The Bombing War: Europe 1939–1945

by Richard Overy

  The two ambitions were not easily compatible, and in the event difficult to achieve. Over the course of the late summer and autumn both forces carried out major raids that became large-scale air battles of increasing severity and intensity as German defences were strengthened and the German Air Force was freed at last to pursue more effective operational tactics. On 12 August 133 B-17 ‘Flying Fortresses’ attacked Bochum in the Ruhr-Rhineland and lost 23 of their 133 aircraft (17 per cent of the force). The next major battle came on 17 August, deliberately chosen as the anniversary of the Eighth Air Force’s first mission in 1942. The object was to inflict a spectacular blow. The targets selected were the ball-bearing works at Schweinfurt and Messerschmitt Me109 fighter production at Regensburg. For this raid American bombers were divided up; those destined for Regensburg under the command of Curtis LeMay would continue on to North African bases, those for Schweinfurt, commanded by Brig. General Robert Williams, had to fight their way there and back. Both cities were the furthest the Eighth Air Force had yet flown into German airspace and involved the largest number of bombers so far dispatched, a total of 376 B-17s – a reflection of a sudden escalation in the supply of both crews and aircraft. The Messerschmitt works produced 18 per cent of all Messerschmitt Me109 production; the ball-bearing works at Schweinfurt produced 45 per cent of the supplies of German ball-bearings. The aircraft set off in the late morning of 17 August and were attacked by German fighters from the moment Allied fighter cover ended, at Antwerp on the Channel coast. The group destined for Regensburg reached the target near midday and dropped 298 tons, killing 400 people and causing the temporary loss of 20 per cent of fighter output. On the way there 12 B-17s were shot down, 12 more were lost in the aftermath as they flew on across the Alps to Tunisia. Poor repair facilities in North Africa left more bombers grounded, and after a long return flight with further losses, only 55 out of the original 146 returned to English bases.161

  The First Bomber Division faced an even greater battle. For more than three hours in the afternoon flight the bomber stream was subjected to persistent fighter attack, losing a total of 36 aircraft. Around 3 o’clock, 424 tons of bombs were dropped on the Schweinfurt works, killing 141 people, destroying two works completely and seriously damaging a number of others. The German Armaments Ministry recorded a temporary loss of 34 per cent of ball-bearing production, though large reserve stocks were available to cushion the blow.162 For the Eighth Air Force the cost for just two targets on their list was exceptional. Together with the 60 aircraft shot down, 176 were damaged and 30 remained in North Africa. Including those lightly damaged, the casualty rate was 71 per cent; counting those destroyed, severely damaged or out of theatre, the loss rate was 31 per cent, levels that could scarcely be sustained for more than a few more raids. The German Air Force lost 28 fighters to the concentrated fire of the B-17s, though in common with almost all air-to-air engagements, American aircrew claimed to have shot down a remarkable 288 enemy aircraft.163 In September a raid on Stuttgart, undertaken while Arnold was visiting Britain, saw the loss of 65 bombers (19 per cent) out of a force of 338, with little damage to the city itself.

  The Eighth Bomber Command nursed its wounds for more than a month before beginning a new series of deep-penetration attacks in October. A raid on the coastal city of Anklam on 9 October cost 18 bombers out of a force of 106, but the most famous battle of the bomber war took place during a second raid on Schweinfurt on 14 October when 65 bombers were lost out of an attacking force of 229, a total loss rate of over 28 per cent of the force. Fighters accompanied the aircraft as far as Aachen, thanks to the addition of extra fuel tanks, which pushed their range to 350 miles, but after that the bomber stream was subjected to hundreds of attacks with rocket and cannon fire from enemy fighters. Eaker wrote to Arnold after the raid: ‘This does not represent disaster; it does indicate that the air battle has reached its climax.’164 But so severe was the risk on any raid past fighter cover that for the next four months operations were only carried out with increasing numbers of bombers on cities within easy range. As a result Kiel, Bremen, Wilhelmshaven and Emden – already the victims of the first year of Bomber Command raids thanks to their proximity – became the recipients of occasional heavy bombing, while the majority of American attacks now took place once again over France. The Eighth Air Force, like the German Air Force and Bomber Command years before, began to think about the possibility of night attacks, and in September 90 B-17s were converted for practice and training in night-flying.165 In three months of raiding the Eighth Bomber Command lost 358 B-17s in combat, achieving the highest loss rate of the war in October. The campaign did not come to a complete halt, but the next raid deep into Germany was made only on 20 February 1944 in very different circumstances.166

