While reading the new book on the Iowa by Lawrence Burr,1 I stumbled across a statement that threw me for a loop. He claimed that after the bombardments of Japan in July 1945, Iowa was ordered back to do gunnery practice at Kaho‘olawe in Hawaii, shooting on July 29th and 30th, and explicitly pointed out that she missed several airstrikes Third Fleet conducted during that time. This was a baffling thing, and I was having trouble believing it had actually happened. Third Fleet was off Japan at the time, and every effort had been made to keep as many ships forward-deployed at sea as possible. A massive underway replenishment apparatus was built, and when that wasn’t enough, mobile bases were constructed at islands across the Pacific. Ships only got sent as far back as Hawaii when they needed serious yard work, more than could be done in the forward-deployed floating drydocks. And absent battle damage, that typically didn’t happen until the ship had been on the line for a year or more. Iowa had been stateside until the end of March for a refit, leaving her as the least likely ship to be sent back.
Iowa refuels from the oiler Cahaba in July 1945
But even beyond that, this story didn’t make sense on several levels. First, doing this would have deprived Third Fleet of a powerful anti-air platform while conducting air strikes on the Japanese mainland. Strikes were flown on July 24th, 25th, 28th and 30th, and further strikes in early August were cancelled due to weather. Given the distance between Japan and Hawaii, even at 20 kts, her maximum plausible cruise speed, Iowa would have taken a week to make the trip, and it’s over 9 days at a more typical 15 kts. Second, even if Iowa’s gunnery during the bombardments was so bad that more training was urgent, there was no reason to send her all the way back to Hawaii. All you need is an uninhabited piece of land, maybe with a smokescreen if you want to practice air spotting and make sure the crew doesn’t cheat. This wasn’t exactly difficult to accomplish in the western Pacific, and by doing it in the Philippines or even off Okinawa, the amount of time the ship is out of action is greatly reduced.2
So at this point, I started looking at other sources. Surprisingly, Iowa’s own website repeats the story, as did Malcolm Muir’s The Iowa Class Battleships, a book that I generally think quite highly of, and that was almost certainly the source for both Burr and the ship’s own website. Muir cites a record from “Naval Gunfire Training Section, Fleet Marine Force, Marine Corps Historical Center”, but something still felt off. If he had cited the ship’s logs, I would have probably been satisfied, if still very confused, but this kind of dash across the Pacific is extraordinary enough that the source being wrong was a serious possibility.
One of Iowa’s SC Seahawks
So I kept digging, and the counterevidence came soon enough. First, the book on Iowa from the Legends of Warfare series had a couple of pictures credited to an Iowa floatplane from raids on the 24th and 28th. Now, there’s no accompanying text outside of the captions and no citations, and even if the photos are genuine, it’s not completely impossible that Iowa left her floatplane behind during the run. But if the purpose of the jaunt across the Pacific was shore bombardment training, it seems weird that they’d leave the key spotting element behind. A brief check didn’t turn up the photos online, so I couldn’t be sure they weren’t mislabeled, but it did send me digging deeper. And I noticed that when I wrote about this period, I mentioned Iowa taking part in the big TF 38 UNREP on July 21-22, which would mean she had to make 20 kts the whole way west at the very least. There are details which probably came from a document Dave Way put together that I was having trouble finding a copy of because I wrote that section north of 7 years ago. But at the very least, I wasn’t going to invent a number of shells transferred out of whole cloth, so it was safe to say she was there for that.
I also figured that a run back to Hawaii would be notable enough to be noted in the cruise book, so I went looking. Unfortunately, the main online repository of such things had only one relevant volume, the WWII victory book, but this had a map of the ship’s operations from 1943 through September 1945. And it certainly looks like the ship only stopped in Hawaii on the way to and from Hunter’s Point for the late 44/early 45 refit, with no sign of a trip direct from Japanese waters to Hawaii.
