Posts Tagged ‘island effect

02
Oct
10

A Village Floating off the Coast

Today’s contribution comes from my friend Kate, the one with whom I was recently on a Michigan wine tour. This is a map which heads an article on the village of Cairanne and the wines which originate there:

Copyrighted by Foodtourst.com. Click to visit site.

There seems to be some sort of notion out there that every map needs a legend. That, somehow, it’s not a map anymore if it doesn’t have one. This is patently untrue. If you know your audience can easily figure out how to read your symbols, you can probably skip it (or, at least, minimize it). Legends are for imparting literacy when your think audience lacks it. They are frequently needed, but not indispensable.

The legend on this map is clearly dispensable. I cannot fathom why the name of the village was not labeled right next to the giant red dot. Instead, the artist created a legend at the bottom to explain what the giant red dot means. His or her choices suggest the following assumptions were being made about audience:

  • Readers have the skills to figure out that Paris, Dijon, etc. are at the locations of the dots found near those words.
  • Those same readers would not understand what it meant if the word “Cairanne” were similarly placed next to a big red dot in France.
  • But they will, however, know what it means if the word “Cairanne” is placed next to a big red dot outside of France.
  • Readers will know that the big red dot outside France is meant to represent the big red dot inside France.

Some of these assumptions are more questionable than others, to put it mildly. In fact, because of the nonsensical nature of assumption number two, the legend makes this map harder to read. As Kate writes, this map had her “confused for almost a minute about whether they thought Cairanne was in Spain.” Probably because she assumed that the artist would label the big red dot in France as “Cairanne” if it were Cairanne. She was confused because she didn’t think that the map artist might have considered her too dumb to figure it out without a legend.

While we are on the giant red dot, I might strongly recommend making it not so giant. Cairanne is a small village. But the dot pattern on the map gives a subtle impression that Cairanne is huge and Paris is insignificant. The artist wants Cairanne to stand out, understandably. But there are better ways to establish a visual hierarchy on this map, for example by changing the colors of the non-Cairanne cities and dots to fade a little more into the background, and making the Cairanne dot the same size as (or only slightly larger than) the non-Cairanne dots, while still keeping it red so that it pops out.

Meanwhile, making its long-awaited return to this blog, it looks like we’ve got another great example of the Island Effect going on here! Just north of France is some water, indicated in white. Just east of France is some land, indicated in white. Thus, France looks like it’s floating off in the sea, lacking any geographic context. Now, I don’t think this is always a problem — it’s perfectly fine to have a map that shows France and nothing else at times. Here, however, the author is very inconsistent in his or her treatment of geographic context. It seems senseless to show some bits of contextual information (the names of some countries) and leave off others (a little bit of land showing where those countries are). It’s also strange to mark Italy, Spain, and Belgium, while leaving Luxembourg, Germany, and Switzerland off the map. Either France’s surroundings are important, or they’re not. To my mind, it should either be an island and the sole thing on the map, or it should be shown in its full European context with all its neighbors. Going halfway just looks sloppy.

Finally, it’s worth noting that the copyright for this map is placed in a bit of an odd position. It’s between the map and the legend, very much visually in the way. I appreciate the owners of the work wanting to ensure they’re credited, but it could be put less obtrusively in the corner.

One Nice Thing: At least the artist thought to include some geographic context. I can imagine a lot of places would just throw an outline of France on the page, with a dot for Cairanne and nothing else. For people familiar with Paris, Bordeaux, etc., this map helps to give them reference points.

Not every map needs a legend. Nor does every map need several of the other common map elements, for that matter. If I scrounge up a few good examples, I may write a post to kick off my Worldwide Campaign to Eliminate Needless North Arrows, and my International Crusade Against Useless Scale Bars.

14
Aug
09

Way Over in British Columbia

Today’s effort comes to us from Kerr Wood Leidal, a consulting firm in Canada:

Detail of map - click for full poster. Taken from http://kwl.bc.ca/docs/CEBC2008-RHAM-FinalAwardofMerit.pdf.

Detail of map - click for full poster. Taken from http://kwl.bc.ca/docs/CEBC2008-RHAM-FinalAwardofMerit.pdf.