  Bomber Command also began to experience higher losses and, despite the advent of the target-finding apparatus Oboe and H2S, continued to hit urban targets with intermittent success. The rate of expansion in the first half of the year was not sustained: operationally ready aircraft within the Command increased by more than two-thirds between February and June, but only by another quarter between July and December. Pilot strength was 2,415 in June, 2,403 in December.167 On 17–18 August Harris was ordered to mount a more precise attack on the German Air Force research station on the Baltic coast at Peenemünde. The raid by 560 aircraft dropped 1,800 tons of mainly high explosive on the research stations and accommodation (also killing 500 Polish forced labourers in a nearby labour camp). Extensive damage was done to the rocket-research programme in one of the few ‘precision’ raids attempted by the Command in 1943, but 40 aircraft were shot down, a loss rate of 6.7 per cent. Harris then shifted his focus to Berlin. The first raid on 23–24 August missed the centre of the capital and hit the southern suburbs, killing 854 people, and destroying or damaging 2,600 buildings. The loss rate of 7.9 per cent was the highest on any raid since the start of the offensive. A second raid on 31 August–1 September cost 7.6 per cent of the attacking force while H2S marking was erratic, causing the bomber stream to drop bombs up to 30 miles from the aiming point. Only 85 houses were destroyed in Berlin and 68 people killed. So heavy were the losses to Stirling and Halifax squadrons that the third attack, on 3–4 September, was made only with Lancasters, but not only did the H2S marking once again miss the main aiming point, 7 per cent of the Lancaster force was lost. The first ‘Battle of Berlin’ petered out until November in favour of less dangerous targets.168 On the night of 22–23 October a second firestorm was created in the small city of Kassel, where H2S marking was for once accurate. The raid report noted that the whole city area ‘was virtually devastated’.169 The death of an estimated 6,000 people was a higher percentage of the city population than in Hamburg. Some 59 per cent of Kassel was burnt out and 6,636 residential buildings destroyed. Armaments production was hit heavily by the destruction of the workers’ quarters, but like Hamburg revived after two or three months to around 90 per cent of the pre-raid level.170 The raid put Kassel for the moment at the top of Bomber Command’s list of the proportion of city buildings destroyed or damaged, ahead of Hamburg, 58 per cent to 51 per cent.171 But Bomber Command lost 43 aircraft, 7.6 per cent of the force, a steady attrition of its strength. Over the course of 1943 the Command lost 4,026 aircraft, 2,823 in combat.172

  The high losses of the autumn months created a growing sense of urgency and uncertainty in both bomber forces, despite the brave face turned to the outside world. The claims that had been made from Casablanca onwards served the cause of service politics as well as Allied strategy; the stalemate developing in late 1943 as the balance shifted away from offensive towards defensive air power exposed the bomber force to the risk of relative failure. Although both forces advertised their success in diverting ever-increasing numbers of German fighters to the defence of the Reich, this was in some sense a Pyrrhic victory, since the bomber forces were now subject to escalating and possibly insupportable levels of loss and damage. Lovett reminded Eaker in September that if Ger
many did not collapse over the winter, the public would think ‘our “full-out” offensive doesn’t work’.173 A paper prepared for Portal in October argued that a failure to demonstrate what strategic bombing could do would have ‘dangerous repercussions upon post-war policy’.174 Both forces wanted the other to help more in pursuit of what was seldom a common ambition. In October Portal told Eaker and Harris that the Pointblank aim to eliminate the German Air Force as a preliminary to a more sustained offensive had so far failed. ‘Unless the present build-up of the G.A.F. fighter force is checked,’ he continued, ‘there is a real danger that the average overall efficiency and effectiveness of our bombing attacks will fall to a level at which the enemy can sustain them.’ The operational evidence showed that Bomber Command had so far devoted only 2 per cent of its effort to fighter aircraft assembly plants (as by-products of area attacks) and Harris was now instructed to launch operations against six principal cities associated with fighter production.175

