Monday, February 24, 2014

The Siege: Big Week

This week is the seventieth anniversary of Big Week, five days of bombing, 20--5 February 1944, the greatest single victory in the history of the USAF. 

But, first, some introductory remarks, because Big Week had an introduction, too. One will encounter persons, who will remain nameless here, who imagine that the strategic air campaign against Germany as the vain and futile culmination of a virtually entirely theoretical approach to war, or perhaps a cult, or perhaps some kind of self-interested pursuit of institutional (that is) air force interests.

But Big Week had an introduction, and I shall introduce it.

For a few months after I graduated for the last time and had occasion to understand just how much society actually valued the skills that it had so generously subsidised my acquiring, I lived in the kind of flop house that a kind-hearted person will run, in a house on a property worth far more than what was left of the carpentry on it. Buildings fall down on their own if they are neglected, except in a magic world where effort and result are decoupled in whatever way we will, just so that it supports whatever thesis we propose.

Not this, but like it. (It's much nicer now.) Huffington Post

Which is not to say that houses can't be helped to fall down. I lived there with assorted persons and a sweet-tempered Golden Retriever. One of the persons in and out of that house did not like dogs, so when he was around he put the dog in her little kennel. He was glad to be shut of her, and she was perfectly content in her kennel.

Anderson bomb shelter, via Wikipedia.

Close spaces make us feel safe. We dream. We dig, imagining that in depth there is safety.

Subterranea Brittanica
We are wrong. The sign points us to the way out. What humans build, humans can destroy.

This is a postwar picture of the Siemens Schukert works at Nuremberg (credit via Crown), although the corporate identity had been subsumed into their "Bramo" trademark, and then been subsumed again as Bramo's early jet turbines were taken over by BMW.

The raid that destroyed this plant was probably the one of 01/01/1945, of which one website proposes that

 This plant was not building gingerbread confections, although the bomb that destroyed it was probably a 4000lb HC "cookie."

Back up a moment, though, to the bomb shelters of Brooklands, which were only effective if people were actually in them. On 4 September, 1940, they were not, as 14 Messerschmitt Bf-110s swept overhead at 1:24pm. A single 250lb bomb punched through the roof of a machine tool shop and detonated on top of a heavy press. A moment later, 83 machinists were dead and 419 were injured. John Terraine's Right of the Line frames this as a raid that "virtually stopped the production of Wellingtons for four days." 

I doubt that the survivors framed it that way. On 24 September, a similar low-level raid Bf 109s hit Vickers-Supermarine Southampton, killing 42, injuring 161, and cutting off municipal services. In the bloodless words of the US Strategic Bombing Survey, this would count as the "loss of a day's production." It was also pretty much the Germans using up the opportunity that a change in tactics always presented. Once barrage balloons were shifted from previous priority targets and proper precautions against splinter damage had been taken, these plants would be very difficult to attack so effectively with low-level tip-and-runs.

This had not escaped the attention of the German Air Staff. On 26 September, 1940, 76 Heinkel He 111s and Ju 88s, "accompanied by 60 Me[sic] 110 twin-engined fighters" managed to drop 70 tons of bombs on a one-square-mile area around the works.Vivid eyewitness testimony describes how a loft where completed wings above a bomb shelter was hit, and how a "blazing tide" of wing ribs followed by burning dope and paint poured into the darkness of the shelter." Most of the 55 killed, although not of the 92 injured, would have been in that shelter (251). 

That was the end for Vickers-Supermarine Southampton, as the works were dispersed. All production at the home of the Spitfire was lost for several months, although fortunately by this time Castle Bromwich was in full production. Even after dispersed manufacture was in full swing, it was calculated in another context that there would have been a loss of 20% of labour on fighter manufacture. (February 1944, 75.) It would have been worse for bomber manufacture, although, fortunately for the RAF, the great Midlands bomber plants that made the Halifax and Lancaster did not have to be dispersed. The German Air Force was never strong enough to make precision daylight attacks into the Midlands, and had instead to resort to area night-bombing campaigns that strove, as second-best to paralyze the general industrial life of a city.

In Big Week, the USAAF did that. (Also, Bomber Command did a fine job of coordinating its night attacks on the "Big Week" targets. But it wasn't beating the German Air Force. It was trying to avoid it, and had not yet come to terms with the fact that it could not do that any longer. a new technological paradigm was needed, but that toybox was going to be kept locked until the night of June 5/6, for fairly obvious reasons.)

