Bombing Auschwitz: U.S. 15th Air Force and the Military Aspects of a Possible Attack

Rondall Rice. War in History. Volume 6, Issue 2. April 1999.

The ongoing historical debate surrounding the proposed bombing of Auschwitz has yet to provide a detailed analysis of 15th Air Force and its capabilities. Without clouding the argument with the political side, and by examining the status of 15th Air Force and German defences during a specific, and likely, bombing period, the facts stand clear. Fifteenth Air Force had the technical means and window of opportunity to bomb the camp with a high probability of success, and an accuracy providing an acceptable risk to the prisoners. The only limitation, and the reason the bombing was not carried out, was the lack of political will and backing.

David Wyman, Bernard Wasserstein and Martin Gilbert, among others, have provided in-depth analysis of the political aspects of why the Allies did not take military measures against the Nazi death facilities around Auschwitz, Poland. These authors, however, only skimmed the surface of the capabilities, constraints and complexities involved had the Allies decided to attack the four operating crematoriums and gas chambers at Auschwitz II (Birkenau). On the other hand, James Kitchens and Richard Foregger have presented arguments against such an attack, while ignoring the fact that, despite the difficulties, the Allies could have carried out such an attack if the political will had existed to do so. An examination of the military’s knowledge of Auschwitz’s mission, the capabilities of various bombers to bomb the killing facilities and possibilities of attacking the crematorium reveal that the heavy bombers of 15th Air Force could indeed have bombed these facilities with acceptable accuracy. However, it would have required bombing during daylight and clear weather, and the missions would have had to be flown between July and early October 1944.

Bombing Requests and Governmental Information Flow

Diversion of bombers from their priority targets would have required political intervention at the executive level. The military leaders would not, on their own, have diverted assets for what they would have viewed as a militarily non-essential mission. However, precedent demonstrates, in the September 1944 air-drop of supplies to the Warsaw resistance, that political pressure could indeed force the military to conduct such a mission, viewed as outside the scope of winning the war, as quickly as possible.

David Wyman provides excellent coverage of the relationships and discussions between the War Refugee Board (WRB) and War Department’s Operations Division (OPD). In fact, the WRB’s mission was to provide executive oversight and coordinate efforts to `rescue the victims of enemy oppression who are in imminent danger of death’. When discussing the possibility of using military forces to carry out its charge, the WRB coordinated with both the War Department’s Civil Affairs Division and the OPD, which was in charge of planning the use of military forces to carry out national objectives. Wyman and Gilbert detail the interaction between the WRB and the War Department, but missing are any earnest discussions about bombing at lower levels of command.

Had the War Department seriously studied the possibility of attacking Auschwitz, or rail lines leading to it from Hungary, during the spring and summer of 1944, it is reasonable to assume that these requests and studies would be in the records of Army Air Force (USAAF) leaders (Generals Henry `Hap’ Arnold, Carl Spaatz, James Doolittle, Nathan Twining and Ira Eaker, to name the major figures) during that time. No such records of this type exist in these generals’ official or personal collections until very late, and any records were less a request for planning proposals and more an afterthought, for information purposes only. At the very least, a serious inquiry into an important operation using air power would have reached General Arnold, since he was the Chief, US Army Air Forces. In addition, it is also reasonable to assume that Arnold would have approached his theatre commanders to ask their opinion, and to request information as to the plausibility of such attacks, available resources, resources necessary for such an attack and chances of success. Arnold remained in constant contact with the air commanders in Europe—especially with Spaatz—because of the latter’s direction of the largest and most important campaign in the young history of US air forces, the Combined Bomber Offensive. The belated information given on 4 October 1944 (as opposed to a serious request for planning action) demonstrates the lack of impetus behind the request from military and political leaders above Arnold. The lack of planning at the theatre and lower command levels indicates that the bombing of Auschwitz was never seriously contemplated or studied. However, had the political will to bomb Auschwitz existed, it would have overcome the objections of the military leaders, as the Warsaw air-drop operations demonstrated.

Polish forces in Warsaw, anticipating the arrival of the Red Army, rose against the Nazis on 1 August 1944. Without explanation, the Soviet forces halted their offensive only 10 km from the city. The personal attention of Churchill and Roosevelt ensured that the Allies attempted to provide badly needed supplies via air-drop—despite the objections of military, especially air force, leaders. The discussions, planning and stated objections of military leaders, mostly from General Carl Spaatz, generated a large amount of message traffic.

After much planning and negotiation, the missions to assist the resistance fighters began on 11 September 1944, and continued for a week, halted only by the refusal of the Soviets to allow further missions. Compared to these, the Auschwitz bombing proposals displayed a lack of urgency and political will. Still, the Warsaw operation aptly demonstrates that the military leaders, when forced to accept missions they did not agree with, would salute smartly and carry them out to the best of their ability when urged to do so by the civilian leadership.

Intelligence and Planning

In order to examine the possibility of bombing, assuming that the political will existed, and that the USAAF had been given the task of planning and carrying out such a mission, its leaders would have needed timely intelligence. First, they would have had to have known about Auschwitz, and its ghastly mission to exterminate Jews and other undesirables. Then, specific target information would become necessary to plan the attack: the specific camp location, detailed location and construction of the four crematoria/gas chamber buildings, and German defences. All of this information was available by the end of July 1944, and further detailed information would have been available.

Before assessing what information the planners required, one must examine the information available to the air commanders during the summer of 1944. From 1942 to mid-1944 the Allied governments’ information on Nazi concentration camps remained fragmentary and unconfirmed. Information relating to Auschwitz (noted in intelligence references by its Polish name, Oswiecim) centered on the I.G. Farben industrial facility, and made little reference to the camps. Also, even when intelligence sources mentioned the presence of camps, they continually referred to them as prisoner-of-war or labour camps. These references indicated that Allied intelligence did not, until 1944, know the true functions of these facilities.