  Harris treated the request as he had treated other ‘panacea’ targets. In July he had been briefed to attack Schweinfurt as the most vulnerable link in the chain of German war production. Alongside claims for its importance he scribbled ‘sez you!’ in the margin, and no night raid was made. In December he was told again that his Command should bomb Schweinfurt at night. He rejected the idea on the grounds that his force even now could not hit a small city with any certainty, and would not waste time on a single target when there were bigger cities to destroy and ‘only four months left!’176 To the Air Ministry’s complaints that Bomber Command should have done more to support the Eighth Air Force campaign against German air power, Harris retorted that he would not do their job for them. The two forces could not be regarded as complementary, Harris continued, since his Command had dropped 134,000 tons in 1943 and the Eighth Air Force only 16,000, much of it on less important targets.177 When RE8 researchers suggested that Bomber Command attacks had by their estimate actually done little more than reduce German economic potential by 9 per cent in 1943 (even this figure turned out after the war to have been over-optimistic), Harris responded angrily that in the cities his force had devastated the proportion must self-evidently be higher. Even this devastation, as the Air Ministry reminded him, had affected only 11 per cent of the whole German population, but Harris was a figure difficult to gainsay even over an issue of real strategic significance.178 He stuck to area bombing in preference to selected targets and as a result made Pointblank harder to achieve, while paradoxically contributing to the stalemate he was trying to break.

  While the two forces argued over priorities, the greatest battles of the bomber war were being played out in the skies over Germany. In all cases the majority of bombers succeeded in carrying out their bombing mission; there was never a point when the bomber did not get through. Although the bombloads were often spread over areas widely distant from the actual target, wherever they landed they did serious damage to the home population, in town or countryside, on which they fell. These aerial battles were nevertheless distinct from the bombing itself. They were fought between the enemy defensive forces – guns, night- and day-fighter aircraft, searchlights, decoys and barrage balloons – and the intruding enemy. At this stage of the war the problems facing both sides, though supported by increasingly sophisticated scientific and technical equipment and weaponry, remained the same as they had been since the onset of the bombing war in 1940: weather conditions, bombing accuracy, the balance between defensive and offensive tactics, operational organization and force morale. These factors profoundly affected what the two bomber commands were capable of achieving.

  In an age where radar aids to navigation and electronic guidance systems were in their infancy, weather continued to play an arbitrary and intrusive role in the conduct of air warfare. The final United States Bombing Survey report on weather effects regarded them as ‘a major controlling factor’ in the operation of Allied bomber forces. For the Eighth Air Force, which relied more on visual bombing, weather prevented any operations for a quarter of the time, while 10 per cent of all aircraft that did take off aborted because of weather conditions. It was calculated from unit records that the operational rate of the Eighth Air Force was only 55 per cent of the potential effort because of the effects of northern Europe’s rainy climate. Low cloud and fog were the main culprits.179 Calculations by the Army Air Forces’ Director of Weather in late 1942 found that the average number of days per month when the sky over the target was absolutely clear was one or two in the winter months, rising to a peak of seven in June, a total of 31 days in the year. Days when the major limiting factors of high wind, ice or more than 3/10ths cloud cover were absent were more numerous, but still numbered only 113 out of the year, again with a low of 6 in the winter months and a peak of 12 in June.180 The air force weather service developed a sophisticated pattern of weather prediction, based on the experience of the burgeoning civilian airline business in the 1930s, providing regular climate data, forecasts of current weather trends, and a precise operational forecast for particular missions.181 Even with reasonably accurate forecasting, weather conditions could alter rapidly and unexpectedly. The combat diary of the 305th Bombardment Group in 1942–3 can be taken as an example: 23 November, Lorient ‘covered by cloud’; 12 December, Lille ‘cloud cover at the target’; 23 January, Brest ‘obscured by the cloud cover’; 4 February, Emden ‘no bombs were dropped due to the clouds’; and so on.182