On the night of 19/20 February, Bomber Command kicked off, technically. 823 aircraft went to Leipzig, 561 Lancasters, 255 Halifaxes, 7 Mosquitos. 78 aircraft - 44 Lancasters and 34 Halifaxes were lost, 9.5 per cent of the force. The Bomber Command War Diaries, now online so that I can quote them directly here (you will notice that this passage is "heavily dependent" on the linked source), notes that Halifax loss rate was 14.9 per cent of those which reached the enemy coast, and that the disappointing, Merlin-engined Halifax IIs and Vs were permanently withdrawn from operations to Germany after this raid.

The intended diversion, 45 Stirlings and 4 Pathfinder Halifaxes minelaying in Kiel Bay, was unsuccessful, and German night fighters attacked the bomber stream all the way to Leipzig, which was still cloud-covered, interfering with bombing and leading to further casualties.

The weather, predicted to break and give five days of perfect flying, was still uncooperative as the morning rolled around. Richard Davis ("heavy dependence" alert!) records that Eighth Air Force Fighter Command reported that 40% of its P‑38 force was affected by engine trouble that, combined with icing, would severely reduce their range. This was a serious matter, as the P-38 was the most important long-range fighter available. Sixteen combat wings of heavy bombers (over 1,000 bombers) would be covered by seventeen AAF fighter groups (835 fighter planes) and sixteen RAF squadrons, but only a single Group of P-51s. 

This was especially troubling, as fighter escort tactics required "relays." It is very uncommon for an aeroengine of the era to be able to operate at anything like as little as 55% maximum rpm, as lean mixture heat (and, in the longer run, lead attack) burned out the engine. Bombers at full war load are likely to trundle along at 150mph IAS, while fighters could hardly throttle below 250mph. Fighter escort thus went in legs, with short-ranged Spitfires covering forming-up and initial penetration, P-47s covering the mid-range, and longer range P-51s and P-38s (hopefully) taking over at the target and return. (Wikipedia reports 94 P-38s, 668 P-47s and 71 P-51s. The Thunderbolt really deserves more respect. The fine details of how the engineering achievement was managed matter a great deal less than the fact that it was.)

Under these tactics, a very large number of escort fighers does not amount to any great excess of "little friends." In this situation, it is all the remarkable that Doolittle was fielding attacks on no fewer than twelve separate plants. Six unescorted bomber wings flew a northern route to bomb targets near Posen and Tutow, while the rest of the bomber force, escorted by the entire fighter force, flew toward Leipzig and Braunschweig in central Germany, while 135 medium bombers, two‑­thirds of which aborted because of weather, attacked airfields in western Europe. 

Only twenty‑one heavy bombers failed to return to base. The Baltic force encountered cloud and area bombed, while the main force bombed visually and seriously damaged four plants manufacturing Ju 88 (night‑­fighter/bomber) aircraft in the Leipzig area and two plants manufacturing Me 109 (day‑fighter) aircraft. Davis cites the AAF official history, as indicating the loss of one month's production of Ju 88s and severe damage to "about" 32% of Me 109 manufacturing capacity.

Bomber Command's diversion the next night was more successful, 552 bombers attacked Stuttgart's engine plants out of 885 dispatched, with 8th AF fielded 30 bomber groups and 17 fighter, with more than 800 bombers and 700 single-engine fighters sortied and 762 bombing, exclusive of 15th AF (and 205 Group RAF) attacking from the south. 

The 22nd saw 8th AF field more than 800 bombers for the third consecutive day, although Bomber Command had had to stand down the previous night. Bombing was somewhat ineffective, but the fighters made high, although of course exaggerated claims.

Davis characterises the raids of the 24th as showing how the tables had turned. 8th AAF sent out over 800 bombers and more than 700 fighters, while that night, Bomber Command would sortie 734 aircraft in Bomber Command's first raid on Schweinfurt, following up on a 266 bomber American raid that day. Bomber Command's diversionary tactics worked this night, while in the sunlit hours before, 1st Bomber Division recapitulated its disastrous raid of the previous fall, but suffered only 11 a/c lost vice 60. 

In justifying the continued strategic air offensive, Carl Spaatz had argued that the Germans would have to be challenged over targets that their fighters had no choice but to defend. Say what you will about "panacea targets" and the dispersal of German industry, but more than two-thirds of the nation's roller bearings were still coming from Schweinfurt on this late February night. By the next morning, that would change, because German roller bearing production would be down by half. 