One of the earliest available official reports on Auschwitz came from the Office of Strategic Services (OSS), the forerunner of the Central Intelligence Agency. The short report, dated 25 August 1942, recorded the number of inmates for Auschwitz at 15 000 and describes them as `mostly intellectuals and middle class elements’. The report did mention inmate turnover as being `very great’ and two Polish workers as having been executed. A separate report, dated 15 October 1943, provided in-depth information on the synthetic oil and rubber works in the vicinity of Auschwitz. Despite this information, very little data existed on the nearby camps.

Intelligence information on-hand in Europe available to flight crews contained much the same information as the 1943 OSS report. However, the Target Information Sheet for the I.G. Farben Complex, used for planning bombing missions, made no mention of the camps (as of 18 July 1944). The file did contain a separate amendment, dated 21 September 1944, covering the labour camp area (Auschwitz III or Monowitz). It provided a grid reference to locate the camp and said, `[it] is now known to be used as a Prisoner of War Camp’. The file also included the reconnaissance photos taken on 4 April 1944 by US aircraft, and a map used by aircrews to show distance and magnetic headings as they approached the target.

The most significant intelligence, the escapees’ reports, came to the WRB in July 1944. The WRB representative in Switzerland, Roswell McClelland, sent two messages to Washington, one to the State Department and one to the War Department (annotated `for the WRB’). Although dated 6 July, the messages arrived at their separate destinations on 8 July and 16 July respectively, according to the stamps on the original messages. Together, they supplied the information on Auschwitz I and Birkenau (Auschwitz II) from the separate escapee reports of two Slovakian Jews who had fled the camp in April, Rudolf Vrba and Alfred Wetzler, and an escaped Polish officer. These reports do not include hearsay, but record only `actual personal experiences’. In addition, McClelland wrote, `Their authenticity seems corroborated by fragmentary reports from different individuals and organizations in Switzerland which have come in the last two years, and particularly as to the composition of transports of Jewish deportees coming from everywhere in Europe.’ The second message, a summation of all three escapees’ reports, provided details on the layout, functions and operations of both Auschwitz and Birkenau.

Historians disagree on the actual date when the complete reports reached the relevant authorities. In any case, the arrival stamp on the messages displays the dates mentioned above. Therefore, had political force been behind an effort to bomb the camp, military planners should have been able to access the information from these messages not later than 18 July. Still, as Foregger and Kitchens both wrote in their separate articles, these messages contained only a summation of the reports, and not the actual reports themselves. The complete reports, and their associated maps, did not arrive in Allied hands outside Switzerland until November, as confirmed by the date on the WRB report.

It remains unclear how much and what kind of information the military commanders in Europe had about the camp complexes around Auschwitz. As mentioned above, the target information file contained none of the information in the WRB reports. Even two months after the WRB reports were in Washington, the target folder still mentioned only a prisoner-of-war camp.

Further down the command line, at the bomber squadrons themselves, crews of 8th and 15th Air Forces do not recall ever being briefed about Nazi concentration camps, their location or functions. In the present author’s interviews with four crew members from 15th AF (727th and 838th Bombardment Squadrons) and one from 8th AF (339th Bomb Squadron), none remembered knowing about the camps during the war from official sources. They all had heard of possible atrocities from open press sources, but were never officially briefed, or otherwise informed. Milton Groban, a B-17 radar navigator-bombardier, flew with 15th AF both in North Africa and Italy, and also served as a staff officer at 15th AF Headquarters in Bari, Italy. He concurred with the other crew members’ recollections, and added that information regarding treatment of Jews came only via the press. He added that personnel at the lower levels (group and squadron) took more interest in these press reports than 15th AF staff officers. Of major interest was the treatment of downed Jewish flyers (he himself being a Jewish-American airman). This concern probably explains why those who flew missions tried to keep themselves informed of Nazi treatment of Jews. The concern became a reality when Jewish flyers were ordered to alter the letter `H’ on their dog-tags (indicating Jewish religious affiliation). The easiest and least noticeable alteration was to imprint a `B’ (Baptist) over the ‘H’.

Despite the claims of Kitchens and Foregger, if Washington had ordered the USAAF to bomb the camp, the available information, when compiled, would have provided sufficient intelligence to begin planning. By pooling the WRB’s information, and the available target intelligence photos from 4 April and 26 June (and, depending when the mission planning took place, the 25 August photos), air planners would have had a significant amount of information immediately at their disposal.

Both Foregger and Kitchens alluded to the inconsistencies between the escapee reports and the reconnaissance photos. Both argued thatthe reports (and later the sketch maps) did not provide adequate information for accurate bombing. Furthermore, Kitchens asserted that only a personal debriefing of the escapees by Allied personnel could have resolved the inconsistencies between the report and the photos. This statement is far-fetched and inaccurate. Intelligence personnel would not have ignored information simply on account of small inconsistencies. Trained to know that human intelligence reports can be fallible, they could easily have compared the reports with the photos and identified the correct location of the crematorium. Had clarification become necessary, additional reconnaissance missions could have been flown.

The question now returns to the `diversion of assets’ so commonly given as the rationale of the War Department’s refusal to bomb. While this phrase usually refers only to diversion of aircraft from military targets, the full impact of the effort that would have been required to plan the attack must also be considered. Kitchens correctly stated that without the proper information at hand, military personnel in-theatre, and even military and civilian workers in Washington and London, would have been diverted from their normal war tasks to obtain the required intelligence. Such an effort would have required at least staff personnel, intelligence analysts and photo interpreters, and probably additional reconnaissance sorties. Intelligence resources, especially photo analysts, remained a constant priority throughout the war. Spaatz wrote to Arnold in March 1944 confirming that, in both the north-west Europe and the Mediterranean theatres, unless more intelligence, planning and communications officers arrived, `our air forces will suffer’.

