Laura Bliss is CityLab’s West Coast bureau chief. She also writes MapLab, a biweekly newsletter about maps (subscribe here). Her work has appeared in The New York Times, The Atlantic, Sierra, GOOD, Los Angeles, and elsewhere, including in the book The Future of Transportation.
Mark Monmonier, the author of How to Lie With Maps, has seen a lot of misleading and deceptive maps. But Trump’s doctored Dorian forecast is a new one.
People take maps at face value because maps generally work. That is despite the fact that maps lie, even good ones. These lies, however, are usually ones of omission and generalization—if maps presented the whole truth of the landscapes they represent, they’d be difficult to use.
But some maps lie for other, less benign reasons, like a certain hurricane forecasting chart that is now fated to go down as one of the most boldly mendacious maps in history.
On September 4, President Trump gave a news conference in which he presented a map of Hurricane Dorian’s projected path along the southeastern seaboard, as projected at the time. But the National Oceanic and Atmospheric Administration map he displayed was doctored, and not well: A black-ink addition had partially extended the cone of uncertainty into the state of Alabama. The Sharpie-d augmentation underscored the several times in recent days Trump had claimed that Alabama stood in the storm’s path, even though by the time of the news conference, no scientific models showed that the state was in danger.
What began as a particularly pointless tempest in the ever-churning teapot that is the Trump era turned into an ongoing administrative drama, with Secretary of Commerce Wilbur Ross reportedly threatening to fire National Weather Service officials after the agency’s Birmingham office correctly contradicted the president by reassuring residents that the state wasn’t going to be seeing any effects from the storm. Then came a highly unusual unsigned proclamation by NOAA that lent credibility to the president’s un-truths and disavowed the NWS’s scientific claims.
Incredibly, this story just keeps going, even as the death toll from the place Hurricane Dorian did hit, catastrophically, rises.
The fact that #Sharpiegate has outlived the storm at its center is in part a reflection of the special power that maps have. Mark Monmonier, the author of How to Lie With Maps (now in its third edition), writes about how American students are rarely taught to be “cautious consumers” of maps, in the way that they’re trained to analyze the credibility of someone’s words. For Monmonier, a distinguished professor of geography at the Maxwell School at Syracuse University and a historian of 20th century cartography, Trump’s map manipulation stands out as the most flagrant example of deceptive cartography that he can think of. And it may be a warning sign of even grosser misrepresentations of facts to come.
We spoke with him on Tuesday morning. The conversation has been condensed and edited for clarity.
You’ve spent your career writing about how mapmakers throw out extraneous detail and make certain subjective choices in order to produce effective maps. Obviously there can be misuses and abuses of this power. Can you think of some more-normal examples of how maps are misleading?
Generally speaking, people trust mapmakers to give them the most reliable representation, and there’s not usually an inherent bias. But there can be sloppiness. One example I see often in digital maps is a failure to recognize inherently meaningful breaks in the values that go into them.
If you’re working with numerical data and have to classify it in five, six, or seven categories, where you position these cut points can make a big difference. You can set cut-points that guarantee numerically compact categories, but which are not particularly meaningful to the map reader. For example, with population change data, there are inherently meaningful breaks: One is zero. Another is the national average, which can help show if places are above or below the average. There are lots of people making maps who use an algorithm to provide them with compact, allegedly accurate categories, but they ignore these, and the maps can end up distorted.
Weather forecasting maps are a special form of cartography with a lot of built-in room for error, right?
When you get into weather forecast maps and models, you’re trying to take a series of assumptions about the atmosphere, the movement of warm and cold air, and moisture. You simulate their movements over time and make certain assumptions.
Meteorologists have now been using modeling where rather than rely on two or three different iconic models—like the American or European model—they instead inject some random errors and run the models to see how stable the results are. If you get basically the same result, you might be reasonably certain that the predictions you’re generating are pretty reliable. And if they’re widely different, you’re dealing with a lot of uncertainty and different quirks in the atmosphere that might make the outcome not as predictable as you might like. When you’re looking forward as many as seven days to track a storm, there is a lot more uncertainty when the storm is farther out—hence this term you see, the “cone of uncertainty.” When it gets closer, the range of options gets smaller.
And in this uncertainty is where Trump’s altered map of Dorian’s path probably originates.
So, in the case of the Trump map, there was a point of origin out over the Atlantic which was the then-current location of storm, and you saw various dots along a projected storm path as the storm would approach the coast. As you got closer and closer, there was a cone of uncertainty that became wider and wider.
That map that Trump was referring to in the news conference was for an earlier forecast, and showed the storm approaching the coast, not going beyond it. The circle seemed to go onto Northern Florida. Somehow he got it in his mind that it looks like it’s got a momentum like a baseball and kept going and going, because he might not understand a whole lot about storms, or about the limits of forecast maps. He had a fixed sense of concern about Alabama, maybe because this is a state that’s given him substantial victories. With this fixation in mind, he thought that he might step in front of things and warn the people of Alabama that there was a storm coming. He was wrong. This guy does not like to be wrong.
So what he did was take a pen and simply draw this additional line on there to show the storm hitting Alabama. By the way, if you look closely the arc that he drew, it doesn’t even connect with the [Alabama] coastline. Then the forecasters in the Birmingham office warned people that they’re not in any danger of being hit, which put Trump out of joint. Then political appointees in NOAA issued guidance to National Weather Service employees. More appalling, Wilbur Ross apparently wanted to fire someone at the National Weather Service, and we’re left with what people refer to as “Sharpiegate.”
Are you aware of any examples from history that come close to this level of overt manipulation in a government map?
Usually, attempts to falsify tend to happen before maps are published, and don’t try to contradict established scientific facts. You can put a spin on something by influencing the appearance of a map before it’s published. You can put a spin on things by determining what is and is not going to be mapped. Something that might put your administration in an unfavorable view, for example: Those maps won’t be part of the plan.
One example [that contrasts this pattern] was a photograph published in 1940 in a French magazine called L’illustration, which basically shows a photograph of some diplomats sitting in front of a map of Europe. Apparently the map didn’t have any boundaries on it originally, so [the magazine editors] penciled some in. This created a problem because it did not show Germany’s recent conquests. It seemed to imply that there was a military reality that was being denied by the French.
But the Trump map is unusual. I cannot find anything truly comparable. We had a map that was already out there that he actually mutilated, and in a very obvious way. This guy shows absolutely no subtlety at all. And then people try to make excuses for him. I have never seen anything like this.
It’s probably indicative of the special respect for maps, that he thought that his alteration of this map would take precedence over the facts that were already printed on it. That’s an unprecedented arrogance that I’ve never seen anything like. This was something where there was a forecast put out and well vetted, using well established standards and with an appropriate use of atmospheric predictions. And this guy decides to alter it.
Trump’s action is certainly not surprising. It’s sort of fully consistent with his behavior. What’s shocking, in a sense, is that some bureaucrats who should know better jumped in. That has ominous signs in terms of other maps coming out from the scientific community and intelligence community. We need to be ever more vigilant of the way in which maps might be altered, disrespected, or suppressed. There might be maps that you don’t see—maps that never get out there.