This one was brought to my attention by a reader, Eliana, who appeared particularly exasperated that this map won an award. It seems that the folks at KWL took first prize in the 2009 Map Gallery competition at GeoTec, which bills itself as the largest GIS conference in Canada. According to the GeoTec site, the winners were selected “based on overall appearance and effectiveness as maps.” So, this means they have to a) look good, and b) communicate spatial information clearly. Longtime readers may recall that one of these things is more important than the other. Though this is not to say that making your map look good is unnecessary, and in a competition like this, it’s a fair criterion for judging.

Let’s start at the legend – of the 8200 or so red spots on this map, each one encodes how much hydropower could feasibly be generated at that site. A red dot means < 1 megawatt, a red square is 1 – 10 megawatts, and a red triangle is > 10 megawatts. This is a non-orderable scheme – squares, circles, and triangles cannot be put into an inherent order. So, it doesn’t make sense to use different shapes for different numbers of megawatts — which can be put into order. Dots of different size or perhaps color brightness (but the same color hue – so, for example, a scheme from light blue to dark blue) would be more sensible here.

The other big problem with this scheme is that if you look at the map, you can’t pick out the squares from the circles from the triangles in most areas. They’re way too small to be able to tell which shape is which without staring or zooming way in, and even then it’s sometimes ambiguous. I will add the caveat that I don’t know how large this map was printed – it may be less of a problem if this thing is two or three feet across. And I’m not even sure if this map really needs to go into this much detail. It does a good job as it stands of showing the general distribution of hydropower sites, mostly clustered along the west of the province. If the authors want to add an extra level of information, about how much power might be generated at each site, their task becomes much harder, because now the dots have to be separate enough, and big enough, for people to be able to tell how they vary. And the reader cannot effectively do that, here.

Even if you could visually tell one shape from another, however, it would still be difficult to pick out the overall pattern of where the 1 megawatt sites are, and where the 10 megawatt sites are. Shape, as a visual means of encoding information, is weak in terms of what we call selectivity. It’s hard to select just one shape, and then try and find the distribution of only that shape. It’s much easier to do this sort of task with something like size – you can quickly see where the big dots are clustered and where the small ones are. A quick example:

Selectivity Diagram

Notice how easier it is to pick out the cluster of small squares near the center than it is to pick out the cluster of triangles near the center - size has better selectivity than shape.

Moving on from the dots, let’s consider a few other, lesser offenses. The labeling has poor contrast with the background, especially at Stewart and Port Hardy. Interestingly, some of the labels have been set in light-colored type, to better stand out against the water, thus demonstrating that the labeler was mindful of contrast issues. But not enough to make them legible against a mass of little red shapes.

Notice the white area in the northwest. That’s part of Alaska. It looks like it’s buried under an ice cap or something, given the color scheme and the fact that it’s flat, while the rest of the map shows terrain relief. I’ve never been to Alaska, so I suppose it could in fact bear a great resemblance to Antarctica. More of concern to me is the fact that the authors have possibly done one worse than the dreaded island effect. Instead of either showing British Columbia as an island with no surrounding land, or showing it in its geographic context, with the USA and other provinces around, the authors have chosen to include just one part of one of the surrounding areas. It looks very odd, and I think it would look better showing just the province, really.

Speaking of odd, it looks like someone has discovered the joy of the “glow” effect in the Adobe suite of graphic tools, because the entire coast of British Columbia is glowing white. Now, a glow effect can be a great addition to a map, but it would probably make more sense to do one a light blue one that looks like shallow coastal water, rather than giving the appearance that there’s been some sort of radioactive disaster off the Canadian Pacific coast.

Since this is a map about hydropower based on the flow of rivers, where are the rivers on this map? You can see a few here and there, in a very light blue, but the hydrography should really stand out more. Maybe not every single creek, but at least the major ones.

Finally, a note on the map projection. The authors appear to have kept the central meridian for this conic projection somewhere far east of the map area – say, around the center of Canada.  For those readers who may be confused by what I just said, I will avoid giving an entire lesson on map projections. Instead, here’s a somewhat related way of thinking about it: Consider your average map of Canada, grabbed randomly from the Internet – the kind with the curved appearance to it. Doesn’t it look like the authors of the map above took BC from the far west end of one of these typical Canada maps, and didn’t bother to rotate it so that it wasn’t tilted clockwise anymore? If it’s the only thing on the map, BC should be centered so that it’s northern border has a shallow “U” shape, instead of curving downward only. The projection on this map just constantly reminds me of the fact that BC is at the far west end of Canada.  Perhaps the authors wanted to keep that suggestion in my mind – “British Columbia: We’re way over here!” Might make a good provincial motto.