Kitchens also wrote that the necessary intelligence-gathering, planning and bombing could not have been done in time to save the deported Hungarian Jews. Even though he did not consider the possibility that, with the proper authorities directing such events, this mission could have taken a higher priority than other tasks, the mass of Hungarian Jewry could probably not have been saved. The final mass deportation of Hungarian Jews took place on 9 July, while the crematorium continued to operate against the remainder of Auschwitz’s large population. The German Central Construction Office at Auschwitz estimated that crematoria II-V could incinerate 4415 bodies each day, with two bodies burned in each oven for 30 minutes. Desiring to increase their capacity, camp authorities ordered that the time be reduced to 20 minutes, and up to three bodies burnt in each, thus almost doubling their capability, to 8000 bodies per 24 hours. While perhaps unable to save the mass of Hungarian Jewry (Yehuda Bauer stated that the Nazis gassed approximately 75 per cent of them upon arrival), bombing any time before November, when gassing operations ceased, carried the possibility of saving many lives. (Bombing accuracy and probability of damage will be discussed later.) The population of the Auschwitz camps fluctuated more than usual between the summer and autumn of 1944, due to killings, inflows from other areas and camps and the beginning of the westward evacuation. Yahil’s sources put the 21 August population at 105 168 in all of Auschwitz’s areas. By mid-October, the number had dropped to around 95 000. At the final roll-call on 17 January 1945, the number stood at 66020 inmates.

Pressure from either Roosevelt or Churchill could have ensured that the effort received the proper priority and was conducted in a timely manner. Without a doubt, to compile the necessary intelligence and plan the mission would have diverted work and assets being used to prosecute the air war against Germany. The full extent of such a diversion cannot be accurately measured, but would probably have been minimal in relation to the overall outcome of the war effort.

So far, examination of available evidence suggests that the information necessary to plan an attack on the Auschwitz killing facilities existed, or could have been obtained. Had such an attack been ordered, the task would most likely have fallen to 15th AF. By operating from air bases in Italy, it was within range to bomb Auschwitz, and it conducted attacks against strategic targets in and around Upper Silesia. Eighth AF, operating from England, did not have the range to reach Auschwitz. For this reason, historians have mainly focused on the possibility of using 15th AF.

Fifteenth Air Force: Capabilities and Missions

Despite being a new command and heavily involved in the Combined Bomber Offensive, by the late summer and early fall of 1944 15th Air Force contained the necessary assets to conduct an attack on Auschwitz. In addition, its operational doctrine contained the necessary flexibility to allow air commanders—even without political pressure—to order the attack. When looking at operational directives and capabilities, the timing of planning an attack on Auschwitz must be kept in mind. As shown above, the required intelligence was not available before late June 1944. However, by July and August enough information was available (or could have been) to plan and carry out an attack. These two months were the optimum time to attack, and offered the best chance to save the Hungarian Jews. While attacks could have taken place from September to December, weather factors decreased the possibility of success. Therefore, this study pays maximum attention to the late summer months of 1944.

The original plans called for 15th AF to fill out its combat complement of units by March 1944. By June 1944, 15th AF controlled five bomb wings (BW) and one fighter wing. The 5th BW included six groups of B-17 Flying Fortresses. The remaining wings (47th, 304th, 55th and 49th) operated B-24 Liberators. All B-24 wings contained four groups except the 49th, which had only three.

Although smaller than 8th AF, the 15th AF represented a sizeable force. On 7 June, 15th AF reported 1146 heavy bombers on hand, compared with 8th AF’s 2786. Of 15th’s total force, 63 per cent (720 bombers) were operational. From 7 June until 30 September, 15th AF averaged 1204 bombers on hand each day. Of these, an average of 914, 76 per cent, were reported as ready for duty. Compared to 8th AF during the same period, 15th AF averaged fewer than half as many bombers on hand, but their operational rate was 12 per cent higher. These numbers also reflect an acute shortage of replacement B-17s during the summer of 1944, due to increased B-29 production at the expense of the Fortress.

With such a significant force available, the next question follows: what were they doing? Specifically, the tasking and bombing directives require examination. At the Combined Bomber Offensive’s inception, submarine bases topped the list of priority targets, due to the Allied concerns over German interdiction efforts in the battle of the Atlantic. Early in 1944, however, attention turned to preparations for Operation `Overlord’, the D-Day invasion of Fortress Europe, and concern about the strength of the German Air Force (GAF). The 13 February 1944 directive from the Combined Chiefs of Staff (CCS) formally placed the Luftwaffe at the top of the bombing list. The first priority became fighter airframe and component production, and installations supporting the German fighter force followed.

As the German fighter threat diminished in the spring and early summer of 1944, Lieutenant-General Carl Spaatz, then the US Strategic Air Force’s (USSTAF) Commander, viewed oil as the most promising strategic objective. The USSTAF commanded all US strategic bomber forces in Europe, then divided into the 8th and 15th Air Forces. On 8 June 1944, only two days after the Normandy landings, Spaatz ordered that the USSTAF strategic aim be shifted to denying oil to the enemy. The order remained in force for the war’s duration. As a result of this order, the Mediterranean Allied Air Force (MAAF) sent a revised bombing directive to its units on 15 June 1944. The information flow for `Pointblank’ (Combined Bomber offensive’s codename) targets flowed from USSTAF, to the MAAF, to the Mediterranean Allied Strategic Air Forces (MASAF), to 15th Air Force. The directive gave special missions, as required and directed by the commander, to support the land campaign in Italy and air support for `Overlord’ as the first two target priorities. The third priority, `Pointblank’ targets, had oil as the primary objective, followed by counter-air and communications targets.

Eighteen days after the 15th July bombing directive became effective, MAAF rescinded it and ordered new priorities. The 3 August directive kept oil as the top priority (Auschwitz’s oil facilities became target number 9), but remaining sources of ball and roller bearings moved to second, and counter-air came third. Although the wording regarding special targets as the top two priorities no longer appeared, the directive allowed for variations from the target list. If necessary, tactical uses of strategic bombers to aid the ground forces could replace the strategic priorities. Also, the directive stated, ‘specific attacks against targets in the above categories or any others may be ordered by this Headquarters from time to time’. Throughout the remainder of August and September, only minor changes occurred in this list. Spaatz, then the Commander, issued a clarification of target priorities on 1 September. He urged using the remaining good weather to intensify attacks against oil targets, and he changed the second priority from bearings to rocket—and jet propelled fighters and aircraft engine factories. A new bombing directive on 13 September duplicated the 3 August directive but added communications targets as the fourth priority.

