Official unofficial neighborhoods, 2017

Time for an update on the classic-yet-ongoing neighborhood mapping project! To recap: we’re on version 2 of a web survey for mapping Boston-area neighborhoods, the boundaries of which can be the subject of much disagreement. We’ve asked people to draw neighborhood extents as they see them, in an attempt to collect enough of these crowd-submitted shapes to be able to say something about the overall popular perception of neighborhood geography.

Here’s the fuzzy map of neighborhoods according to a snapshot of some 2,300 submitted polygons in our data, as of March 2017.

Boston neighborhood consensus

This is a fun update over previous editions mainly for two reasons: it includes areas outside the city limits of Boston, and it is not limited to a pre-selected set of neighborhood.

The way this map works is:
1) Lay a somewhat-fine grid over the area. These hexagonal cells are something like 75 meters across.
2) For each cell, find the most commonly named neighborhood intersecting it.
3) Measure consensus on that cell as the number of submissions for that neighborhood intersecting the cell, divided by the total number of submissions for that neighborhood. Thus we can say something like, “75% of people who drew the South End agree that this location is part of the South End.”
4) Ignore any neighborhoods with fewer than 5 submissions. This weeds out all the joke/rude submissions, but also is just an arbitrary minimum for measurable consensus.

The consensus map does a decent job of showing where boundaries are hard and where they’re fuzzy or uncertain. Without that, a simple map categorized by name looks like this, in hasty random color screenshot form:

Boston neighborhoods categorized

It’s also worth doing a quick comparison to official boundaries. Here’s one with official lines for Boston and Cambridge overlaid.

Boston crowdsourced vs official neighborhoods

Boston does a pretty good job, with a few exceptions that I notice:

  • People prefer to subdivide downtown and the waterfront area into smaller neighborhoods.
  • Official Mattapan intrudes on Dorchester a bit—exactly how much is hard to say from our data, but certainly the bit beyond Morton Street looks too far.
  • Official Hyde Park eats a slice of Mattapan, even some bits with 100% Mattapan consensus in our data.
  • Official Jamaica Plain appears to intrude on Roslindale.

Much of Cambridge, however, is quite different, as people tend to think of the city in terms of squares like Central or Harvard, while the official boundaries put the squares between neighborhoods.

Keep the submissions coming! It would be great to see more neighborhoods pass the threshold and appear on this map, and as always it’s fascinating to see how the shapes evolve and settle. Thanks for everyone’s participation so far!

Posted in General | Tagged | 1 Comment

The Boston Projection

Boston is the center of the world

Cartography Twitter has become abuzz with some local map news: last week Boston Public Schools ditched the classic Mercator projection in favor of the Peters projection for classroom maps. A trivial-sounding story to some, it’s interesting because it’s change motivated by social justice—seemingly bold but in fact a late shot in a decades-old cartographic controversy. So, perhaps we can briefly share a bit of background on map projections with Boston-area map readers.

The first thing to know is that there is no such thing as an undistorted flat map. Every single one is “wrong” in some way or another. A globe is the only way to represent the world without significant distortion. Any attempt to flatten out the round Earth sacrifices accuracy in at least one way: size, shape, and/or direction.

The Mercator projection, which you see in things like Google Maps, and which you probably knew from classroom walls, is known for gross distortion of size at high latitudes. Thus Europe appears bigger relative to, say, Africa than is true; and Greenland looks to be the size of South America when in reality it’s something like one eighth the size.

The true size of Greenland

Although areal distortion is a side effect of the Mercator map’s original purpose—it preserves angles and was useful for navigation—some have argued that its exaggeration of northern latitudes promotes a very Eurocentric view of the world. Other map projections that maintain size accuracy (equal-area projections) are viewed as socially just, giving due attention to parts of the world long oppressed by European colonialism. In the 1970s Arno Peters made some waves promoting the projection which will now grace the walls of Boston schools, and which famously (to cartographers) made an appearance on The West Wing. BPS is a little late to the party, but has adopted the map for essentially the same reason: to remove an imperialist image and promote a different view of the world.

Many cartographers are critical of this projection, as the claims behind it are not novel and even erroneous, with some frustration that it proclaims “truth” yet necessarily trades one form of distortion for another. (Notably, shapes are badly distorted.)