One Nice Thing: The terrain relief is not just a useless bit of decoration – it’s useful, because hydropower potential is affected by terrain. You can see the river valleys and everything (if not the actual rivers, unfortunately). So, the relief here is both a nice aesthetic component, and conveys information relevant to the topic at hand. A win-win.

Before I leave off, I’d like to thank all of you who have been writing in to me and submitting maps you have encountered. It’s a big help to have other eyes looking for these things. I may not end up using every submission, but I appreciate them nonetheless.

05
Aug
09

Concerning the Value of Human Life

Today’s effort is brought to my attention by one of our readers, Robin, who previously alerted me to the problems of Moon Maps. This map appears in a post on the OKTrends blog, where the authors analyze the geographic distribution of how people answer various questions on the dating site OKCupid. We’ll take one of them as an example:

human_lives_map

Obtained from http://blog.okcupid.com/index.php/2009/07/13/sweet-ass-american-trends/. Legend, title, and map combined by author into one PNG file.

Side note: the darkest red, for the highest percentage “No,” shows up as a single pixel at DC. So, I cannot claim that the map doesn’t follow the legend. Though, the DC dot is so unnoticeable it might as well not be. Putting an area on a map which is too small to be seen is a serious problem. Better for the authors to include an inset or two, showing the smaller states in a more magnified fashion (as we say in the business: at a larger scale).

There’s a fundamental error that a lot of people are prone to making with these maps (and, in fact, I didn’t notice myself doing it until I’d spent quite a while looking at these). If you look at the map above, or any of the others in the OKTrends post, there is always one state at each end of the scale. In the above example, Nevada is the brightest green, matching the far “Yes” end of the scale, and DC (thought you can scarcely see it) is the darkest red, matching the “No” end of the scale. This does not mean that 100% of the people in Nevada answered yes and 100% of those in DC said no. It just means that more people answered yes in Nevada than anywhere else. Maybe in DC 2% of the population said “yes,” and in Nevada 5% did, with all the other states in between. But this legend makes it appear as if DC uniformly said “No,” and Nevada “Yes,” with the vote being split in all other states. It is unintentionally misleading, which is the most tragic cartographic sin. The authors wanted to convey some information in good faith, but their communications became twisted and false.

The Yes-No labels on the legend are not the only problem. It’s the color scheme. It’s a diverging color scheme – the green and the red are opposite ends, and states shade toward one end or the other. There are many good reasons to use these schemes, but in this instance, it just makes it look like there are solid No states and solid Yes states, instead of states where a handful of people said yes, and those where a slightly larger handful of people said yes. The color scheme should be one hue, changing in lightness (say, from a pink to a dark red), and the map should show % yes votes (or % no) votes only. This still lets you pick out trends – Nevada has more Yes votes than other states, but it doesn’t mislead you in to a panic about the amoral citizens of Nevada and how they disregard human life. You can tell that “yes” is unpopular, but less so in certain areas.

Another reason to ditch the red-green scheme, which the authors use uniformly throughout their maps: It doesn’t always make sense when the options are other than yes and no. Here’s a map from later in their post:

Obtained from http://blog.okcupid.com/index.php/2009/07/13/sweet-ass-american-trends/. Legend, title, and map combined by author into one PNG file.

Obtained from http://blog.okcupid.com/index.php/2009/07/13/sweet-ass-american-trends/. Legend, title, and map combined by author into one PNG file.

These two colors, green and red, have certain connotations of positive and negative in much of the english-speaking world, so here, the scheme is a problem, unless you’re going with an implied value judgment where “Right to Bear Arms = Good” and “Right to Vote = Bad.” The people on the OKTrends blog don’t quite strike me as that type. Nor the type to greenlight an unequal valuation of human lives. And, of course, the legend does again make it look like Idaho is 100% behind giving up the right to vote. I’m guessing (and it is just guessing) that if you looked at the data, the vast majority of respondents in each state said they’d rather keep the vote and lose their guns. But Idaho happened to have the most people who were more fond of guns than voting (hard to hold that against them, given the political climate). Probably just a few percent. But people will come away from this map with the sentence stuck in their mind “Idaho strongly prefers to lose the right to vote.” I have belabored the point, so let’s move on.