The evidence presented clearly shows the capability, in aircraft and in command discretion within the target priorities, of 15th AF to bomb Auschwitz. By the summer of 1944 the command controlled ample aircraft, and those aircraft had sufficient range and payloads necessary for such a mission. Also, the bombing directives allowed the flexibility for commanders to direct attacks against special targets.

Analysis of German Defences

By the late summer and early fall of 1944 German defences, like the Nazi military machine overall, displayed signs of weakness but still represented a sizeable threat. The threat to any bombing of Auschwitz came from Luftwaffe fighters along the route, and anti-aircraft guns at the target. Although the Germans could have met an attack on Auschwitz, this threat was not singly enough to deter air commanders from planning such an attack. Also, as analysis of similar bombing missions shows, Allied losses would not have been significant.

Those who support the non-bombing, most notably Kitchens and Foregger, tend to overestimate the German air defences in 1944, while those critics of the failure to bomb the crematoria, led by Wyman, use an overly simplistic analysis. The Germans, like all Second World War combatants, used two main systems for air defence: fighter aircraft and anti-aircraft guns. On the Nazi side, the former was in decline by the summer and fall of 1944, while the latter became the most effective. Each will be examined, in turn, by looking at the threat perceived by planners, as well as their actual effect on bombing missions on and around Auschwitz.

Writers supporting a bombing effort correctly point out that the USAAF controlled the skies of Europe by mid-1944. However, they wrongly characterize the notion of `air superiority’. While Allied aircraft could fly over the Continent with a higher degree of safety than in 1943, this did not translate into being able to fly with impunity. German fighters still represented a threat and continued to attack Allied aircraft. However, aware of dwindling resources, especially fuel, they marshalled their forces and attacked only when they saw a possible advantage or when targets they viewed as essential for defence became threatened.

Spaatz, in his weekly summaries to Arnold, mentioned the GAF tactics on two separate occasions in early and mid-1944. On 4 March he wrote, `The enemy has been unable or unwilling to send his fighters in strength against [operations of the past week]. [I] believe he will withhold them except when weather conditions over Germany indicate good visual bombings over vital areas.’ Spaatz summarized the enemy fighter tactics in his 8 May message. `German fighter forces’, he wrote, `can no longer oppose our attacks indiscriminately but must, for purposes of conservation, select occasions when they have a tactical advantage.’ Therefore, the time-frame and area of attack must be analyzed to make possible an accurate analysis of German fighter opposition. Among other duties, estimating and reporting on enemy air activity within 15th AF fell to the Assistant Chief of Staff for Intelligence (designated A-2). The A-2 staffs produced several reports on enemy fighter defences for the period in question. A 10 July report analyzed enemy fighter opposition to the Allies’ 7 July attack on Blechhammer (another important oil target located just over 40 miles northwest of Auschwitz, and contained a sub-camp of Auschwitz providing slave labour). The Blechhammer attack represented a concerted effort of 8th and 15th AF to attack oil targets in the region. The total effort sent more than 1000 B-17s and B-24s, with German fighters downing only 25 bombers. Due to the importance of the targets, the Germans sent at least 400 sorties against the formations (approximately 300 aircraft, but some landing and attacking again). The intelligence report indicated 225 fighters assembled along the Vienna-Gyor line (Gyor is approximately 70 miles east of Vienna, or half-way between Vienna and Budapest). From Vienna to the target area, between 100 and 125 more aircraft attacked the formations. Finally, 40 to 50 fighters, flying for the second time against the mission, intercepted the bombers near Gyor. Although a significant effort by the Luftwaffe, it did not react in the same way when bombers attacked similar targets in the same area only six weeks later.

The first attack on the Auschwitz oil refinery occurred on 20 August 1944 as part of a larger attack on Polish and Czechoslovakian oil facilities. The attack plan called for five of the six B-17 groups of the 5th BW to attack Auschwitz with 28 bombers each, for a total of 140. In addition, 10 of the 12 groups of B-24s in the 47th, 304th, and 55th bomb wings would target the refineries at Szolnok (Hungary), Dubova (Czechoslovakia) and Czechowice (Poland). The three B-24 groups of the 49th BW would attack Szolnok-Rakoczifalva airfield. The plan included using five groups of fighters from the 306th Wing for escort. The intelligence annexe estimated that the enemy could oppose the attack with up to 110 sorties in the Budapest-Vienna area, and could reinforce the effort with 30-35 fighters from the Munich area. In addition, target area opposition could include 50-60 fighters, at most. Milton Groban, one of the mission’s radar-bombardiers, wrote in his diary, `we were expecting plenty of Luftwaffe because we were going into their stronghold.’ However, on 20 August, the Luftwaffe disproved his and 15th AF’s expectations.

The 15th AF A-2 report on the attacks called enemy opposition to the 20 August attacks `meager’. After analyzing the separate reports from all fighter and bomber groups, the A-2 concluded that the formations saw only 35 to 45 enemy aircraft. Of these, 15 to 25 made a single-pass attack at the bombers as they passed east of Budapest, and the escorting fighters drove them away. Only two Me-109s attacked near the target areas. They targeted a straggler, but did not succeed in bringing the aircraft down. All bombers from the 5th BW returned to their base, and the remainder of the bombers reported no losses to enemy fighters. In fact, the 97th BG claimed its bomber crews shot down one Me-109, a second Me-109 kill was reported as probable and a third one was damaged.

The decline in enemy activity from the 7 July Blechhammer attack to the 20 August attacks around Auschwitz should not be interpreted as a trend. Only two days after the Auschwitz oil attacks, 15th AF again attacked targets in Vienna and Blechhammer, drawing a reaction from 150 enemy aircraft. Also, Auschwitz was attacked again on 13 September when the 55th BW tasked a `normal effort’ for all four of its B-24 groups against the oil refinery. Of the 115 bombers tasked, 101 arrived over the target area. Two of the groups reported seeing no enemy fighters, one reported seeing nine (six over the target area) and the other reported seeing three. None of the fighters engaged the bombers. The mission and intelligence reports correspond with aircrew recollections. All six crew members interviewed said that, although the Luftwaffe was not as aggressive and feared as they were prior to 1944, they remained a threat to be respected. In addition, all agreed that flak was the primary danger to the bombers. While the Luftwaffe threat declined as 1944 wore on, the flak became more numerous and `damned accurate’.