The defeat of Big Week reverberated through the whole system of the Air Defence of the Reich. As Allied and Axis prepared for the culminating battles of the spring that would precede the invasion, the Germans diverted all available pilots (and fuel) to the battle. Germany would never have a more favourable opportunity to inflict aircrew and aircraft losses on the Allies on favourable terms as in this campaign. The Battle of Britain, in spite of the losses to aircraft production inflicted by the German raids with which I introduced this discussion, is reckoned as an Allied strategic victory for precisely this reason. In the long run, the loss of Stirling, Wellington and even Spitfire production occasioned by the battle was irrelevant against the damage done to the German air arm. 

Now let's look at Ninth Air Force's Big Week, via Wikipedia):

20/02: 35 B-26 Marauders bomb Haamstede Airfield, The Netherlands, as a target of opportunity, after about 100 B-26s abort attacks on other airfields because of weather.
21/02: 18 B-26s bomb Coxyde Airfield, Belgium; weather causes almost 190 aborts. The Ninth Air Force's Pathfinder Squadron (provisionally activated on 13 Feb) takes part in this operation, its first venture into combat. 185 aircraft scheduled to attack other airfields in the Netherlands and France in the afternoon are recalled because of bad weather.
22/02: 66 B-26s bomb Gilze-Rijen Airfield, The Netherlands; bad weather causes 100+ others to abort.
24/02: 180 B-26s attack NOBALL (V-weapon) targets and Rosieres-en-Santerre, France. Bad weather makes bombing difficult and causes 34 other B-26s to abort
25/02 191 B-26s bomb Venlo, Saint-Trond, and Cambrai/Epinoy Airfields, France in a morning raid as a diversion in support of the VIII Bomber Command heavy bombers over Germany; 36 abort, mainly because of a navigational error. 164 B-26s dispatched against military targets in France during the afternoon are recalled because of bad weather.

Every one of these "tactical air counter-offensive" raids used as many pilots, navigators, bombardiers, and much of the same ground crew as a heavy bomber mission. The abort rate, which will be familiar from such hits of low-flying anti-airfield bombing campaigns as No. 2 Group's contribution to the Battle of France. (War Diaries! Bowyer!), reflects the fact that it is actually harder to navigate a small plane than a big plane because of lack of space and instruments and such. 

In sum, the Battle of Britain  was a defeat for Germany. So was No. 2 Group's campaign against the advancing German army in 1940. So, if anyone cared enough to keep a scorecard, was the 9th Air Force and Fighter Command's attempt to engage themselves in supporting the combined bomber offensive with precise, low-level attacks on Axis tactical air assets. Forcing the enemy to shoot large quantities of light AA off and patch holes in runways is not a satisfactory pay-off for the loss of an aircrew with cumulative thousands of in-flight training time.

Big Week, by contrast, was a victory for the Allies. Why? What made it different from the Battle of Britain?

i) Scale of aircraft production. At the same time as I concede the importance of all of those great factories, I want to argue that this advantage is greatly exaggerated. Far too many of the planes produced did not count for various reasons. At the sharp edge, the Allies did not have enough planes, while the Germans did.

ii) Scale of aircrew training. This was a much more significant advantage. The Allied planes that were available could be manned, and manned well, while the Germans ran out of pilots during the campaign. The pilot supply is normally singled out as the limiting factor on the RAF's resilience in the Battle of Britain and was a well-established constraint on air power. (Not every "air strategist" of 1940 had his head in the clouds of pure theory. They could also look at what happened in WWI).

iii) Technological obsolescence. The German fighter arm was flying an old high altitude fighter and a new low altitude fighter, and this, at last, came back to bit them. "Fokker panic" is not a nice thing. 


iv) The Allies were not just dropping bombs on random places to make the Germans come up and fight and lose an unfavourable battle of attrition, because that's what you do with a wasting asset in the months before the invasion that is going to determine the outcome of the war.

No. The Allies were attacking the factories that built the planes and the engines that powered them. We have already seen how the roller bearing shortage  undermined the DB605 and doomed attempts to build a suitable high altitude fighter. (2) We can see now, with access to German records, that the exaggerated claims of air-to-air combat losses were not what "defeated" the German fighter arm in the spring of 1944. It was, rather, the belated and necessary decision to disperse the industry. German fighter production would roar back in the summer of 1944. And it would be too late.

The Combined Bombing Offensive worked. It turns out that you can bring the house down on people, and even the coziest little kennel provides only the illusion of security.


  1. I think that ii is by far the most important of the factors here. And this is actually related to my favorite chart to explain the way that World War Two went, Table 64 from the League of Nations 1937-8 Statistical Yearbook on crude oil production in 1937.