Note that there are two states in the lower right corner which are not actually filled in with any color. Thanks for playing, Hawaii and Alaska! It’s OK — they’re just happy to be there, given how often they’re left off of US maps.

The shapes of the states are poorly generalized, and may in fact have been drawn freehand. Possibly just traced loosely, instead. But it has that feeling to it. Many of the states look slightly-to-moderately wrong. Look at my beloved Michigan – it’s become an amorphous blob. Saginaw Bay is entirely missing. It’s a sizable body of water. Like, nearly as big as some states. Also, I’m pretty sure Ohio doesn’t look quite like that. And Maryland seems to have taken over part of Virginia. Wyoming, though, is just as rectangular as ever.

I would be tempted to complain about the map projection, but since I think it’s freehanded, it doesn’t exactly have one.

Note the line around Michigan, showing the water border between the US and Canada as it passes through the Great Lakes. Except, the line only shows the border as it passes through two of the lakes. The actual border passes through two additional ones, north of Ohio, Pennsylvania, and New York – but you don’t see that section here for some reason. So, we have part of a Canadian border. And no Canada, by the by.

The map suffers from the dreaded island effect, where the US looks like it’s floating off in the ocean with no land nearby, and Canada looks like a great and forbidding sea, much like the Gulf of Mexico or the Atlantic Ocean. I might not call this a problem for this map otherwise, but putting a portion of the Canadian border in, and then not actually drawing Canada, makes this a serious issue.

Nitpicky: Both maps have a few artifacts here and there, like in Louisiana in the bottom map. A few pixels the wrong color.

One Nice Thing: It’s an unclassed choropleth, which I rather like. Most choropleth maps organize data into classes – say, one color will represent all states with yes votes between 0% and 5%, and the next color will be for states from 6% to 10%, etc. — the states are grouped. In an unclassed choropleth, such as those above, each state is given a color in exact proportion to the number of votes for one choice or the other. A state with 1.1% yes votes gets a different color from one with 1.2%, because they have different values. They’re never grouped together into the same color, as they would be in a classed choropleth. There’s a whole debate as to which is better, classed or unclassed, and I will not get into it here. I will just say that I am a fan of the unclassed choropleth, so I think it’s a Nice Thing.

Addendum: I am a bit embarrassed that I didn’t notice this at first, but these maps are very unkind to the millions of people with red-green colorblindness, as several commenters on the original OKTrends post mentioned. Here’s an approximation of what a red-green colorblind person sees when they look at one of these maps:

Run through http://colorfilter.wickline.org/ with protanopia filter

Run through http://colorfilter.wickline.org/ with protanopia filter

Designers cannot ignore such a vast population, and must take color vision anomalies into account.

04
Aug
09

True Confessions of a Trained Cartographer, Part 1: My First Map

(Editor’s note: Daniel Reynolds is a colleague of mine in the University of Wisconsin Cartography Laboratory. He has graciously agreed to act as a guest poster, publicly criticizing some of his own works for your edification and entertainment  — DH)

Okay, so the title is a lie. This wasn’t my first map. Or even my second. But this was the first map I made where I wasn’t trying to copy a lab step by step. I cringe every time I see this thing.

4850_map1

In my introductory cartography class at Utah State University, the first assignment was to do a combination choropleth and dot density layout with the given data in ArcView 3.3. Why not something else, something not quite so dated given that it was 2006? Or perhaps something with better tools to make stuff pretty? Well, the instructor had never used ArcGIS 9.1 (the then-current standard at USU) since he hadn’t been working in the geospatial field for a number of years. All that is to say that I didn’t have the best software to produce a good looking map.  My excuses stop there…

When I made this map, I didn’t know much about color theory. That is unfortunate, as I’m pretty sure I picked one of the worst possible color schemes. By that, I mean the choropleth makes me want to stab my eyes out. I’m pretty sure it was a default scheme that the software spit out. Color  is, in my opinion one of the easier ways to categorize a map as a ‘GIS map’ or a ‘Cartographic Map';  a good number of the default color schemes in the GIS packages I have used are passable, but need work. Other schemes make Cynthia Brewer cry…like the one I used. While it makes sense to use a diverging color scheme to highlight areas of population growth and population decline, I managed to abuse this concept pretty thoroughly. A diverging scheme is meant to highlight data in relation to a critical value. A good diverging scheme is made of two hues changing in saturation and/or brightness as you move away from the critical value.

