NASA Creates Insanely High-Res Map Of America's Trees, And Offers A Lesson In Information Design

We’ll start with a cool picture, but stick around. Things get really interesting.

According to Robert Simmon, art director of NASA’s Earth Observatory, the best science communication goes in layers. You start out with something striking and relatable, and then you invite your audience to go deeper, revealing more and more with each new layer. For example: This satellite map of America’s forests, built from 265 million segments and intended to be a tool for understanding exactly how much carbon our trees can absorb. It’s our infographic of the day, but it’s also a story about the subtle choices that go into good visual information design.

Click to zoom.

Layer 1: This is a map of where the trees are in the lower 48 states. The data is incredibly high resolution, accurate to within 30 meters, a scale which means that forest managers and researchers can track the disruption caused by things as small as a parking lot or a large building. This stuff is important because forests are one of the largest carbon sinks in the world. Managing our carbon stock is going to be a key part of fighting global warming.

Layer 2: As you zoom in close, you can see fascinating patterns. In some places, the trees almost look like a topographic map, as they trace the patterns of a mountain range. In other places, checkerboard patterns reveal the impact of human clear cutting. Even though this is a map that seems to ignore people, the impact of humanity can still be seen.

Layer 3: The map was created based on data from the Woods Hole Research Center’s (WHRC) National Biomass and Carbon Dataset (NBCD). It’s a baseline estimate of "basal area-weighted canopy height, aboveground live dry biomass, and standing carbon stock" (in other words, trees) for the year 2000. It combines data from a variety of sources, most notably from satellites which scanned the ground. This reveals the lasting importance of the space program: It allows us to look more closely at our own planet. Earth Observatory published the map as part of a feature story on mapping and understanding the locations of the planet’s trees. Earth Observatory’s mission is to "share with the public the images, stories, and discoveries about climate and the environment that emerge from NASA research."

Layer 4: Though WHRC had released its own map of the data, Simmon wanted to make a new one. "My role is to make this imagery easy to interpret for non-experts," he says. He wants you to be able to quickly look at an image and get it. ("Hey, a map of trees!") But the more you look at it, the more you can learn. With this criteria in mind, there were problems with the WHRC map. "There are two things about it that I don’t really like," he says.

The first has to do with coloring. Both maps go from beige to increasingly intense green as density goes up. But with the WHRC map, the highest densities are represented as red. Red is not a color we normally associate with trees. This is, to say the least, not very intuitive. "Scientists do this all the time," he says. "They do not understand how vitriolic I get about it."

Even worse, the switch from green to red is abrupt and doesn’t accurately represent the degree of change. "The objective with a map is not to get specific values out of it. What maps are really powerful at is showing patterns and relationships." You want to have the visual relationship match the numerical relationship. You want the shift from 10 to 20 to look as big or as small as the shift from 50 to 60. Otherwise, you risk overemphasizing certain areas or falsely implying patterns where none exist.

The second problem comes in the areas where there just aren’t all that many trees. In the WHRC map, they fill those areas of no data with a topographic map. "It’s distracting," he says. "They throw out the topographic data in regions with actual data, so you can’t compare it." It ends up being a background texture. Similarly, the offshore water is a dark blue and Canada is white, which pulls the ocean into the foreground, complicating our interpretation of the data. "To overcomplicate maps is a mistake," he says, "Only show the one parameter and remove any potential distractions. You can throw all that stuff out without any negative repercussions at all."

Layer 5: For all the problems with the WHRC map, Simmon says it’s actually one of the better examples of scientific visual design. "Their palette is vastly superior to what most of what NASA and other research institutions put out."

Click to zoom.

Simmon sees a community beset by problems owing to legacies of early software. "The scientists are pressing buttons in their software and getting the results," he says, while the software was made by people "doing things to optimize it to be as fast as possible or work on an 8-bit monitor and the code was just never updated." As a result, we get a visual tradition based on computational expediency which becomes what the scientific community is accustomed to seeing.

Simmon says that there is a great deal to be learned by looking at the work of cartographers. "A lot of these presentation problems that scientists struggle with all the time were solved by mapmakers 30, 40, 50, 60 years ago or longer," he says, "In a lot of cases I’m just applying other people’s older techniques to these computer data sets. There’s a lot of low-hanging fruit that could be improved."

Layer 6: Good information design for the public is about clarity and impact. Simmon says that his primary goal as an information designer is to create "almost an emotional reaction with people." He wants to make that first "getting what it is" as easy as possible while retaining as much information as possible in the display so the more you look at it, the more you see. "It’s a classical design sense of creating a hierarchy of information."

[You can follow Robert Simmon on Twitter at @rsimmon.]

Add New Comment

11 Comments

  • Steve_Wolf

    As a graphic designer, I particularly liked (and appreciate) the layer 5 discussion about making sure the representation conventions you show show the data intuitively. It drives me nuts when I see, for example, graphs in newspapers that don't show value axes all the way to zero or, worse, present several graphs with an article that APPEAR comparable until you look to see that the grid lines in different ones represent different units. Good designers pay as much attention to content as they do to form, imho.

    Footnote: Robert Simmon, puh-leeze learn the different between "criteria" (plural) and "criterion." In layer 4, your sentence should read "With that CRITERION in mind..."

  • gregorylent

    man, look at the clear cuts in washington, and considering that the entire country has been logged, some places multiple times .... yikes

  • Alex Lingeman

    Impressive project. I think the legenda also plays a vital role in understanding the biomass. My first presumption was that the trees in the northwest could be of a pine-like species, and the colour also related to area-specific trees.

  • WHEA7

    Could you provide a link to the full-res image, if it exists?  I'd love to have it as an example picture for an interface I'm working on.  Thanks!

  • qianniciello

    "This unbelievabLE satellite map of America’s forests.." But otherwise, this is awesome! The patterns you can find are truly amazing.