Japan Fights On After Nagasaki: Or, What Do You Do With Your Nukes?

By April of 46 that leaves you with 45 weapons at least. I’d pursue the invasion with planed tactical use of nuclear weapons, both at the beaches as well as in major operations where Japanese units would be forced into heavy action. Armed forces would be transformed into feelers, prodding and pushing as far as they can without meeting heavy resistance. Once such resistance is noted nuke it until it’s neutralized then move on leaving civilians and such to military police and provisional administration. The Japanese would probably deploy biochemical weapons against US troops so my tactic of dispersal and “feelers” would minimize losses on allied side.
The source I had wasn't so precise about 1946 production levels, but I think 10-12 bombs per month for the first months of 1946 would not be unreasonable. If you husband all the post August production, it could well be over 50 bombs (though I think it very unlikely that that would happen).

But even allowing for some limited use of bombs for occasional attacks on cities, or for use in MAJESTIC (assuming it actually does happen), it's not impossible that the U.S. could have a few dozen bombs to spring all at once, or in rapid succession, for CORONET come spring. I think The Red did that on a more limited scale in his timeline. Now, you would also have to ramp up conversion of B-29's to Silverplates if you want to do that - not too hard to do, but it would have to be put in train at least a few months before. I think you'd also need to train more pilots for the job, too.

I have to think being hammered by dozens of 20kt range fission bombs on the Kanto Plain within the space of several days would be pretty demoralizing even to the most fanatical army hardliner.
 

nbcman

Donor
Yeah, I can't recall what Red had for the casualty estimate, but it wasn't any worse than 1 in 5 or so, if memory serves. Horrible, terrible, crippling, sure, but not virtual annihilation.
I don't think that the Author gave an estimate of Japanese casualties:

"I might do a rough calculation but as I've said in the past I'm sceptical as to whether these types of statistics really benefit the TL. "


I didn't see any overall Japanese casualty estimates in my Kindle version.

EDIT: Clarified that the estimates were for Japanese casualties.
 
Last edited:
Yeah, I can't recall what Red had for the casualty estimate, but it wasn't any worse than 1 in 5 or so, if memory serves. Horrible, terrible, crippling, sure, but not virtual annihilation.
Here is a quote from the TL about Japan:
The death toll from starvation at this point had now eclipsed military and civilian deaths related to the war at over ten million
 
Here is a quote from the TL about Japan:
Thanks. So that's about 15% or so.

EDIT: If that's in ADDITION to combat and bombing deaths, then maybe it's closer to 20%, total.

Which, depressingly, is not implausible for a war that drags deep into spring 1946 with MAJESTIC + CORONET + continued atomic bombing and Operation STARVATION.
 
Last edited:
You could have pockets of resistance all over Japan, East Asia and the Pacific holding out into 1947 or even later, because there's no one whose authority they recognize to tell them to lay down their arms.
As from the holdouts on bypassed islands for decades
 
Yeah. Just multiply that by ten. Or a lot more.

It would be very messy.
I’ve always wondered how long it would take and how bloody it would be to defeat the Japanese forces not located in the Home Islands (3.5 million troops in Asia and the Pacific) if Japan didn’t surrender in August.

Here is a breakdown from BobTheBarbarian on the strength of Japanese forces by location in August 1945:
Bonin Islands: 23,600 (15,000 Army and 8,600 Navy)

Okinawa (Ryukyu) Islands: 52,100 (40,900 Army and 11,200 Navy)

Taiwan (Formosa): 190,500 (128,100 Army and 62,400 Navy)

Korea: 335,900 (294,200 Army and 41,700 Navy)

Sakhalin and the Kurile Islands: 91,000 (88,000 Army and 3,000 Navy)

Manchuria: 665,500 (664,000 Army and 1,500 Navy)

China Proper: 1,124,900 (1,055,700 Army and 62,900 Navy)

Central Pacific Ocean Islands: 106,900 (48,600 Army and 58,300 Navy)

Philippines: 127,200 (97,300 Army and 29,900 Navy)

French Indochina: 98,200 (90,400 Army and 7,800 Navy)

Thailand: 107,500 (106,000 Army and 1,500 Navy)

Burma and India: 71,500 (70,400 Army and 1,100 Navy)

Malaya and Singapore: 134,700 (84,800 Army and 49,900 Navy)

New Guinea: 33,800 (30,200 Army and 3,600 Navy)

Other: 390,300 (286,100 Army and 104,200 Navy)
How many Allied casualties would it cost to defeat all these forces (while there’s intense fighting in Japan proper)?
 
