. Subtle nuances that fade in the background to the untrained eye, but trends nonetheless. If nothing else, these trends offer a faint possibility that maybe, just maybe, Mother Nature may be following a game plan all along.

…Conditions [for lake-effect snow to form] are so difficult to achieve in one given place that lake effect only occurs in four locations on the entire planet: the southeast shoreline of the Hudson Bay, areas just east of the Great Salt Lake in Utah, the northernmost Japanese island of Hokkaido, and of course, the Great Lakes.

Lake snows generate downwind of the Lakes. Sure, lots of people live near a Great Lake, but only a few lucky ones live downwind. Downwind though, is very much a relative term. One day, it takes a west wind. Another, a north wind. But that’s just one piece of the puzzle. Pace yourself. Meteorology follows from this point on.Read more here.



Images, North America

Image of the Day — Unique New York


Last a month a report in the New York Times suggested that Michael Bloomberg, the mayor of New York City from 2002 until 2014, has been thinking about running for President of the United States as a third-party candidate, and may be willing to spend as much as a billion dollars of his own money to do so. Today, on the sole day between the end of football season and the start of ex-Iowa primary season, Bloomberg himself confirmed that report.

According to MarketWatch, this is “the first time Bloomberg himself has said he might run, though his surrogates have told other outlets the former New York City mayor and founder of Bloomberg LP was considering such a move. ‘I find the level of political discourse and discussion distressingly banal and an outrage and an insult to the voters,’ said Bloomberg”.

The Bloomberg strategy is a fairly simple one: first you take Manhattan, then you take D.C. The idea would be for him to secure huge amounts of donor money and media support available in New York City, as well as the 5.4% of America’s electoral college points that you get by winning New York state in the general election, and then use those assets in order to lure people outside of New York to vote Bloomberg on election day too, hoping that enough Americans will not want to vote for a non-centrist candidate like Donald Trump, Ted Cruz, or Bernie Sanders.

While it still seems quite unlikely that Bloomberg would attempt to pull this off, it cannot be ruled out entirely, in particular because of the following constitutional catch: if none of the candidates in the general election wins more than 50% of the electoral college votes, then the president of the United States is chosen instead by a state-majority vote in the House of Representatives, wherein the congressmen and congresswomen representing each state vote amongst themselves to determine which presidential candidate their state desires. The candidate which has the backing of a majority of states then becomes president, while the vice-president is chosen by the Senate.

In such a vote, Alaska’s one congressman would have as much power as all of California’s dozens of congressmen put together. This vote would probably favour right-wing establishment candidates, since most states and congressional districts in the country tend to be relatively right-wing, Congress is currently controlled by a Republican majority, and congressmen tend to be establishmentarians. If it ever came to this, then, a candidate like Bloomberg might have something of an advantage over one like Cruz, Trump, Sanders, or, perhaps, Clinton.

Could it ever come to this? Well, it did in 1824, when “four candidates ran for president… Andrew Jackson got the most votes from Americans and the most votes in the Electoral College, but not a majority, so the race was turned over to the House of Representatives voting as states who picked John Quincy Adams instead.” Crucially, if Bloomberg could secure a victory in New York state in the general election, that alone might make it relatively difficult for one of the other two candidates to win an electoral college majority.

If, for example, Jimmy Carter had lost New York in 1976 to a third-party candidate, Carter still would have gotten more votes than his Republican opponent Gerald Ford, yet would have fallen short of the electoral college majority needed to avoid turning the vote over to Congress. Had a third-party candidate won New York, Pennsylvania, and Iowa instead of Obama in 2012, Obama would not have won a majority in the electoral college, which would have meant that the Congress would have been able to vote to select the president of the country instead.

What Bloomberg’s public presidential mulling-over really indicates, then, is the enduring power of both the state and city of New York. This is actually not just a Bloombergnagian phenomenon: Hilary Clinton served as one New York state’s two senators from 2001 to 2009, Bernie Sanders was born and raised in Brooklyn, and Trump in Queens. (Even Chris Christie has influence over bridges that reach New York). For a while there was also some presidential buzz about New York’s current governor Andrew Cuomo, who’s father Mario was also a longtime governor of New York and came relatively close to becoming president in 1988 and 1992.

At this point, in fact, the only leading candidates who do not have ties to New York are those who have ties to Florida: Rubio is a Florida senator and son of Cuban immigrants, Jeb Bush used to be a Florida governor and is near-fluent in Spanish, Cruz’s father is from Cuba (though Cruz is himself a senator representing Texas), and Trump has usually lived in Florida when not living in Manhattan.

New York is also a somewhat peculiar state, politically. Though on the one hand it is the heartland of liberal America (along with California, of course), on the other hand it has politics that are in some ways quite Republican-esque. It has, for instance, groups that are strongly pro-finance (because a lot of its money and jobs come directly or indirectly from Wall Street), pro-Israel (because it is home to an estimated 26% of America’s approximately 6.8 million Jews …some of whom would also be happy to see Bloomberg or Sanders become the first Jewish president rather than see Hilary become the second Clinton president), and pro-security (because it has been the main terrorist target in America, not only on 9-11 but in many other cases as well).

New York also has a sizeable right-leaning “upstate” region, in which potentially significant shale energy reserves are located. Unlike in nearby states, such as newly gas-rich Pennsylvania, New York has not yet been allowed to develop these resources. Some upstate New Yorkers may therefore be hoping for a president who will support the removal of the fracking moratorium in the state, so that they too can participate in the regional shale bonanza.


The last time New York voted for a non-Democratic candidate was in 1984 when, along with most of the other states in the country, it approved of a second term for Ronald Reagan. In the following election of 1988, however, New York was one of just 10 states, and the only populous state, to vote against George H. W. Bush, who had been Reagan’s vice president. Will New York vote against a candidate from the Democratic Party again in 2016? Probably not, but of course we will soon find out.