No matter what side one takes, the fact is that a map projection is one of many tools in the cartographer’s kit that can be used to subtly—or boldly—promote a certain agenda or worldview. And as everyone knows, the only accurate world maps are those that correctly show Boston at the center of it all, right?

(adapted from Mike Bostock’s Projection Transitions)

Posted in General | 1 Comment

Much Sass State

Can you believe it’s almost the 400th anniversary of the the Pilgrims arriving at Hot Lumpy?

South Shore and Cape anagrams

Maybe you’re more of a North Shore type, and are excited for summer so you can visit the beach at Meth Cranes By-The-Sea or Wry Nut Probe for a swim in the Anal Cat Notice.

North Shore anagrams

Yeah, it’s a Massachusetts anagram map. Designed by our inner 12-year-olds.

Massachusetts Anagram Map

Anagrams were all generated by the amazing Internet Anagram Server (or I, Rearrangement Servant).

Posted in General | Tagged | 4 Comments

Myriad marathons


Happy Patriots’ Day, everyone!

Today thousands of runners are making that famous trek from Hopkinton to Boylston Street. That’s The Boston Marathon, the only one that counts, but there are many ways you can run 26 miles and 385 yards to reach that finish line. We thought it would be fun to put them all on a map. So there they are above: 581 different 26.2-mile routes to the finish line from all directions—plus the one official course.

If you protest that the map surely doesn’t show every possible marathon route, you would of course be correct. The number of unique routes of marathon distance is practically infinite. You could run all kinds of crazy routes and barely even leave the neighborhood! How about a spiral marathon?

Spiral marathon route!

What’s on our map is only a sample of routes that start outside the city, as the actual Marathon course does, flowing together as they approach Boston. We came up with a set of around 4200 candidate starting points, threw them at MapQuest’s open directions API both for walking and driving directions, and kept the routes that turned out to be close to the right length. We also did our best to toss any routes that use limited access highways. Best not to run on those.


Whether you’re running the Boston Marathon today or one of the myriad others, or are just taking a spin around the neighborhood, good luck and happy running!

Posted in Events | Tagged | 2 Comments

A day of traffic

If you’ve driven a car at all around these parts, you’ve certainly spent some time as a part of slow or stopped traffic. Despite how it may feel when you’re parked on the Southeast Distressway, Boston is, relatively speaking, not so bad. There are other cities in the US that have it worse. (See data such as the Urban Mobility Scorecard from Texas A&M.) Outside of special events or bad weather, what we have is a fairly predictable rush hour cycle. Traffic is slow for a few hours in the morning and early evening, and generally fine at other times.

Or so it appears to this layperson, at least. Not wanting to claim any expertise, I will leave this subject with a simple map of the ebb and flow of traffic in Boston. Here’s approximately 20 hours of traffic yesterday, a seemingly ordinary day, according to Google Maps.

The automated screen captures for that animation were collected using a script by Alyson Hurt of the NPR visuals team who was mapping something legitimately fascinating: traffic map as weather map. Next time the internet is abuzz about a winter storm, look at a live traffic map and you’ll get a pretty good idea of where it’s snowing.

Posted in Transportation | Tagged | Comments Off on A day of traffic

The island nations of Boston

The island nations of Boston

Some neighborhoods of the Boston area are actual islands, or were at one point. Others, however, can feel that way even when connected to each other by land. Water isn’t the only thing that can create a gulf between neighborhoods; sometimes it’s created by features of the urban landscape and the experience of passing through them.

To me, one of those things is an unpopulated area. Even if two pieces of the city are contiguous with no real empty space between them, if there’s a zone between them where nobody lives (perhaps a commercial district or even a nice park), the experience or idea of passing through there increases the distance between the two neighborhoods in my mind. Thus a neighborhood that is physically close becomes a “somewhere else” to me, and I remain on my island.

The above map aims to represent that experience, imagining land and water as defined by population. Any unpopulated areas (according to census data) become water, population density becomes the terrain, and Boston becomes a rugged world of islands and lakes.

Posted in General | 1 Comment

Map your neighborhood! (Again!)

When you think of your neighborhood, you probably think of many things—people, institutions, cultures… and geography. In your mind it probably has a spatial extent, or a center, or some fuzzy definition of what is “here” and what is “there.” Share that geography by drawing your Boston-area neighborhood on this map. Why do it? Quite simply, because if enough people contribute to the map, it will be a fascinating data set that shows we collectively see our city. It’s mostly for fun, but possibly meaningful to policymakers and such.