  Bomber Command was less affected by the weather because area bombing could be carried out in weather conditions that were less than ideal, but British experience also showed that ‘average good visibility’ was only available between five and nine nights in the summer months and three to five during the winter.183 To the end of the war, Harris continued to cite the weather as a principal explanation for why Bomber Command could not switch to attacks on precise target systems.184 With the arrival of electronic aids to navigation, bombing could be carried out through cloud and smoke, though the return to a base suddenly shrouded in fog caused regular accidents. Weather did make severe demands on pilots as they struggled at night to cope with the elements, as the following recollection of the last night of Operation Gomorrah illustrates:

  we set course north to the targets and flew into thunderstorms resulting in heavy icing. We could hear the ice breaking off the sides of the aeroplane and the propellers, and then we were losing power virtually on four engines with heavy icing and I then lost control of the aeroplane … It started to descend with the weight of ice and at that stage we were hit by anti-aircraft fire … The aeroplane by this stage was completely out of control with icing … a very frightening experience.

  On this occasion pilot and crew survived.185

  Of all the problems posed by the generally poor weather conditions over Germany, maintaining adequate levels of bombing accuracy was the most significant. For the Eighth Air Force clear weather was essential if precise targeting was to survive. In ideal conditions the early raids on France showed an average error range of 1,000 yards, but conditions were seldom ideal. An investigation carried out in February 1943 highlighted problems of poor accuracy and, alongside weather and enemy defences, blamed poor onboard teamwork and inexperienced pilots. It was found that the method of bombing in squadron (or Group) formation reduced accuracy substantially for the last formations to bomb; up to July 1943 an average of only 13.6 per cent of bombs fell within 1,000 feet of the aiming point, but for the last formation to bomb the figure fell to only 5 per cent.186 Anderson pressed his commanders to focus all their efforts on putting more bombs ‘on the critical points of the targets’. ‘It is evident,’ he continued, ‘that our bombing is still not up to the standards that it can be.’187 Cloud, industrial haze and smokescreens made this inevitable. The American offensive began without a Pathfinder force and no electronic aids to navigation. In March 1943 Eaker had asked Portal to supply both the Oboe and H2S equipment for use by the American formation leaders. Portal offered equipment for only eight air
craft and training facilities for the crews. By the early autumn there were three Pathfinder units forming, one with H2S, two with the American version of the technology, known as H2X.188 Eaker found the temptation irresistible to use the new guidance system to allow bombing when the target was obscured. In September 1943 the Eighth Air Force undertook its first deliberate blind-bombing attack on the German port of Emden. The bombing was scattered around Emden and the surrounding region, but the fact that it was now possible to arrive over the area of the target in poor weather introduced the American air forces to area bombing.

  Escalation in this case was dictated by the technical impossibility of bombing accurately for more than a few days a month. From September 1943 onwards American bombers were directed to attack city areas through cloud in the hope that this would hit the precise targets obscured by the elements. Sensitive to opinion, the raids on city areas were defined, like Bomber Command, as attacks on industrial centres or, increasingly, as ‘marshalling yards’.189 The distinction for most, though not all, American raiding from British area bombing was intention. The civilian and the civilian milieu were never defined as targets in their own right, even if the eventual outcome might make the distinction seem merely academic. Around three-quarters of the effort against German targets between 1943 and 1945 was carried out by ‘blind bombing’ using H2X equipment; from October to December 1943 only 20 per cent of bombs dropped using radar aids came within five miles of the aiming point, an outcome little different from Bomber Command in 1941–2.190 High levels of accuracy were reserved for primary industrial targets in good weather, where accuracy levels increased from 36 per cent within 2,000 feet in July 1943 to 62 per cent in December. American practice became one of selective precision when visibility was good, and less discriminate attacks when visual conditions were poor.191 The result, as with Bomber Command, was to increase the number of German civilians killed and houses destroyed by American bombers, which now carried much higher incendiary loads for blind-bombing raids. Even precision bombing resulted in widespread damage to the surrounding civilian area. The USAAF in-house history of radar bombing was more candid than the public image: ‘Neither visual nor radar bombing ever achieved pin-point bombing; both methods were, in effect, methods of area bombing in the sense that a certain percentage of bombs fell within an area of a certain size, the rest falling without.’192