In response to Map of the Ancient Mississippi a few of you left comments to the effect of “no one should live in floodplains.” It’s an appealing notion, but I think it’s unrealistic. Anne Jefferson of Highly Allochthonous published an epic post yesterday outlining the benefits and risks of building in a flood plain, and the need to balance them: Levees and the illusion of flood control. It’s well worth a read. Plus, she dug up some great archival imagery, like this map of the 1927 flood:
Some additional details on the map are in this article from the National Archives: The 200th Anniversary of the Survey of the Coast.
Here’s a sneak peek at the site-wide redesign we’ll be launching next week. Let us know what you think (not that we can change anything at this point, but we’ll keep any critiques in mind for version 3.1). [Design primarily by Cuban Council, implementation by Paul Przyborski & Kevin Ward.]
Sticking to the flood theme, here’s a recent map from the U.S. Army Corps of Engineers showing the predicted travel time for water in the Morganza Floodway.
It’s a reasonably good map, with one big flaw: the colors are more appropriate for categorical data (such as a geological map of different rock types) rather than sequential data (like elapsed time). There’s no natural progression from one color to the next, so to work out the ordering one has to rely on the position of adjacent bands of color, or look repeatedly between the map and the key. A palette that varied from light to dark, dull to saturated, or both, would be easier to read at a glance. Like this:
Since I’m still not ready to do a long series of posts on colors, here’s a handout I wrote for the 2008 Access Data Workshop that covers more of the basics: Use of Color in Data Visualization. (PDF) If you’re really interested in the topic just skip what I wrote and go straight to the references:
While poking around looking for imagery of the 1973 flooding on the Mississippi, I discovered some fascinating, very high resolution aerial photography. Images of Morgan City—a community near the mouth of the Atchafalaya River that was damaged by flooding in 1973, and will likely be hit by rising water in the next few weeks. Both are false color, but the May image is either a poor scan or the originals have degraded. These photographs, and thousands like them, are available from Earth Explorer, the online USGS archive of satellite data and aerial photography. John Mcphee’s essay “Atchafalaya” from his book The Control of Nature beautifully describes Morgan City and the 1973 flood.
Historic flooding along the Mississippi River gives me an excuse to show another of my favorite maps, from Geological Investigation of the Alluvial Valley of the Lower Mississippi River.
Each color represents an old channel, dating back 1,000 years or so. Those that correspond to historical records are dated, while older channels are ordered according to the principle of superposition (newer sediments on top of older ones). Many of the surface layers are still visible in satellite imagery. High-resolution, geolocated maps from Cairo to the Gulf, along with major tributaries, geological maps, and cross sections, are distributed by the U.S. Army Corps of Engineers.
Since I was (slightly) critical of the New York Times graphics department yesterday, I’ll show a Times map that is one of my favorite visualizations today: a map of shaking intensity and slip during the 1906 San Francisco Earthquake.
Aside from being elegant, it’s data-rich without being cluttered, and multivariate. It shows shaking intensity (in two dimensions), fault location (in two dimensions), slip along the San Andreas (one dimension), urban locations (two dimensions), and has three layers of direct labeling (cities, faults, and slip). The key is adjacent to the data, and written in clear language. Compare the map from the Times to the equivalent from the USGS:
The New York Times version both has more data and is more readable. Why? First off, the Times uses a subdued, discrete, color palette that shifts through a small range of hues. This allows other information to be layered into the map. The USGS, in contrast, uses a rainbow of hues at maximum saturation. The rainbow palette [at some point I’m going to do series of posts on color palettes, but for now i’ll just point to Color Brewer to explain the importance of appropriate color maps in data visualization] uses the full spectrum and is so saturated that there are very few colors left over for additional data—notice how the red fault lines on the USGS map disappear in areas of intense shaking. The dark ocean in the USGS version also competes visually with the data-rich areas of the map.
The New York Times also forgoes the use of terrain data in their map. It adds a lot of great context to the image (the San Andreas creates a distinct valley), but at web resolutions shaded relief is very difficult to pull off. The high frequency contrast distracts from the other data in the map. At larger sizes, or printed at high resolution, the terrain could be added back in (but it still needs to be treated carefully—see Relief Shading for advice). As a result, fine details in the faults and the boundaries of urban areas remain legible.
My favorite part of the Times map is the inclusion of the amount of slip at various locations on the San Andreas: it’s so much more informative than merely indicating the epicenter. Earthquakes—especially big earthquakes—aren’t points. They occur over a two dimensional area (reduced to a one dimensional quantity on this map). Increased slip correlates with increased shaking, which tells a story. That’s what makes this map so powerful. It’s a narrative of cause and effect, which is a quality of the best data visualizations.
Last week, Jessica Ball of the American Geophysical Union’s Magma Cum Laude blog pointed out this map of natural hazards in the U.S., published by the New York Times:
The map has many of the design virtues common in graphics from the Times (clean, focused on the data, clearly labeled, small multiples), but when I first viewed it I had trouble parsing the data. It took me a while to figure out that hazard was indicated by color, and the size of each circle denoted ancilliary information (population).
It turns out that our eyes & brains perceive area more strongly than color. Here’s a list of ways to visually compare quantities (described by Bang Wong in his Points of View column for Nature Methods), in order from easiest to hardest:
I suspect I’m not the only person who would consider the size of the circles more important than the color, and assume size was correlated with hazard. I had to carefully read the key to figure out the proper way to read the map. Perhaps using size to encode hazard, and color (or opacity, or icons) for population would have worked better.
I recently had the opportunity to attend & give a presentation at the 2011 International Symposium on Remote Sensing of the Environment in lovely Sydney, Australia. (OK, not so recently—the conference ended on the April 15th. I blame jetlag.) Just over 60 people turned up for the talk, which was mostly about our visualization of Eyjafjallajökull. (Which means I had to attempt to learn how to pronounce “Eyjafjallajökull.” Luckily I don’t think anybody made it from Iceland to Sydney to critique me.) If you’re interested, I’ve posted Keynote and PDF copies (originally prepared for last years’s fall AGU).
During sessions many of the talks were a bit on the technical side for me, so I ended up preparing material for NASA’s portable “hyperwall”—9 HDTV screens linked together. It’s useful for either high-res animations, or small multiples. With relatively thick bezels on the monitors, simultaneous display of 9 images worked particularly well:
Compare the small multiples on the hyperwall with the sequential display of the same images from the Earth Observatory’s World of Change.
During my talk I promised the audience that I’d update the blog more frequently, so there should be more posts in the future. I sometimes struggle for good ideas, so if you have a suggestion, please drop a note in the comments.