Last edited:
I’ve always wondered how long it would take and how bloody it would be to defeat the Japanese forces not located in the Home Islands (3.5 million troops in Asia and the Pacific) if Japan didn’t surrender in August.
The first question is whether it's still Hirohito who is making the radio broadcast to make them lay down their arms.

If it is, I think you'd still get the great majority to do so, even if it's in November or February, say.

If Hirohito is dead, and the throne has passed to another member of the dynasty, there will be a certain percantage of commands, I suspect, who would refuse it, because they will doubt it's legitimate.

If, as in Decisive Darkness (the timeline by The Red), there is no longer a recognized central authority, but rather a number of competing ones, then it's going to take a while to get them all to give up.

If we're talking spring 1946, though, a lot of what's on Bob's list will be wiped out, or whittled to a remnant. The choice for Allied militaries will be which ones are worth combat to mop up, and which ones might be best left to starve out. The ones in the hills of Luzon and Malaya will have to be dealt with forcibly because of the danger to large civilian populations, but there is less urgency with the garrison on, say, Rota.

It will be messy, though. It would be a bleaker postwar world than the one we got.
 
Start exterminating Japanese cities and military installations. Starting with those that were the most important to logistics
 
It's also worth keeping in mind that whatever else is going on, mass conventional bombing of Japan is going to continue. On August 13, in fact, LeMay hit Japan with a big 1,000+ bomber raid. To give an extra nudge to the peace faction.

Meanwhile, by autumn, LeMay's XXI Bomber Command would be joined by the Commonwealth Tiger Force, a sizable (initially, 22 squadron) contingent of RAF Bomber Command, to be operating out of bases in Okinawa.
 
Two more books just got added to my reading list. Frank and Giangreco. They should be available from my local library or via interlibrary loan.
 
Two more books just got added to my reading list. Frank and Giangreco. They should be available from my local library or via interlibrary loan.
Speaking Giangreco, I just came across a passage (p. 201-03) I'd forgotten about: Marshall did a lot more than just ask Groves about tactical use of nuclear weapons. He had his staff drawing up a detailed plan for using both atomic weapons and poison gas on X-Day - and he was preparing to advocate for it aggressively to Truman, to the point of urging that all further attacks on Japanese cities cease immediately, so that a stockpile could be built up for X-Day:

Giangreco p202.png


I would really love to see the planning documents Marshall was recalling to Pogue.

Further up, Giangreco notes the reasoning Marshall was preparing to use on Truman: "We'd dropped two bombs in rapid succession. But it didn't do the trick. The shock value is gone now. Bombing another city every week is not going to shock them more." And then he could play on Truman's anguish, expressed in that letter to Senator Richard Russell, about killing "all those kids." So, we use the bombs on invasion targets instead. Ironically, as Giangreco notes, this almost certainly would have ended up killing more people - many of them Americans, thanks to radiation.

This raises two questions for me, one minor and one major. The minor one is that a strategy employing nine atomic bombs on MAJESTIC would have been more than the projections I quoted for bomb production by the end of October. Even if you save the Kokura bomb, that only gets you seven. Perhaps Hanford could have accelerated the rate slightly and it's not showing up in the sources I've seen.

The major one is the looming debate (on top of the others!) on atomic weapon use within the administration that appears to have been starting to gel in early-mid August. Marshall was preparing to push for all bombs to be reserved for MAJESTIC. But it appears to me that Spaatz and Groves would prefer to keep hitting cities. Who would win the debate? Would Truman split the difference? Giangreco doesn't quite say, but the vibe I get is that the deck was stacked in Marshall's favor.