diverging

Instead of following this sensible approach, I used four hues (although I think at the time I felt that orange/yellow and green/blue fit together reasonably well). This results in a color scheme that could be interpreted as categorical even though the data beg for an ordinal scheme. On top of all that, I don’t think there is enough differentiation between some of the colors (the two blues, the two greens). A little less obvious is that I really goofed up the relationship between the color scheme and the critical value. The second lowest category (0-11%) should be part of the color values above the critical break. In other words, it should be some shade of green. As it is, the critical value appears to be 11% instead of the more sensible 0%. While we’re examining those values, I should probably point out that the break values were likely determined by the software without much thought on my part.  My best guess is that I used a standard deviation classification. Why? I don’t know!

A few last parting shots at the choropleth map:

  • Not only did I fail to include surrounding landmasses to avoid the island effect, but I also managed to leave an enormous amount of empty space on the page.
  • The legend drives me nuts. Again, I didn’t effectively use the space at hand.  I think it would be much more aesthetically pleasing if it were more compact (make the legend title two lines instead of one).
  • The labels could be much smaller and less distracting. They disappear on the darker states.
  • The topic of the map is population change, but we have no frame of reference. Is this change from 1930 to 2030?

On to the dot density map:

Part of the assignment was to symbolize the population density of hurricane-impacted states on a county level. While I think this is an appropriate scale for the map in question, I had no clue at the time.  I remember the concepts of dot size versus dot value being very vague and somewhat hard to comprehend. The basic goal with a dot map is to pick a value that is easy to wrap your head around and a size that allows you to display that value effectively. I won’t go into more detail here. I managed to pick a decent value but then made the dots so small that it looks like someone tipped over a pepper shaker on my map.

My other major problem with the dot-density map is that the data set we were given is not appropriate for the given scale.  The coastlines are quite jagged which is a result of using a dataset that has been generalized (poorly in my opinion) for a smaller scale. (Yes, I do mean smaller scale).

Lastly, the island effect is avoided this time around, but the fill color ended up with a bizarre and entirely unnecessary pattern. I’m not quite sure why this happened as I don’t recall choosing this.  My best guess is that there was some image integrity loss since it is stored as a jpeg.

One Nice Thing: I had the presence of mind to generalize the legend values in the choropleth rather than using precise but essentially useless decimals.

PS.  At the time I made this map, I was really proud of the background color. I had never really used image editing software (other than doodling in MS Paint) so realizing I could draw a box, make it blue, and drop it behind everything else made me very happy. Unfortunately, the particular shade I chose makes me gag.

PPS. If you are wondering why we mapped predicted population change for the entire US with population density of hurricane impacted states, I have no clue.

(Editor’s note: To avoid leaving the Internet with the impression that Mr. Reynolds still regularly makes maps worthy of such criticism, I leave you with a link to his portfolio, where you can see his current, and much better works. — DH)

29
Jul
09

On the Canadian Adventures of David Wilkins

David Wilkins, the 21st United States Ambassador to Canada, stepped down a few months ago with the election of Barack Obama as President. While both countries eagerly await his replacement at Rockcliffe Park, let’s have a look at this handy map which shows where Mr. Wilkins traveled during his nearly three years in the country:

Obtained from http://ottawa.usembassy.gov/. No author listed.

Obtained from http://ottawa.usembassy.gov. No author listed.

This map can be found on the official website of the US Embassy in Canada, by going to the page for the Ambassador.

The obvious fault, first: Putting big red dots on top of labels tends to defeat the purpose of having those labels. A lot of the things I talk about on this site might not be immediately obvious to a lot of people making maps – good color choices, projections, etc. But it’s hard to imagine how the problematic nature of obliterating text escaped the notice of the vigilant staff of the US Embassy. Maybe they’re trying to subtly insult the people of Saskatchewan or Labrador. I’m sorry, I mean “Lab  dor.” Misread the map for a second.

Speaking of Labrador – it’s not a separate province, last I checked. Not sure why it’s labeled separately from Newfoundland, since all the other labels on this map are provinces. If they’re going to label major physical features, why not label Baffin Island while they’re at it?