But then I finally tracked down the document Dave Way put together in the deeper recesses of my hard drive, drawing primarily from Iowa’s war diary, and there was a lot of detail about operations with TF 38 over the relevant window, and nothing at all about a trip to Hawaii for gunnery training. It looks like this whole thing originated with Muir, and a report that he either misread or that was just wrong.3 And I wrote this post largely to correct the record. Simply put, Iowa did not do gunnery training at Kaho’olawe in July 1945. She spent the entire month in the Western Pacific.
1 Burr is not a good writer, and the book is not very good. Don’t buy it. ⇑
2 Also greatly reduced is the risk of submarine attack, although Indianapolis wasn’t sunk until July 30th, so this may not have been foremost in the minds of Pacific Fleet planners. ⇑
3 I am not trying to beat up on Dr. Muir, who did a lot of excellent work in his book. I originally suspected someone might have confused July 1945 with July 1944, but Iowa does not appear to have been near Hawaii in July 1944 either. I tried to get in touch with Dr. Muir, but was unable to. ⇑
Comments
The Unauthorized History Of The Pacific War podcast recently did a great ep on the battleship raids along the Japanese coast - these guys get down into the weeds in details (always entertainingly so), and never mentioned a thing about any of the battlewagons having accuracy issues. Believe me, they would have. ;)
My recollection is that gunnery wasn’t great from any of the ships, but that’s mostly because that kind of thing is quite difficult, with a side-effect of not training on it all that much.
Seems the BBs in the Pacific during WW2 broke down into two rough groups. Old Standards that did mostly shore bombardment and Fast BBs doing escort for the CVs. While there was plenty of cross pollitaion, those where their core duties.
Remember hearing and reading the Fast BBs did not do a good job of shore bombardment at Saipan. From what “Unathorized History....” says, they got better by the time they went to work on mainland Japan.
That raises the question of why the old battleships weren’t doing this bombardment if they were more skilled at it.
(Warning: this contains data from Wiki, which also repeats the claim that Iowa was at Kaho’olawe, and speculation.)
These bombardments were from long range (~30kyd) to reduce the risk of mines, so some of them may have required the Iowas’ longer maximum range. (Standards shooting HC have a maximum range of ~29-37kyd, mostly depending on how worn the individual gun is rather than which type it is, and Iowas ~36-42kyd.)
This long range, and poor visibility (from the weather and/or smoke from fires started by the bombardment itself), would also have reduced accuracy whichever ships were doing it.
Given that much of the point of this bombardment was to provoke Japanese land-based aircraft into revealing themselves (by attacking the bombarding ships) so they could be shot down, as opposed to actually doing damage, good AA and deck armor may have been more important than shore bombardment skill.
The main thing the old battleships were doing at this time seems to be training for the planned landings on Japan: around 5 doing that (mostly at Leyte Gulf, confirming that if Iowa did need a training facility then there were closer ones than Hawaii), 3 under maintenance/repair/upgrade, and 1 returning from these. 2 were patrolling around Okinawa, USS Nevada apparently was with the force off Japan but didn’t participate in the bombardment (no reason given), and USS Colorado is unaccounted for.
Yeah, the early fast BB bombardments weren’t great. I may have to dig more into how accuracy was this time. Part of it was a much softer target, and maybe lower standards.
@muddywaters
The reason is simple: the fast battleships were part of TF 38, and it was easy to detach them from the carrier force and send them in. The slow battleships were not, and it would have required separate arrangements to bring them in. Might even have still been with 7th Fleet, which would require getting permission from MacArthur. Also, they were slow, and that’s not something you want if you’re doing something like this, because it greatly increases the amount of time you spend close inshore, where trouble is more likely.
I actively disbelieve this. She was maybe capable of 20 kts, and the slowest ship otherwise with TF 38 could make 27. Having her cover the oilers doesn’t even make much sense given the lack of surface threat and the way that the replenishment groups were handled.