The revised A-2 Intelligence Plan assumed that, due to the declining capacity of the Luftwaffe, the Nazis would utilize anti-aircraft (AA) guns to their maximum capacity. A-2 also believed that the concentration of guns could increase due to a fairly steady availability of AA guns concentrated within the Nazis’ constantly shrinking territory. The document also noted that, with the Third Reich trying to defend its precious oil supplies and production capabilities, these areas contained some of the heaviest AA defences. At the time of the report, Blechhammer was the eighth most heavily defended target in Europe.

Flak defences around the Auschwitz I.G. Farben facility totalled 316 heavy guns in late August. The 20 August tasking order’s intelligence annexe includes those guns around Vitkovice (35 miles south-west), Crakow (20 miles east), and Katowice-Gleiwitz (15 miles west) as being target area defences. These guns could target aircraft during their bombing run, during their time in the target area or during their post-strike recovery prior to exiting the target area.

The mission reports, however, revealed that the flak defences did not take a toll on the attackers. Two of the groups (99th and 2nd) reported the defensive fire as accurate, but only moderate in intensity. The remaining three groups of B-17s called the flak inaccurate and moderate. The difference in reporting does not represent inaccuracies by one or the other group. As in all large-scale bombing missions, the formations of aircraft could extend miles on either side, and appeared over the target area at different times. The groups arrived over the target separately, but all dropped their ordnance between 1030 and 1100 hours. The flak downed no aircraft, but 45 B-17s received minor damage, and one was severely damaged. In addition, flak wounded one man in the leg.

Less than one month later, 15th AF’s visit would meet with a different fate from the Germans’ guns. Although not providing a specific count, the 13 September’s pre-strike intelligence annexe specified that `heavy guns have been added since [the 20 August attack]’. All four B-24 groups reported the target area flak as accurate and intense. Of the 101 aircraft flying over the target area, nine did not return to Italy. Four aircraft went down in the target area, four more made forced landings in Yugoslavia, two headed towards Russian territory and one ditched in the Adriatic. In addition, eight B-24s recovered with severe flak damage, while 44 more reported minor damage, and two crewmen were injured. Only 37 of the 101 aircraft returned unscathed.

The crematoria at Auschwitz-Birkenau were slightly more than four miles from the oil facility. Although camp commandant Rudolf Hoss and the escapees from Auschwitz all mentioned AA guns being in the camp, the distance between Birkenau and the refinery would still allow the guns protecting the oil facility to engage bombers over Auschwitz. German flak guns ranged from 88 to 128 mm. The most numerous was the 88-mm Flak-36. Its maximum ranges were 14860 metres (9.2 miles) horizontal, and 10600 metres (6.6 miles) vertical. Still, one must consider whether German gunners would have targeted the bombers when they were not attacking the oil refinery. It would be reasonable to assume that, if they did, because of the increased range from their positions, the damage would not have compared to the casualties suffered on 13 September.

Clearly, neither the fighter nor flak threat represented such a danger to a proposed attack against the Birkenau crematorium that air leaders would have argued against it. Bombing accuracy, however, presented a larger threat—not to the bombing forces, but to those the bombing would have attempted to save.

Assessing the Target for Attack

Before analysing accuracy, the weight of effort required to destroy the gas chambers and crematoria requires examination. Weaponeering, a term used in today’s target analysis, defines a target’s characteristics and the number of bombs and aircraft needed to achieve probable destruction. Weaponeering will also provide insight into how much of a `diversion of resources’ the attack would have demanded. Selecting and defining the target is the first step in weaponeering, followed by examination of target construction. From the type and construction, the targeting officer selects the type and number of weapons best suited to destroy the target. The number of bombs needed determines the number of aircraft required to carry them. The expected damage is expressed in a `probability of damage’.

By 1944, Auschwitz II-Birkenau contained all of the operational gas chambers and crematoria. Crematorium I at the main camp (Auschwitz I), and its attached gas chamber, ceased operation in December 1942. In 1942 two small cottages near Birkenau, called Bunkers One and Two by the SS, took over the majority of the gassing operations. These small brick buildings could not have been targeted. Surrounded by woods, Bunker One measured only 49 x 21 ft and Bunker Two, 56 x 27 ft. In addition, intelligence sources did not know of these buildings’ existence. Being converted houses separate from the main area, they would have received little attention from photographic analysts unless other sources revealed their purpose. Neither Vrba or Wetzler knew of their operation, and their report identifies only the four large crematoria at Birkenau as being in operation.

Birkenau’s four main killing facilities were located on the camp’s west side. Numbers II and III, built to almost identical specifications, contained three parts, two of them underground. The `dressing-room’, where victims disrobed, measured 162 x 26 ft and an outside entrance gave access underground. The gas chamber, also underground, measured 99 x 23 ft. Being underground did not necessarily hide their existence. If a photo technician specifically looked at Birkenau for prestrike reconnaissance, it is possible he would have seen these. Combining the known intelligence from the escapees’ reports—therefore knowing of the existence of these facilities—a photo-interpreter could have located them. Examination of aerial photographs using modern techniques shows the outline of both, and the gas vents of the killing chamber. The above-ground crematorium buildings, 99 x 37 ft, would serve as the aiming-point. These calculations, taken from the original plans, differ slightly from those in Foregger’s article. Foregger used figures from a study by Dino Brugioni and Robert Poirier (CIA analysts who, in 1979 published the reconnaissance photos and explanations). These measurements came from analysis of the aerial photographs. Obviously, personnel of the day could not access the original plans. Therefore, for weaponeering purposes this study will also utilize Foregger’s figures for facilities II and III. He used 352 x 41 ft as the overall size of each building complex. Like the first two, Crematoria IV and V, located 1746 feet from III, represented a `matched set’. Their complexes measured 220 x 42 ft (interestingly, here Foregger’s figures match the plans).