We’ve done crowdsourced neighborhood mapping before and were pleased with the response. Longtime payers of attention to this space will remember the project we started a couple of years ago:

Boston neighborhoods consensus

See some maps and discussion of that project here, here, and here. It was a fun project, and we were able to see some cool results. But that project was limited in several ways.

So we’d like to do it again… but better! If that’s all you need to hear, head straight on over to the new version of this survey and draw your neighborhood. Or read on to see why we’re doing this again, what we’re doing differently, and how the results can be much more meaningful.

The original idea was a reaction to regularly written news articles and internet comments about disputed boundaries between neighborhoods in Boston, and the fact that no single official set of boundaries exists. Fine, we thought, if you all say everyone else is wrong, show us what you think is right. And although the result of this, expectedly, seemed to represent fuzzy notions of place, the whole concept really revolved around questions of division and dispute.

Now we’d like to approach this as a general question of geography and the image of the city. The entire premise is much more open and, we think, much more in tune with what neighborhoods actually are—personal spaces, subjectively defined. We don’t want to ask “where is the line between these two places?” but rather “what is this place?”. Although in the end the map will still be built upon lines and discrete shapes, the data should now be more versatile and meaningful. Thanks to our carto-friend Nick Martinelli, there’s an easy implementation of a survey to address some of the new questions. With all that in mind, below is a summary of some of the improvements in this version of the project, which we think will provide a lot of new and interesting possibilities.

A wider scope

It was clear early on that our original project was very narrowly defined, and thus only narrowly meaningful. It was restricted to the city of Boston and a specific set of neighborhoods. While this did address some particular questions, with those limitations it’s impossible to get a real picture of the subjective internal boundaries of the broader “city” of Boston and its surroundings. Not everyone in Boston identifies with the a neighborhood in our list, and not everyone here lives in the city of Boston. So the new version leaves everything much more open. There is no predefined list of neighborhoods (you provide the names), and it’s meant for any and all cities/towns in Greater Boston.

A wider audience

We were always very hesitant to draw any real conclusions from the first project, and a big reason for that is that it is a self-selection bias. Quite simply, our data did not represent residents of Boston—it represented people who look at this site or other sites that linked to it. We don’t know exactly who those people are, but we’re pretty confident that they’re not a representative sample of people who live in the city.

The new version certainly doesn’t solve that by itself, but we’ve taken a couple steps to make it easier to reach more people. For one, it’s mobile-friendly, so anyone on a smartphone can contribute. Second, the drawing method is, we hope, a little simpler. Instead of clicking a multitude of vertices, you just trace a shape. We imagine that this would work well if someone were to, say, carry around an iPad to survey people. Just put the tablet in front of someone and ask them to draw on the map with their finger.

Not just borders

Not everyone thinks of their neighborhood in terms of where it starts and ends. For many, it’s a collection of points—perhaps a square, or a store, or a church—nodes and landmarks in the Kevin Lynch sense. In an attempt to capture neighborhoods as collections of landmarks, the map lets you drop a point marker as something that defines a neighborhood. It might be what you consider the central point of the neighborhood, or it might just be some place that you strongly associate with the neighborhood. What it means is up to you!

Authoritative data

Another shortcoming of the first version is that we had no way of knowing how trustworthy a submission was. It could have come from someone who has lived in the neighborhood for fifty years, or it could have come from someone who’s never even been there. Unless it was obviously a random scribble, we had no choice but to treat everything equally. The new survey, however, includes a couple of optional questions asking respondents how long they have lived in the neighborhood and city, if at all. This may be helpful in analyzing and mapping data, for example allowing us to give more weight to longtime residents, or to map how boundaries may have shifted over time.


Neighborhoods aren’t just shapes or points on a map. They are collections of experiences. A fun new feature here lets people share stories or comments about the neighborhoods they’ve drawn. You might use it to explain why you drew the shape the way you did, to share an anecdote about the place, or any number of things. There’s only so much a line on a map can say, so we want to make sure there’s a way to tell the rest of the story.