Fortunately, I can check my belief on that, because I have a book on Nevada in my library (Silver State Dreadnought). It turns out that, no, Nevada was not with the fleet off Japan, and Wiki’s sources don’t even suggest that she was. She was covering minesweeping operations in the East China Sea off Okinawa. The book has her there from July 2nd to August 8th, with no jaunts to join any part of TF 38. Also, she was barely capable of 18 kts at that point, which is definitely a no-go to running around anywhere you might need to leave in a hurry. As before, my instincts were correct. Not even sure where Wiki invented that from, because I don’t see it in their sources.
(Also, there was belief aboard Nevada that they’d host the surrender as the only ship to get underway at Pearl. So far, every battleship in the Western Pacific I have details on thought they’d have hosted if not for Truman’s unreasonable bias in favor of Missouri.)
Does it make sense to interpret Saipan as basically live target practice, i.e. that having them not-very-effectively bombard a real enemy was considered a better use of resources than having them go to a practice range?
(Saipan was June 1944, so some of the old battleships were busy in Normandy, but others did arrive the next day.)
Presumably being bad at shore bombardment also makes it take longer to do a given amount of damage, but I don’t know which is the dominant factor. And if the plan is to run away when you see the threat, then it is a question of raw speed not time spent nearby.
Yeah, there’s probably a lot of truth to that interpretation of Saipan, and most other bombardments from the fast battleships early in the war (Mili, Ponape, etc.)
I don’t think that “given amount of damage” was the plan. Based on Iowa’s performance, if she was firing 1 shell/gun/min, the bombardment lasted about half an hour and used up ~25% of her shells. I would guess that it was “shoot for a given amount of time/shells, whichever comes first, then leg it”. Real-time BDA is really hard against land targets if all you have is WWII optics. (Oh, and at least one bombardment was done completely blind, so there was no way at all to check effects until later.)
I don’t think that makes sense either. A quick skim of the action reports doesn’t give timing, but I doubt it was over an hour, during which time the ships could have moved at most 27 nautical miles. That’s really not that much against an air threat, so time in danger is going to be dominated by moving into and out of the bombardment area, not the length of the bombardment itself.
You don’t have to be observing the damage in real time to have a general idea, from either scored target practice or less formal estimates, that (made-up numbers) 30min of Iowa shooting = 15min of Standard shooting.
(And yes, it’s unlikely to be that much longer without a break because battleships only carry ~1 hour of ammunition, so it’s very possible that the Iowa still does it faster when transit time is included.)
Also agreed that some bombardment jobs are easy enough that you use what’s easily available, even if they’re bad at it.
(By “live target practice” I mean a slightly stronger hypothesis, which I’m not claiming to have proof of: that part of the point of these easy jobs was to make these ships better at shore bombardment, so they could do harder ones later.)
I wasn’t thinking of running away as for avoiding air threats (as you note, that generally doesn’t work).
(Though should I take that to imply that “much of the point of this bombardment was to provoke Japanese land-based aircraft into revealing themselves (by attacking the bombarding ships) so they could be shot down, as opposed to actually doing damage” is also wrong?)
I was thinking of running away as for slower-moving or fixed short-range threats (torpedo boats, land-based guns, etc), though I don’t know whether those were a significant risk in this case. (There isn’t much that’s actually between Standard speed and Iowa speed except other battleships, unlikely by this point, but running away from torpedo craft gives you more time to shoot them.)
Granted that you don’t need to have a formal BDA process to be able to say “we expect this ship to be X better, so will need fewer shells”. I don’t think that such logic was actually behind the decisions here. My guess is that whoever was in charge of the battle line bugged Halsey to let them shoot at Japan, and eventually Halsey decided to go ahead with it.
For any value of “go from a point well offshore in to shoot at shore targets, then head back to safety”, I think the higher speed of the fast battleships is going to dominate shorter shooting time for the Standards. The main reason to use the Standards is if you want more damage than the fast ships can deliver for magazine depth reasons, and I don’t think that was really what was going on.
