After the atomic bombs were dropped on Japanese cities, it didn’t take long for the US public, to start drawing what it would look like if atomic bombs went off over their own cities. PM, a New York City newspaper, may have inaugurated the genre with its August 7, 1945, issue, when it took what scant facts were known about Hiroshima and superimposed the data onto the Manhattan skyline:
This impulse — to see what the bomb did to others, and then to apply it to one’s own cities — worked on at least two levels. In once sense it was about making sense of the damage in intuitive terms, because maps of Hiroshima don’t make a lot of intuitive sense unless you know Hiroshima, the city. Which very few Americans would.
But it’s also a recognition that atomic bombs could possibly be dropped on the USA in the future. The atomic bomb was immediately seen as a weapon of the next war as well as the present one. It was a weapon that would, eventually, make the United States very vulnerable.
Considering how many non-atomic bombs the US dropped on Japan during the war, it’s a little interesting that nobody has spent very much time worrying about what would happen if someone firebombed the United States. Why not? Because the U.S. has never imagined that any other nation would have the kind of air superiority to pull off sustained operations like that. No, if someone was going to bomb us, it would be a one-time, brief affair.
When the US did invoke American comparisons for firebombing, it was to give a sense of scale. So the Arnold report in 1945 included this evocative diagram of Japanese cities bombed, with American cities added to give a sense of relative size:
So I was kind of interested to find that in the final, late-1945 issue of IMPACT, a US Army Air Forces magazine, contained a really quite remarkable map. They took the same data of the above map — the Japanese cities and their equivalent US cities — and projected them not on Japan, but on the continental United States.
It’s the only attempt I’ve seen to make a visualization that showed the damage of the ruinous American air campaign against Japan in such a vivid way:1
The correspondences between US and Japanese cities were chosen based on the US Census of 1940 and presumably a Japanese census from around the same period. The above map isn’t, the text emphasizes, a realistic attack scenario. Rather, it is meant to show this:
If the 69 U.S. cities on the map at right had been mattered by Jap bombers free to strike any time and anywhere in this country, you can vividly imagine the frightful impact it would have had upon our morale and war potential. Yet this is precisely what the B-29s did to Japan.
What’s remarkable is that this isn’t some kind of anti-bombing screed; it’s pro-bombing propaganda. Both of these images are bragging. The text goes on to emphasize that if someone were really targeting the US, they’d hit industrial centers like Detroit, Philadelphia, and Pittsburgh — to say nothing of Washington, DC, which is conspicuously absent and unmentioned.
IMPACT was classified “confidential” during the war, meaning it had a circulation of about 10,000 airmen. It’s a pretty wonderful read in general — it’s a vociferously pro-Air Forces rag, and is all about the importance of strategic bombing. As one might expect, it de-emphasizes the atomic bombings, in part to push back against the very public perception that we have today, where the last two major bombings are emphasized and the other 67 are forgotten. On the above maps, Hiroshima and Nagasaki are unremarkable, easily in the crowd.
I thought it would be interesting to copy out all of the data (city names, damage percentages, and look up the US Census data) and put it into an interactive visualization using a Javascript toolkit called D3. If you have a reasonably modern browser (one that supports SVG images), then check it out here:
One thing you notice quickly when putting it this way is how large some of the metropolises were versus the relatively modest of most of the other cities. The idea of someone bombing out 55% of Sacramento, or 64% of Stockton, or 96% of Chattanooga, is kind of mind-melting. Much less to consider that a New York City minus 40% of its land area would look like.2
You can also see how cramped Japan is compared to the USA (they are at the same scale in the above image, though the projections are a bit tweaked for the layout). Even that could be more emphasized, as the text does: because Japan is so mountainous, its inhabited area is only roughly the size of Montana. So it’s even smaller than it looks.
Still, for me it’s just remarkable that this mode of visualization would be used in an official publication. These guys wanted people to understand what they had done. They wanted people to know how bad it had been for Japan. They wanted credit. And I get why — I’m not naive here. They saw it as necessary for the fighting of the war. But it also shouldn’t have been surprising, or unexpected, to those at the time that people in the future might be taken aback by the scale of the burning. Even Robert McNamara, who helped plan the firebombing operations, later came to see them as disproportionate to the US aims in the war:
This sequence, from Errol Morris’s Fog of War, has been one of my favorites for a long time. But it wasn’t until recently that I realized its source was one of these maps used for postwar boasting. It’s an incredible re-appropriation, when looked at in that light. A document meant to impress an audience, now being used to horrify a different one.
- Regarding the image, I scanned it out of a reprint of the IMPACT issue. Because of the crease in the center of the pages I had to do some Photoshop wizardry to make it even — so there is a lot of cleaning up around the center of the image. The data hasn’t been changed, but some of the state outlines were retouched and things like that. Similar Photoshop wizardly was also applied to the Arnold Report image to make it look clean. I suspect that the IMPACT image may have come first and the Arnold report image was derived from it, just because the IMPACT caption goes into details about methodology whereas the Arnold report does not. [↩]
- But don’t confuse “destroyed” with casualties — I don’t have those numbers on hand, though if I can find them, I’ll add them to the visualization. The nice thing about D3 is that once you’ve got the basics set up, adding or tweaking the data is easy, since it is just read out of spreadsheet file. The maddening thing about D3 is that getting the basics set up is much harder than you might expect, because the documentation is really not aimed at beginners. If you are interested in a copy of the data, here is the file. [↩]