Foregger used wall and roof thickness to determine bombs and fusing. Assuming that any planning effort would have used the Vrba-Wetzler report and reconnaissance photographs, it is doubtful that planners would have known these precise elements of construction. However, Foregger’s analysis finds that this construction required 500lb general purpose bombs with delayed fuses. The 500-pounder was the standard weapon used by the USAAF, and would probably have been the weapon chosen anyway. The delayed fusing allowed for penetration before exploding. However, even an impact on the crematorium or the above-ground vent area without delayed fusing would have caused damage. In addition, Groban concluded that, for a target partially above and partially below ground, ordnance crews would set a mixed fusing. Some bombs would detonate on impact, and others would allow some penetration.

Foregger’s calculations (of which he gives no details) for facilities II and III concluded that these two buildings would require 400 bombs for a 64 per cent chance of damaging each complex. Half as many bombs would provide a 50 per cent chance. Facilities W and V, being smaller, gave only a 60 per cent probability with 400 bombs, and 39 per cent with 200, for each complex. A typical heavy bomber carried 10 bombs (as did the raids against the Auschwitz refinery). Therefore, it would require 160 bombers to obtain at least a 60 per cent chance of damaging all four buildings, As a comparison, the September raid on the I.G. Farben oil refinery tasked 115 B-24s, and only 101 reached the target. The tasking represented a normal effort for all four groups of the 55th BW.

A comparison can be drawn, again using the oil-refinery attacks as a reference for bombing effort. If the 55th BW used 28 bombers from each of the four groups, each group could target one of the facilities. Only 280 bombs would be targeted for each facility, giving a less than 50 per cent chance of a hit.

Since Foregger did not provide details on his calculations, an independent study was done using three different methods. The first two methods will use the figures for bombing probabilities from the Handbook for Bombardiers (the same book Foregger used). A final method will examine numbers produced by modern computer-based software using ballistics for the standard 500-lb bomb.

To weaponeer completely a target from scratch is a very detailed process. Figures and formulas for ballistic error, bomb and fuse reliability, stick length (distance of bomb impacts on the ground), intervalometer settings (time between release of bombs) and slant range (angular distance from bomber to target) all affect bombing probabilities. From these numbers, a Range Error Probable and Deflection Error Probable can be calculated and assist in finding the Single-Shot Probability (SSP). The SSP gives the number of individually sighted and released bombs necessary to achieve at least one hit with a prescribed degree of assurance. Fortunately, the Handbook for Bombardiers provided tables whereby, if certain elements were known, others could be derived from charts.

The first method begins by selecting the probable range and deflection errors. Using a 20000-ft bombing altitude, the range and deflection errors probable are 201 and 250. Next, the Vulnerability factor (Vf) for each is determined. Using the dimensions of crematoria I and II as the allowable errors (since any error beyond these would miss the buildings), the range Vf = 0.20 which translates to a Range Single-Shot Probability (RSSP) of 0.1073. The deflection Vf = 1.408 gives a 0.6584 Deflection Single-Shot Probability (DSSP). Multiplying these two figures gives a SSP = 0.07 for the first two targets. Using the same methodology for crematoria IV and V calculates to an SSP of 0.05 (smaller due to smaller buildings).

Using the smaller SSP, the charts show that if for each ten bombs dropped (one bomber’s load), there existed a 40.131 per cent chance of hitting the target one or more times. Furthermore, 105 bombs would give a 99.5 per cent chance of at least one hit on one building. If one group of heavy bombers attacked each building with 25 bombers each (250 bombs), there was a 99.5 per cent chance that five or more bombs would strike each building.

These figures probably represent high numbers. The original range and deflection errors probable (from table I in the Handbook), when calculated to a Circular Error Probable (CEP), come to a 392-ft radius—far better than any Second World War heavy bomber. Therefore, a second method using the charts must be calculated.

Circular-error probable (CEP) describes the probability of one-half of the bombs falling with a circle of designated radius. During August and September 1944, the CEP of all aimed bombs in 15th AF was 800 ft. Using the width of the smaller crematoria (220 ft), the target radius would be 110 ft. Again using the Handbook, these figures correspond to a Vf = 0.1375 and a SSP of less than 0.01. This SSP corresponds to the third method’s findings.

The final method involved using current targeting software used a computer program to check the SSP. The program would be valid for such computations, since the 500-lb bomb’s ballistics remain virtually unchanged. Using various intervalometer settings and altitudes between 18000 and 20000 ft, the SSP computed to 0.01, the lowest output available. Using the 0.01 SSP, the Handbook’s charts show that every 10 bombs, or one fully loaded bomber, will give a 9.56 per cent chance of hitting the target at least once. Also, 220 bombs on each crematorium (or 22 bombers—less than the normal 15th AF method of using between 25 and 28 bombers per group per attack) would give a 90 per cent chance of one hit or more.

These calculations consider bombing accuracy in terms of the probability of hitting the target. Many other factors must be considered, including bomber methods and tactics. Different tactics and altitudes could increase the probability of success and decrease the number of bombs possibly striking the barracks areas. Therefore, a further analysis of accuracy and probability for collateral damage remains necessary.

Bombing Methods, Tactics, and Accuracy

The importance of bombing accuracy has played a central role in the historical controversy surrounding any bombing attempt on Auschwitz. With the noted inaccuracy of Second World War bombing, critics discount the idea that bombing would have saved prisoners because Allied bombs would have decimated the camp’s occupants. However, close analysis of 15th Air Forces bombing during the period from August to September 1944 displays a high degree of accuracy—by Second World War standards. In addition, by using a three bomber front, and only bombing during clear weather with each aircraft’s bombardier acquiring the target, Auschwitz’s killing machinery could have been hit with a high degree of accuracy while limiting the danger to the prisoners.

What military personnel called `precision bombing’ during the Second World War bears no resemblance to what is meant by that term today. Instead, it meant that a specific target was being attacked rather than an area target. The precision arose from planning to bomb a specific target, and selecting a desired Mean Point of Impact (MPI). The alternative was area bombing, or `carpet bombing,’ in which bombardiers did not select specific targets but dropped their bombs on geographical areas.