    The US produced 60% of the world's crude in 1937. Venezuela produced another 10%, the rest of the Western Hemisphere produced 6%. The USSR produced 10%. Egypt through Iran combined for about 6%.

    At best, the Japanese empire could reach up to 3.5% of the world's 1937 oil production. Unless Germany could capture the Caucasus, the best that they could reach would be about 3% of the world's 1937 oil supply[1].

    That difference in fuel production meant that the US could fly 800 four engine bombers, along with 800 fighters, and hundreds of twin engine bombers, and the RAF could fly hundreds more four engine behemoths, all with reasonably well-trained crews, while the Germans faced a choice of whether to move the panzers, supply the panzers, defend the Reich, or train the fighter pilots, using their insufficient supplies of fuel. [2]

    Using that interpretative framework, the panacea was not Pointblank but the Oil Plan (and the Transportation Plan limiting their ability to ship oil). That's why Germany's massive production of fighters in 1944 (almost 100% increase on 1942 German total aircraft production) occurred while the Allies gained air dominance: they didn't have enough fuel to train more pilots. Now, admittedly the numbers produced were a lot of obsolete ME-109's and FW-190's that were behind the performance of the top Western Allied fighters, but that seems to be more due to Speer's decision to concentrate on producing sheer numbers of aircraft rather than converting factories to newer models. But even if they had produced more advanced fighters, absent enough oil they couldn't contest Allied air dominance over the FEBA.

    In other words, trying to take on the world with less than 10% of the world's oil wasn't a great idea.

    [1]: Yes, Germany massively expanded it's synthetic oil production during the war, but that was not nearly enough. German synthetic oil production peaked, according to USSBS, at about half of 1937 Romanian production.

    [2]: This was the other thing that crippled Italian military performance, besides their unwillingness to produce top-quality engines as you correctly noted in a previous blog. The Germans refused to share enough oil to train their pilots, move their tanks, or sail their ships enough.

    1. a couple things.

      Oil production, granted oil fields, is a function of investment as much as proven reserves, and this was far more true in 1937 than today. The United States in 1937 was still benefiting from the investment boom of the 1920s, and the figures going forward from 1937 are likely to be less impressive. In terms of potential, the United States as of 2012 has 22 billion barrels proven reserves, Rumania 1.05, and Hungary plus the Greater German Reich almost 800 million --and the European reserves were better known and more accessible than the American, which include Alaskan reserves.

      I'm not arguing with your point here, Chris, just observing that the oil case was less clear in 1944 than it might seem. American air planners were correct to see it as a chokepoint in the German war economy --just as the Germans did! But how accessible was it to the means of attack available? How many targets, located where, will have to be attacked with what weight of bombs (and with the aircraft implied by the weight of bombs to be determined) in order to execute the Oil Plan? That the Oil Plan succeeded is true. But it only did so after the Invasion, and there is good reason to think that the Invasion, by bringing OBOE installations forward, made the Oil Plan offensive possible in its turn.

      By the panacea argument, ACM Harris means that on a given day in the ETO, it is going to be impossible to attack a given target by visual means. If we concede that when visual means are available, we should deliver a precision attack against a panacea target, most bombing sorties are still going to be area attacks. Given limited bombing opportunities, Harris wants to argue, why not concentrate efforts on a single strategic aim? I don't want to be seen arguing Harris's case. It's pretty much discredited by events, after all, but what events tell us is that the Invasion is going to be a key factor in the success of the bombing offensive, just as the CBO contributes to the success of the invasion.

      What do we need to make the Invasion successful? Above all, air superiority. Remember that the blueprint for achieving air superiority over a land offensive is very well known by 1944. You "lean forward" against the enemy, deliver bombing attacks against targets that they must defend, and then wear them away in air-to-air combat.

      It is the "vital targets" that is the challenge here. Having observed the failed German air offensive-preparatory-to-invasion that was the Battle of Britain, air planners knew exactly which target had enjoyed that status in 1940. Not "oil targets" (successfully engaged in the ports offensive) but rather aircraft factories. Had the Germans shut down Castle Bromwich, they would have shut down Fighter Command. They couldn't, and they didn't, but that doesn't change the fact that the calculus was clear.

      So that's the agenda of Big Week. Force the German aircraft industry to disperse by making effective attacks on the big factories. Once that is accomplished, the CBO will have the luxury to decide on the next step and its next focus.

      But that is the first step.

    2. ChrisM, do you have those figures in terms of exports or of spare capacity? The US had 60% of production, but it probably also absorbed a lot of that?