The provinces are filled in using different colors. This is a perfectly reasonable idea, to help tell them apart. But, for some reason, British Columbia (BC), Nova Scotia (NS), Prince Edward Island (PE), and Newfound and Labrador (NL / Labrador) are all the same color, whereas the other nine provinces and territories are given different colors from each other. Perhaps the maritime provinces on the east coast have joined forces with British Columbia for some nefarious purpose, and the US Ambassador is secretly trying to alert the world without exposing the fact that he knows.

The dots are different sizes. Insofar as I can tell, this is only so that they fit better – there are many small ones in New Brunswick (NB), for example, to avoid overlaps. But, then there are still small overlaps here and there, anyway, such as in Manitoba (MB). It would be better if the dots were of uniform size – else, it implies that certain places are more important than others, or that the Ambassador visited some places more than others. Smaller would be better – since the ones in NB are quite legible, and there would be minimal overlap.

Me being unreasonably nitpicky: The abbreviations are not all standard two-letter Canadian postal abbreviations. They could be using their own system, but I’m not sure why they would. NWT should be NT, NF should be NL, etc.

Finally, it’s arguable whether or not the island effect is a problem here. Personally, I think it would look better if Canada was placed in a geographic context – especially one that shows the US, since this is a page for the US embassy. Emphasizing that relationship makes sense here, and it seems a missed opportunity.

One Nice Thing: The linework is well-generalized. It’s not overly detailed, but has a clean simplicity appropriate to the purpose.

30
Jun
09

The Cruelty of War

Another Wikipedia map (don’t worry, they won’t all be from there), showing battles of the American Civil War, based on National Park Service data:

Amazingly enough, it turns out that the red-coded counties, coded for “Eastern Theater,” are, in fact all in the east! The color coding has no particular use here, except to show you the difference between the NPS-defined theaters of the war. Why not just draw a border around the zones, so that you don’t need a confusing rainbow of colors to tell you roughly nothing that isn’t apparent by looking at where the filled-in counties are?

Also, why is “Mult. Years” the darkest color? This map’s color scheme suggests that “darker is later” – 1863 is darker than 1861. Making a county that saw battles in multiple years (about half of the ones on the map) even darker than the color for 1865 makes it seem like they were fighting the war there well into the 1870s. A media conspiracy has kept it secret.

Quick tip: Let’s say you’re making a map of something that happened, say, 140+ years ago. Using modern county and state borders might be ill-advised.

It looks like the US is tilted backwards. Maybe it’s a commentary by the author – “Look! The US is falling over…a house divided cannot stand!” Actually, what probably happened is they used a sinusoidal projection, which is good for showing the whole world at times, but not so good for showing one country at high latitudes.

One Nice Thing: The color scheme for the years within each theater makes some level of sense – the colors are arranged in a light-to-dark pattern as the years go on. Excepting, of course, the color for “Mult. Years.” Good cartographic sense.

30
Jun
09

What Have I Done?

It seems only fair for me to start this blog out with one of my own works, which still graces the Wikipedia page on the Kalamazoo River:

Kalamazoo River Map

Yeah.

The labeling is troubling, and the whole thing has this “I threw it together in MS Paint” look (which, actually, I didn’t), but the real problem is the inset in the lower left.  See, there’s a problem on a lot of maps called the island effect, where you just show one state or country in your map, and completely leave off the other geographic context. Look at the area to the immediate west of Michigan. It’s brown. That’s where Lake Michigan is. Also, Wisconsin. Apparently, Wisconsin (not to mention other bordering states) blends seamlessly and stealthily into the Great Lakes. So does Canada – they have, in fact, hidden their entire country in camouflage in preparation for the big invasion.

Also, in the inset, Michigan is surrounded by the same brown used for land on the main map…making it look more like a hole in the middle of the land rather than an island. And like there’s no water around it at all, only barren wasteland.

To be nice (to myself), I did make this some time before I actually had any cartographic training, or even access to the proper tools. I’ve been planning for years to replace this with a better effort. Someday, perhaps that will happen.

One Nice Thing: The city label size varies by city population. Though, you know, it would be nice if I indicated that in some way.




Follow

Get every new post delivered to your Inbox.

Join 71 other followers