See results

Another new thing here is that you can see what everyone else has drawn so far. You might use this to guide your own drawing (although steering clear of the influence of others is preferred) or simply to browse and compare your map to the crowd’s. You can also vote, as it were, for neighborhoods that people have drawn, helping lend more weight to those. You’ll also be able to see any stories that people have attached to their neighborhoods. The raw data itself is also much more accessible to everyone.


Finally, and not insignificantly, this new version is a big upgrade in technology. Whereas the first one used Google Maps and a disaster of a database, the new one uses open-source maps on the front end (built on Leaflet and map tiles from Stamen) and the super powerful and easy-to-use CartoDB to store and display data. The drawing itself uses an unholy hacked-up combination of Leaflet.FreeDraw and Leaflet.Editable. If you’re not one for map technology, the upshot is that all this will make it much, much easier to submit, extract, and analyze data in different ways.

One of the best consequences of the original project was that it inspired Nick Martinelli to develop a similar project for several cities in the Pacific Northwest. Nick built everything in a much better and more reusable way, and posted his code to GitHub. That, in turn, is what we’ve used for our new version, along with a few modifications. So, thanks Nick! If anyone else out there is looking to do something like this, take a look at his project. You can get it up and running in almost no time.

Help draw THE map of Boston!

Enough talk! We want to know what Greater Boston looks like in the eyes of its residents. Please take a couple of minutes to draw your neighborhood at:

We will make some composite maps once there’s some data to work with, of course, but you too can look at the data and makes some maps! You can always get up-to-date polygon data and point data. (Warning: you may encounter some rude shapes in that data. This, we have found, is an inevitable consequence of asking people to draw things on the internet.)

Posted in Geography, Projects | Tagged , | Comments Off on Map your neighborhood! (Again!)

Boston is 46 Percent Water

This is not a a paradigm-obliterating viral map. Boston is 46% water and this shows where it all is. Most of the water is offshore. Some of it is inland.

But what if you don’t care where it all is? That little chart in the corner shows the same proportions. It’s a geography teaching tool, without employing—er… geography.

Comparative Geography: Land v. Water

A few years ago, I stumbled on a genre of dusty old geography books: the elementary text book. One of my favorites, Methods and Aids in Geography (1888), features numerous comparative geography illustrations.

At the time, for example, the population of all of New England and New Jersey was equal to that of London and its suburbs. The book also sports maps (which we’ve seen recently, ad nauseam, like number 7 here) comparing state areas to country areas, countries to countries, continents to continents, planets to… ugh, you understand.

Of all the great gems in this book, there is one chart that especially caught my eye, a sort of proto-treemap, comparing the world’s land and water areas.

I like the simplicity of this chart and since Massachusetts is so watery, I decided to reproduce it for our fair state.

I crammed all of these guys…



Inland Water

Offshore Water

Into this chart:

You can be the judge, but I think if you’re looking for area comparisons, this works better than a map. Like, say, this map:

Bonus Map!!!!11!!11!

Land v. water by town (using the stats from census county subdivisions).


Posted in Geography, Historical | 2 Comments

Fixed on Boston, Researchers Enhance Pale Blue Dot

Twenty-five years ago last month, the famous Pale Blue Dot image of Earth was born when astronomer Carl Sagan convinced NASA scientists to turn the Voyager 1 camera around and snap a pixel-sized portrait of our home planet from 3.7 billion miles away.

Now, researchers at the Massachusetts Institute of Technography have developed a recursive spatial unmixing algorithm that allows them to infinitely zoom in on, and enhance, any image.

We asked them to try out their algorithm on the Pale Blue Dot–with their with their sights set specifically on Boston–and incredibly, they obliged! Below is the result of a multi-billion dollar effort in image processing: a snapshot of Boston history, when it was time to make the donuts.

Posted in awesome, General | Comments Off on Fixed on Boston, Researchers Enhance Pale Blue Dot


It’s the end of Winter 2015 and, well, we don’t need to tell you that it has snowed a bit. Last Sunday we celebrated breaking the all-time seasonal snowfall record with 108.6 inches.

Below is an animated map of SNODAS-modeled snow depth during the heart of this winter, when we saw approximately 300 feet of snow fall in the span of three weeks. What’s kind of interesting here is that after a while the big snowstorms become barely noticeable. It mirrors the ground experience of snow depth kind of maxing out at “a metric s***load” — when Mom called after each storm asking how much snow we got, the answer was always “I can’t even tell anymore.”

