AFSWP (Armed Forces Special Weapons Project — became DASA, IIRC) and AFOAT-1 often worked together, with AFSWP being credited for AFOAT-1′s work as a means to provide cover to it.

NRDL (Navy Radiation Defense Laboratory) did a lot of work on fallout shelters, but also other defensive measures against radiation like measuring instruments, etc. The founder of the Program in Arms Control, Disarmament and International Security here at the University of Illinois Urbana-Champaign, Arthur Chilton, served at NRDL while in the Navy. Since it’s one of my few published works so far, here’s a link to the ACDIS history:
http://acdis.illinois.edu/aboutacdis/history.html

RAND did lots of work for the Air Force, including modeling of fallout dispersal.

My take is that the different plume maps were produced from data generated by different sampling and monitoring methods. AFSWP’s map may be from Air Weather Service sampling aircraft operating in support of and/or training under the direction of AFOAT-1. NRDL’s data may have come from surface vessels and — possibly — from drones not involved in direct sampling support of the test task force, which was usually done by the Air Force. The RAND map may be from a model and based on the other data collected. Pretty much just speculation on my part in the absence of more context.

One thing’s for sure despite the variance. It was bad day for fallout.

Side-note: It’s pretty clear that the BRAVO map was declassified almost exclusively because of the Lucky Dragon incident. (I wonder if it would have been if it had just been Marshallese who were exposed.) However, it’s a very generalized map — and it’s not even clear that it’s strictly accurate.

Here are three different maps of the BRAVO fallout produced by different sources: AFSWP, NRDL, and RAND. They’re each pretty different — the AFSWP one looks the most like the declassified map (though it is more specific). The NRDL one shows a much “fatter” distribution. The RAND one shows a significant “hot spot” of radioactivity further out from the blast, which significantly complicates any assumptions about the fallout being linearly distributed over distance.

Which of these is correct? I don’t really know — but it’s interesting to me that there were so many, and how different they were from the “official” one that was declassified.

In 1954, fallout was:
1. Officially, nothing more than a “nuisance.”*
2. A highly classified source of intelligence.

* A 1950 atomic weapon effects card issued to military personnel described fallout as a “nuisance.”

CASTLE BRAVO was as much a compromise of an intelligence method as it was an environmental disaster. And anything that could spread 7,000 square miles of lethal to near-lethal fallout is way more than a “nuisance.” It’s a little hard to get your mind around how secret most information was about fallout was then. There was still very little actual data available on it even as people began clamoring for fallout shelters in the late 50s as the realization of what it took to defeat it sank in.

The dispersion maps are deceiving though, as they only show one side of fallout’s life cycle, its dispersion. True enough, that’s what you’re worried about if you see a mushroom cloud upwind from your location. It’s what happens next to fallout (or radiation escaping from Fukushima) that such maps fail to show — the re-concentration of the various isotopic constituents of fallout by a variety of biological and environmental means.

Think about it for a moment. Dispersion is good, right? If it continued to infinity, even better, especially assuming that graphical representation meant this stuff was really being diluted and spread evenly throughout the environment…

But it’s not.

Plutonium is pretty nasty stuff. Inhale a tiny chunk and you may have sealed your fate. The models do a good job of roughly mapping the dispersion side of things, but they really don’t account for that tiny bit of Pu, which is concentrated and will stay concentrated at that point enough to do potentially lethal damage to your body. This was referred to as the “hot-point problem” and is a relatively unexplored issue that is one of the unfinished bits of science that might tell us more about the threats posed by radiation.

Moving on to my main point here, we need more maps and graphics that illustrate the various ways in which fallout and other radiation are re-concentrated once dispersed. Japan was very lucky that most of the plume from Fukushima headed out to sea, with much of it falling over the ocean where the currents further dispersed it. But eventually even the ocean begins to re-concentrate radiation, as the various and sundry food chains inevitably do.

On land, the most telling example is that of I-131. With a short half-life, it was considered relatively inconsequential in the scope of fallout risks, until the 1957 Windscale reactor accident demonstrated the ways in which it was quickly re-concentrated. I-131 is deposited on grass at the last stage of its dispersion. Cows eat grass, concentrate I-131 in their milk. Babies drink milk, where I-131 is concentrated in their thyroid at clearly dangerous levels.

Guess what? Until this process was understood, measuring I-131 in the environment showed it was usually present at levels that were considered far too low to be of concern. But this was obviously misleading, as we now know. So are dispersion maps in the absence of their connection to the environmental processes that inevitably reverse the process of dispersion.

I hope the post didn’t come off as alarmist — I try to be realistic about these things.