For the most part, RAF Bomber Command conducted the Combined Bomber Offensive’s area night bombing while the US strategic force practised daylight, precision bombing. The RAF experimented with attacking individual targets in 1940, but believed the effort too costly and ineffective. Believing that the bombing technology and intelligence lagged too far behind bomber capabilities to be of use, the RAF opted for attacking German cities and their large industrial areas. The USAAF entered the war believing in the ability of a heavily armed bomber to attack specific installations with `pinpoint’ and `pickle-barrel’ accuracy. To achieve such accuracy, the attacks had to take place in daylight. However, during battle conditions, and because of technological limitations, such accuracy did not become commonplace. The USAAF adopted a `target area’ to define accuracy. It considered bombs falling within a circle with a radius of 1000 feet around the aiming-point, or MPI, as being within the target area. As will be seen, the daily and weekly accuracy reports used this circle to establish accuracy.

One important factor requires clarification. Bombing accuracy statistics and analysis beg examination only because of the proximity of the prisoner barracks to the gas chambers and crematoria. Had the killing facilities been detached from the camp, large-scale bombing could have been accomplished without considering the deaths of those whom the bombing attempted to save. Again, reconnaissance photographs and the escapees’ reports would have shown planners the necessity of minimizing bomb deflection errors. This proximity would have meant, not only that the accuracy of the bombers required examination, but that certain methods of bombing had to be excluded because of their gross inaccuracy.

`Blind bombing’ meant bombing without the bombardier visually acquiring (i.e. seeing, positively identifying and specifically targeting) the target or any other reference point. Usually, this meant bombing from above cloud cover, but the increase in the Germans’ use of smoke generators created additional target acquisition problems. These generators were used during the August and September attacks on the Auschwitz refinery, and achieved their desired effect. Not only did the smoke partially obscure the target during bombing, but four of the five groups on the first mission could not report accuracy results. Aware of European weather conditions, the military looked for ways to improve upon bombing accuracy when weather (or smoke) obscured the target. As earlier missions against I.G. Farben proved, smoke generators could affect bombing accuracy. However, the distance between Birkenau and the oil plant meant that it was doubtful that any smoke screen could have obscured the camp.

The Massachusetts Institute of Technology’s Radiation Laboratory improved on an RAF radar-bombing system, designated H2S. Eighth AF first used the American version, H2X, in November 1943. The system (also known as AN/APS-15, Pathfinder, PFF or `Mickey’) became a valuable tool, but did not provide pin-point bombing in limited visibility. Both the August and September attacks against the Auschwitz refinery used the Pathfinder as a bombing aid. Milton Groban was closely associated with the Pathfinder induction into 15th AF and training of its crews. He wrote a training manual on the system, Bombs through the Undercast. The book, and the training he directed, proved so valuable that he received the Bronze Star medal for his efforts. Although, the H2X system aided navigation and allowed the bombers to bomb with improved confidence, it remained unreliable. Using the system effectively required additional training, intelligence support, and operator proficiency.

One way to train the operators included using `pictures’ taken by the scopes of targets, and having the operators identify the target. Radar methods, being in their infancy in the Second World War, represented crude images compared to today’s systems. To improve on accuracy, the air forces developed a system called PFF-Synchronous bombing. The system required close communication between the radar-bombardier and a visual bombardier, and was usually used when broken clouds or smoke partially obscured the targets. The visual data, being more reliable, supplemented Pathfinder information. The PFF-Synchronous method improved in accuracy as the cloud cover diminished, hence additional visual updates.

Visual bombing remained the most accurate way to bomb. Radar bombing could not have been used to attack a target such as the facilities at Auschwitz-Birkenau. A simple examination of the radar scopeview of the Auschwitz area proves this point. This mission’s inherent concern for collateral damage to the prisoners meant that the bombers would have relied on the most accurate methods. Therefore, the attack would have required good weather.

Bombing tactics greatly affect bombing accuracy. Bomber formations varied from three- to 12-bomber fronts. Target nature and size, among other factors, determined the formation employed. Using the dimensions and length of the bomb run, the axis of approach and target line presented a target area approximately 350 ft wide and 2800 ft long, perpendicular to the buildings’ lengths. The three-aircraft front would not only have minimized the width of the bombs possibly falling on prisoners’ barracks, but provided the best accuracy for smaller targets. Operational analysis demonstrated that compared to a nine-aircraft front, a three-bomber front would place 46 per cent more bombs against a 400-ft-wide target. Also, crossing the targets did not drastically decrease accuracy. For a 400-ft-wide target, attacking down the length improved accuracy by only five bombs per 400 dropped. Also, the three-wide formation provided the smallest pattern width, 880 feet. Because of the size of the oil refinery, most formations used a six-aircraft front. However, target size and the importance of bomb width dispersion would dictate a smaller, three-front formation when attacking Birkenau.

So far the evidence demonstrates that any attack against the Birkenau killing facilities would have required a daylight clear-weather attack, by a small-front bomber formation, attacking perpendicular to the target. Although representing the best available tactics, these measures would not have guaranteed that the barracks area would not receive damage. Analysing the 15th AF bombing statistics from the summer and fall of 1944 provides the best means of anticipating collateral damage.

The following statistics do not represent a circular-error probable (CEP). Kitchens used CEP, and this can cause confusion in the statistical analysis when comparing and measuring accuracy. Therefore, this study will adopt the direct percentage method used by 15th AF. If 15th AF quotes an accuracy rate of 40 per cent for a 1000-ft circle, then out of 100 bombs dropped, 40 landed within 1000 ft of the MPI.

From July to December 1944, 15th AF averaged putting 40.1 per cent of its bombs within 1000 ft of the MPI. In addition, the best months were July to September, with 41.7 per cent accuracy. The best wing, 47th BW, placed 49 per cent within the 1000-ft circle; the worst wing, 304th BW, produced only a 22.4 per cent showing. The only period exceeding the accuracy rate for this quarter was March 1945, when 15th AF placed 51.6 per cent of its bombs within 1000 ft.