Meanwhile, in the 1950's, MAJESTIC veterans would be wondering about all the strange cancers that their war buddies were coming down with...
 
I think @The Red 's timeline, Decisive Darkness, remains the most compelling timeline of a "Japan Fights On" scenario, even if I might quibble with details. Not least because his outcome is a more unsettling one between the extremes: The problem is that it drags on long enough, with enough destruction, that there is no longer a recognized emperor or central command authority left to surrender when the time comes. So surrenders come here and there, from different factions and commands. This was a live fear by senior Truman Administration and military leaders in 1945. You could have pockets of resistance all over Japan, East Asia and the Pacific holding out into 1947 or even later, because there's no one whose authority they recognize to tell them to lay down their arms.

I can't recall what his final body count was, but it was substantial without being overwhelming. 10-20% or something like that. Which is horrible enough by itself.



I wouldn't disagree with this.

Whether Louise would kill MAJESTIC til spring is not something I am in a position to suss out. Matt Wiser's research aside, the Army did a study 8 months after the war which said it would have pushed it back just 2 weeks. (See below, page 255 from John Ray Skates' The Invasion of Japan: Alternative to the Bomb.) Matt says there was a 1985 Defense Department study which reached the same conclusion. I'm a bit skeptical, but I really can't say I've researched it enough. (Does anyone have daily weather results for Southern Kyushu in November 1945?) Of course, every week you push it back, the weather gets worse. It would make the English Channel feel like Walden Pond.

View attachment 573829
But the screenshot you posted says 30-45 days not 14 days
 
Speaking Giangreco, I just came across a passage (p. 201-03) I'd forgotten about: Marshall did a lot more than just ask Groves about tactical use of nuclear weapons. He had his staff drawing up a detailed plan for using both atomic weapons and poison gas on X-Day - and he was preparing to advocate for it aggressively to Truman, to the point of urging that all further attacks on Japanese cities cease immediately, so that a stockpile could be built up for X-Day:

View attachment 574327

I would really love to see the planning documents Marshall was recalling to Pogue.

Further up, Giangreco notes the reasoning Marshall was preparing to use on Truman: "We'd dropped two bombs in rapid succession. But it didn't do the trick. The shock value is gone now. Bombing another city every week is not going to shock them more." And then he could play on Truman's anguish, expressed in that letter to Senator Richard Russell, about killing "all those kids." So, we use the bombs on invasion targets instead. Ironically, as Giangreco notes, this almost certainly would have ended up killing more people - many of them Americans, thanks to radiation.

This raises two questions for me, one minor and one major. The minor one is that a strategy employing nine atomic bombs on MAJESTIC would have been more than the projections I quoted for bomb production by the end of October. Even if you save the Kokura bomb, that only gets you seven. Perhaps Hanford could have accelerated the rate slightly and it's not showing up in the sources I've seen.

The major one is the looming debate (on top of the others!) on atomic weapon use within the administration that appears to have been starting to gel in early-mid August. Marshall was preparing to push for all bombs to be reserved for MAJESTIC. But it appears to me that Spaatz and Groves would prefer to keep hitting cities. Who would win the debate? Would Truman split the difference? Giangreco doesn't quite say, but the vibe I get is that the deck was stacked in Marshall's favor.

Meanwhile, in the 1950's, MAJESTIC veterans would be wondering about all the strange cancers that their war buddies were coming down with...
Giangreco's book was an interesting read but some of his conclusions I had to wonder about. Like the idea that after reaching the intended stop lines the Americans would be sucked into advancing some kilometres further north in pursuit in the Japanese retreating towards the local redoubt on Kyushu.

I didn't feel as if his explanations for that were very compelling as the Allies were already bypassing Japanese strongholds that were not strategically needed so why would they abandon that on Kyushu when they had achieved their objectives?
 