There’s more to the story, of course, than sheer wonder at big numbers. Those of you who don’t work from home (like Andy) or live somewhere else (like Tim) faced daily misery and probably turned on your neighbors in the Great Parking Wars. But back at a broader level, a fascinating story still unfolding from this winter is the political ramifications of boatloads of snow. It’s been a test for the new governor. It’s shaping public opinion on the 2024 Olympics. And, most notably, it’s caused dustups and debate about public transportation.

This is a map blog, so we won’t dive into the political issues, but suffice it to say that the havoc wrought upon the MBTA is one of the biggest stories of the winter. You all remember the following maps from Sara Morrison—can we even call them facetious when they were in fact totally accurate?

We here at Bostonography are of a more analytical bent (but don’t get us wrong, we love poop emoji), and although we weren’t on the ball enough to look into transit data as the winter progressed, we’d be very interested to see the MBTA delays and outages quantified, especially in a spatial context. If anyone out there is looking into that, please show it off!

Rail transit woes made the biggest headlines, but at some point we did remember that we’ve been automatically mapping bus speeds every day for a couple of years now. Generally it’s very difficult to notice a difference in these maps no matter the circumstances, but here, we thought, we might finally see a trend due to snow-clogged roads. Below is five weeks’ worth of maps, with major snow events marked. Apart from the days where service was suspended entirely, sure enough there is some noticeable reddening in February as the city reached its breaking point. But even then, the effect is much smaller than you might expect. (Takeaway: MBTA buses are always slow, no matter what.)

But just how much show did Boston get this year really? Can old-timers still claim crazier days full of barefoot walking to unheated schools through neck-deep snow? Well, they can claim whatever they want, but this year was incredible. Two of the top ten biggest snow storms in Boston history passed through this winter (a claim that only 1978 can also make). Sure, none of this year’s storms topped the list, but once you get to two feet, what the heck is the difference anyway?

For a quick comparison, let’s take a look at the aftermath of the second and sixth-largest storms in Boston history, a Feb. 6-7, 1978 storm (27.1 in.) and a Jan. 26-28 storm from this year (24.6 in.). We were able to grab clean Landsat images from a few days after each. Here, we compare Martha’s Vineyard.

Boring, right? Different sensors (and band combinations—the 1978 is a false color image) aside, the images just look vaguely snowy. Ok, maybe the 2015 image looks a bit snowier. But that could be because it had a bit less time to melt (or because there wasn’t already ten million inches of snow on the ground in 1978!). Still, looking at these images gives you a sense of what our landscape looks like when it gets blanketed by a large snowfall.

Boston is just out of the 1978 Landsat 2 image, but we get a nice view of it in 2015. The Charles: covered. The Common: covered. Rooftops: covered! Boston was officially snowed in. And it would be for another few weeks.

Looking over the rest of the state in this Landsat image from February 1, you notice dozens of jagged lines. The snow has acted as a sort of terrestrial fingerprint powder, highlighting lines running through our landscape that you wouldn’t otherwise see: utility right-of-ways. That’s right, those gross, overgrown, humming-power-line highways are kinda pretty from space.

But these single snapshots don’t give a very good sense of the process of accumulation or what things might have looked like at different times during the winter. To try to get a sense of these things, we turned to MODIS. Thankfully, though the MODIS instruments are a coarser resolution (250m) than Landat (15m), they take pictures of every spot on earth almost twice every day! At that rate, you can do things like look at so-called winter storm “Nemo” as it passed over Massachusetts two years ago.

With an image or two being taken every day, it is also possible create composite images showing conditions over chunks of time. Here, we have used Charlie Loyd’s Wheather scripts to reduce cloud cover and reveal areas with consistent snow cover over the last three winters. The brightest whites are areas with the most consistent snow cover for that month. Just look at how snowy it has been this year compared to the last two years! Even Cape Cod and the islands remained covered throughout all of February.

It may be snowing in New York at the moment (not that we would know anything about that), but as of 6:45pm tonight, it’s spring. So, happy spring, folks. May the blinding memory of the winter of 2015 become ever rosier in time. Go Sawx.

Posted in General, Seasonal | Tagged , | 1 Comment