Drawing the circle even smaller, to 600 ft, the command put 20.9 per cent of its bombs `on target’ between July and September. In the smaller target area, the 55th BW surpassed the 47th BW in accuracy by 0.2 per cent with 25.2 per cent. As with the 1000-ft circle statistics, only March 1945, with 28.8 per cent, surpassed the July-September period. The high accuracy rates correspond to 15th AF’s most intense flight activity. Bombers flew 20 845 sorties in August 1944, and posted 15th’s highest effective sortie rate (percentage of bombers airborne to those arriving in the target area and dropping their ordnance), 88 per cent.

A combination of factors, including the aforementioned dramatic decline in German air defences and effectiveness and favourable European weather conditions, could have aided bombing accuracy. The relatively stable and complete manning of 15th AF aircrews also warrants mention. June 1944 to November 1994 represents 15th AF’s most stable wartime aircrew manning. Prior to June, the command was still in the process of receiving the manpower to bring it to its authorized strength. After November, owing to the reduction in combat losses and turnover, 15th AF had a surplus of aircrews that actually contributed to its decreased bombing accuracy, due to inadequate training and experience.

Overlaying 600- and 1000-ft circles over a diagram of Birkenau aids in demonstrating these accuracy statistics when applied to the camp. The 600-foot radius does not include any additional buildings, and the 1000 foot circle adds nine barracks in Block BIIf, and one-third of the `Canada’ loot storage facility. Using the earlier computations requiring 400 bombs to have a 60-64 per cent chance of hitting the target, this means that, between July and September 1944, 233 bombs would have fallen somewhere outside 1000 ft. If 15th AF had sent one bomb wing to attack Birkenau, with each group targeting one building, then 163 bombs would have fallen outside the circle.

These statistics do not tell the whole story. Using these alone (as Kitchens and Foregger did), one could argue that bombing could decimate the camp’s housing areas. But one must recall the statistics presented earlier, and overlay a three-bomber pattern width of 880 ft (this also came from August and September 1944 mission statistics). Overlaying this width on the diagram demonstrates a greatly reduced threat of bombs devastating the camp.

Also, by analysing other attacks’ bomb plots, the majority of the stray bombs fell along the flight path. In other words, they impacted long or short of the target rather than at great widths from the target.

These statistics demonstrate the feasibility and probability of success while limiting the amount of danger to the inmates. Still, these are only statistics. Any attack on the camp carried with it a probability of killing inmates who might have survived the war (and in some cases did). Only a minor error in deflection aiming (in this case, aiming bombs to the right of the intended target) could lay the pattern of bombs down the centre of the camp. Even the tightest pattern would allow stray bombs that could hit the camp. Also, even with a good run described above, the `Canada’ storage area would have been in danger. Many prisoners worked there, and the area, although dangerous, provided inmates with the best chance of securing food and of bartering items of any work detail.

Above all, however, the four buildings at Birkenau represented the largest and most apparent danger to all inmates, and to people still arriving. Kitchens and Foregger point out that the bombing would also take out the rail link where, on a day-to-day basis, a large concentration of prisoners arrived. However, the majority of those arriving, like the Hungarian Jews, went straight to the gas chambers. In addition, these represented the perceived weakest and those not fit for work (young, old, and many women). Although the survival rate for the remainder was low, it was higher than the survival rate of the gas chambers.


With a political push for action, the air leaders would have carried out the attack. Certainly any effort represented a `diversion of assets’, but such diversions were not unprecedented. As shown, the total effort would have required one mission by one group of heavy bombers. In order to minimize collateral damage to prisoners, Birkenau could not have been attacked at night or in bad weather. However, any targets it replaced could be attacked in less than optimum weather with systems such as Pathfinder.

Authors who have considered the military aspects of a possible attack tend to generalize all aspects of any such operation. Kitchens and Foregger use broad representations and generalities to portray the mission as impossible or foolhardy. They used bomber accuracy statistics and information on German defences from various dates and regions. However, as this study has demonstrated, analysing the situation under specific probable circumstances provides the most effective means of defining the possibilities and characteristics of any possible attack.

When narrowed to between July and October 1944, the historical record reveals that German defences rarely took a heavy toll on attacking bombers, and strategic bombardment accuracy achieved a high level. During these months the Luftwaffe mustered its resources, yet did not affect the bomber missions against the Upper Silesian oil facilities. German flak guns, while ever-present and with improving accuracy, did not significantly reduce attacking formations, and these months also saw the second highest level of 15th AF’s strategic bombing accuracy. The command’s statistical analysis unit detailed how lower altitudes and varying formations could offer increased precision.

Regardless of the bombing platform used or the statistical accuracy then available, there existed a chance that the attack would kill camp inmates. In order to minimize the collateral damage, the military leaders would have had to pick the best available asset under the existing time constraints. Using 15th Air Force would have been the easiest means to attack the gas chambers and crematoria, but it also would have required a large effort with questionable accuracy.

If 15th AF had been selected, the mission could have only been performed within a rigid set of circumstances to limit the possibility of decimating the camp. The bombers would have had to use a three-bomber front, with each bombardier acquiring the target, and under clear weather. Only these tactics and conditions could have helped limit stray and inaccurate bombs from hitting the barrack areas. Under these criteria, and using 15th AF’s statistical accuracy for August and September 1944, the Allies stood a very good chance of destroying or damaging the Birkenau facilities while limiting the possibility of killing prisoners.

Without decisive political intervention to attack the crematoria (nonmilitary, but essential targets), Allied military leaders continued with their plans to achieve the Third Reich’s destruction by the best and fastest means possible. Also, the German reaction to either temporary or permanent inability to use their high-capacity industrial killing machinery cannot be accurately gauged. Their reaction could have included using earlier but less efficient methods requiring additional manpower—trenches and firing squads. Or, with the ever-closer approach of the Red Army, they might have begun the death marches back to the Reich a few months earlier. Whereas bombing offered only a possible reprieve for camp prisoners, destroying Hitler’s grip on Europe guaranteed a means for saving the remaining inmates.