The problem is that I haven’t looked into this stuff since 2017, so my thoughts are basically “whatever Iowa Part 4 says”. But with that disclaimer out of the way, “we’re going to dangle the battleships as bait for air attack” doesn’t seem right. For one thing, as noted, at least one bombardment took place in weather too bad for air spotting, which is also going to make air attack and providing air cover pretty difficult.
I think this significantly understates the difference that the speed makes. The Standards by this point could maybe make 20 kts, as they were old and not necessarily that well-maintained. I’m pretty sure the fast ships were in better shape here, so say 26 kts. Practical top speed of a cruiser/torpedo boat/whatever is likely to be around 30 kts, depending on sea conditions. So you have 10 kts advantage over the Standards and 4 over the fasts, which in turn means a lot more tactical options for the fast ships. Not just “we can run away for longer”, although more than halving the closing speed does mean that, but also stuff like making torpedo solutions significantly harder.
We probably agree more than this looks like (this often happens to me).
Wiki does give some durations, but given how many other things they get wrong, I’m not sure whether to believe them: ~2 hours of shooting to fire ~30 shells/gun = one every ~4min (~1/8 of their nominal maximum rate), and (for a different one) 6+ hours visible from the enemy shore, which they considered dangerous. Shooting at less than maximum rate to allow for spotting plausibly improves accuracy, particularly if multiple ships are firing at the same target, but I don’t know whether that slow is plausible.
And yes, agreed that bombarding at night/in poor visibility looks like trying to avoid air attack, not trying to provoke it.
If you mean that running away from the torpedo itself reduces its range so the attacker has to get closer, yes and was vaguely counting that in “more time to shoot them”.
Also relevant: how much skill is there in shore bombardment, i.e. how much room is there for practice to be useful?
Original (somewhat later) 5″/38 and general instructions.
As usual, the fire control system automatically corrects for the ship’s motion, both roll/pitch and forward movement, so targets you can see (and that are near sea level) are an easy point-range-shoot.
However, shore targets are often not visible to the firing ship, requiring indirect fire, i.e. using offsets from a visible reference point(s) and/or spotting by off-ship observers. This was not natively supported (the Mk48 moving map the Iowas would later have didn’t exist yet), so required some manual steps.
The time shooting sounds plausible. You can’t just park a battleship somewhere, and the mechanics of getting ships into and out of firing position could easily mean the entire bombardment takes two hours and the gun crews are only working for a quarter of the time each. I’d guess that the time visible from shore is a bit high. A battleship should be visible maybe 12-15 nm from someone on the shore, a distance that it will take ~half an hour to cross at full speed. Double that for an observer of about the same height. It’s hard to see being visible for two hours on either side of the bombardment making sense, although I suppose that depends heavily on the shape of the coast nearby.
I remember reading a tiny bit about the process, and it was quite complicated, particularly before they got the ballistic cams for the 1900 lb shells, so there was more manual correction there. (Which may or may not have been part of the issue at Saipan, I can’t recall.) But it’s been a very long time since I looked into this. I suspect that skill was a very significant part of it, particularly practice in quickly and accurately doing the various manual steps required. Automated systems take less skill, as evidenced by PoW at Denmark Strait.
At a guess, typically bombardments during/after a landing are the difficult ones (the ongoing land battle makes it urgent, and a bad shot might accidentally hit your own troops), while ones well before a landing are the easy ones. (Assuming that in both cases, you have access to plenty of ammunition and the enemy is unlikely to have the means to seriously threaten a battleship.) The examples above seem to fit that: at Saipan the old battleships arrived before the actual landing started, while Mili and Ponape were bypassed.
Also on the height of the terrain: Japan has lots of 1-2km high mountains (horizon of ~110-160km, so visibility is likely to be the limiting factor), but I don’t know whether they had portable enough communications to usefully put observers on them. The 6+ hours was at Muroran, which is also in a ~100km wide bay.