Yokosuka hadn't been bombed, and apart from a few strays during the Doolittle Raid wasn't really at all.
Its also in Tokyo Bay, so the Government and Emperor would be hard pressed to miss it.
 
But the screenshot you posted says 30-45 days not 14 days
And you are correct, Chris. As near as I can figure, lack of caffeine and sleep generated a conflation of this with a later DoD study. The general tone of optimism about MAJESTIC caused me to skip over that last sentence in the main paragraph when it came time to posting the attachment.

So on the question of timelines, the 1946 Army study is at variance with the 1985 DoD study of MAJESTIC (and the typhoon's impact on it) cited by @Matt Wiser. It does, however, raise the question, which Skates does not discuss, of how an X-Day pushed back to the Dec. 1-15 timeframe, would still have been viable given how weather conditions off Kyushu typically deteriorate by that point. In fact, if I am not mistaken, Chris, you criticized Giangreco in The Red's timeline thread for relying on the 1946 Army study without even mentioning the more optimistic 1985 DoD study. It would seem that the same criticism could be levelled here at John Ray Skates. It seems hard to believe that neither Giangreco or Skates were utterly unaware of it, so it remains a puzzle to me.

In fact, Giangreco sticks with the 45 day delay estimate in a lecture he gave at the University of Kansas in 1998 - and again, without exploring whether mid-December weather would have forced a long postponement:

The Divine Wind, or Kamikaze, of a powerful typhoon destroyed a foreign invasion force heading for Japan in 1281, and it was for this storm that Japanese suicide aircraft of World War II were named. On October 9, 1945, a similar typhoon packing 140-mile per hour winds struck the American staging area on Okinawa that would have been expanded to capacity by that time if the war had not ended in September, and was still crammed with aircraft and assault shipping- much of which was destroyed. US analysts at the scene matter-of-factly reported that the storm would have caused up to a 45-day delay in the invasion of Kyushu. The point that goes begging, however, is that while these reports from the Pacific were correct in themselves, they did not make note of the critical significance that such a delay, well past the initial - and unacceptable - target date of December 1, would have on base construction on Kyushu, and consequently mean for the Honshu invasion, which would have then been pushed back as far as mid-April 1946.​
If there had been no atom bombs and Tokyo had attempted to hold out for an extended time- a possibility that even bombing and blockade advocates granted- the Japanese would have immediately appreciated the impact of the storm in the waters around Okinawa. Moreover, they would know exactly what it meant for the follow-up invasion of Honshu, which they had predicted as accurately as the invasion of Kyushu. Even with the storm delay and friction of combat on Kyushu, the Coronet schedule would have led US engineers to perform virtual miracles to make up for lost time and implement Y-Day as early in April as possible.​
Having said that, this also underlines (to me, at least) how Giangreco generally takes a somewhat more pessimistic view of DOWNFALL's prospects than Skates (or for that matter, Richard B. Frank) does. I am unable to resolve the analysis of the 1946 and 1985 studies (since I have not even read them) to offer a conclusion about which was closer to the money. And of course we would also need to look at the historical weather off Kyushu in November and December 1945 to know what actually would have happened with a November 15 or a December 15 X-Day - something I have not done, either.

I am reminded, by the way, that The Red opted for the 1985 study result of a 14 day delay and posited an X-Day on November 16 (though he trims back Marshall's plan of 8-9 nukes in support of MAJESTIC to just 3).

EDIT: Your second post:

Giangreco's book was an interesting read but some of his conclusions I had to wonder about. Like the idea that after reaching the intended stop lines the Americans would be sucked into advancing some kilometres further north in pursuit in the Japanese retreating towards the local redoubt on Kyushu.
I saw that, but I feel reluctant to evaluate it without more study. I am tempted to wonder if it is of a piece with his generally more pessimistic take on DOWNFALL. Perhaps I might say that the willingness of MacArthur to push beyond the stop line might depend on how the fighting had actually progressed, what casualties he had suffered, what the state of Japanese defenses in the area around the stop line was, etc. That said, we all know that MacArthur wasn't averse to jumping past stop lines!
 
Last edited:
Top