Blackett's War (44 page)

EVEN BEFORE
the recommendations of the Convoy Conference and the reorganization of the U.S. antisubmarine command structure could be implemented, a remarkable turnaround in fortunes had occurred in the Atlantic. On March 18, 1943, after two weeks of frantic efforts, the Bletchley code breakers resumed reading the U-boat Enigma signals by employing their longest shot yet; piecing together cribs from short-signal messages transmitted by the U-boats to report convoy sightings or their position or fuel supplies, they managed to break back into the system, in
many cases by making truly inspired guesses at the contents of the messages to build up a crib to run on the bombes. Signals that reported a U-boat’s location might be cribbed by direction-finding fixes on the boat that independently established its longitude and latitude; the signature within the signals that gave the boat’s number was in many cases cribbed with help from the Operational Intelligence Centre’s tracking files and by putting together a catalogue of each boat’s individual radio fingerprint—the background radio noise unique to individual transmitters and the slight individual eccentricities in each radio operator’s “fist,” the way he varied the length of his Morse code dots and dashes or paused between letters or words.
⁴³

In late March the first American escort carriers began appearing in the Atlantic; these were small ships that carried twenty-four fighters and navy attack planes and extended air cover across the Atlantic to the convoys. Shipboard direction-finding equipment was now paying off as well; these sets allowed escort destroyers to run down the line of a radio bearing to intercept a U-boat before it got in range of a convoy, increasing by probably 50 percent the number of U-boats sunk by destroyers in February, March, and April according to a study by Rodger Winn’s Operational Intelligence Centre. It was air attacks, however, that now were really beginning to tip the balance. In April and May, fifty-six U-boats were destroyed, thirty-three of those in whole or in part the result of air attacks.
⁴⁴

On May 19, Dönitz, who earlier in the year had been promoted by Hitler to commander-in-chief of the German navy with the rank of Grossadmiral, broadcast to his commanders an extraordinary do-or-die order:

It took just four more days for reality to sink in even to Dönitz. On May 23 he withdrew his U-boats from the Atlantic convoys and ordered a “temporary shifting of operations to areas less endangered by aircraft,” the South Atlantic and the coasts along Gibraltar, Brazil, and Africa. He noted in his diary that not long before he was sinking 100,000 tons of enemy shipping per U-boat lost; that exchange rate was now down to 10,000 tons
per U-boat. “Thus losses in May have reached an intolerable level,” he acknowledged.
⁴⁶

In June, the U.S. Navy’s code breakers in Op-20-G did what Dönitz’s security investigation earlier in the year had failed to, and tracked to the source their growing suspicion that there had to be a leak somewhere in their own code systems. As early as August 1942, Bletchley Park had warned that German radio messages contained information obtained from reading the Allied convoy code. But GC&CS had only one cryptanalyst assigned to checking the security of their own ciphers and the British naval officials in charge of code production were slow to react and convinced themselves that some stopgap changes introduced in December had mostly fixed the problem.
⁴⁷
The American code breakers now carefully matched up decrypted Enigma signals advising the U-boats of “expected convoys” with British and American messages that had been transmitted shortly before in Naval Cypher No. 3. It had required a huge fight to get the security-conscious Admiral King to agree to release copies of the American convoy signals to his own navy’s investigators, but once they finally got hold of them the match was perfect. The Germans’ precise information about convoy movements clearly had come directly from the U.S. Navy’s own signals. On June 10 a new convoy code, Cypher No. 5, was ordered into service. For the rest of the war, the Allied convoy signals remained unbroken.
⁴⁸

The ASWORG scientists also had been denied access to the Enigma secret on King’s orders, but they figured it out themselves—incidentally providing an example of what Dönitz could readily have done had he had an organized scientific staff of his own, which he did not (at least until late in the war, when it was too late to make a critical difference). Morse had assigned Jacinto Steinhardt, a Ph.D. chemist who joined ASWORG in November 1942, to carry out a study of how well Allied radio direction fixes contributed to locating and attacking U-boats. Steinhardt decided to first see how good the data was, and so compared the direction finding reports to the actual locations where U-boats were subsequently located. The DF fixes were unbelievably accurate; in fact, Steinhardt calculated, ten times more accurate than ought to be possible through direction finding alone. Morse and Steinhardt had little trouble deducing the true explanation—that decoded German messages were the real source, which for security reasons were being disguised by the U.S. Navy as DF reports.

At his next meeting with Admiral Low, Morse reported this curious result “with a straight face.” He said:

Enigma decrypts throughout the spring were revealing something of potentially even greater significance than the U-boats’ whereabouts: the growing demoralization and even panic of their crews. In his May withdrawal order, Dönitz vowed the battle would soon be rejoined in the North Atlantic, “the enemy’s most sensitive area.” The U-boat arm, he swore, would soon “subdue the enemy by a continual bloodletting which must cause even the strongest body to bleed to death.” But such heroic language was apparently lost on the men under him, now being killed by the thousands a month: Rodger Winn noted in an intelligence report that in April and May, for the first time in the war, U-boat commanders had failed to press home attacks on convoys, even when they were in a favorable position to do so. Enigma decrypts showed that a sharply increasing number of boats were declaring “mechanical difficulties” during their outbound transit of the Bay of Biscay and turning back to port.

Dönitz made the rounds of the bases giving pep talks, but the mood was morose. “No more parties were given to celebrate the start of a campaign now,” recalled one commander. “We just drank a glass of champagne in silence and shook hands, trying not to look each other in the eyes.”
⁵⁰

The battle of the convoys was, in fact, all but over.

IN PRESSING FOR
stepped-up air operations in the bay back in March 1943, Blackett had anticipated this development, foreseeing that the withdrawal of the U-boats from the North Atlantic would offer an even more compelling reason to focus the anti-U-boat offensive on the bay; it would also offer an opportunity to deliver the coup de grâce to Dönitz’s force. As long as the U-boats remained engaged in attacking the convoys, there was some truth to King’s insistence that the “convoy area” was the proper focal point for the application of air power: the convoys were the bait that drew the U-boats into the killing zones of escorting aircraft. But, as Blackett argued with remarkable prescience in a March 23 note to the naval staff, that situation would surely not last. At present the use of VLR aircraft to provide
air cover to shadowed convoys was undeniably the best way to employ them; measured in terms of both ships saved and U-boats sunk, each flying hour produced ten times greater returns in that mission than when employed patrolling the bay. But the comparison was “misleading,” Blackett cautioned, and certain to change:

Blackett attached an analysis that he and Williams had carried out on the number of VLR aircraft needed to thoroughly cover the bay. It was definitely very much fewer than the numbers that would be needed to protect ships all across the Atlantic and Indian oceans once the U-boats shifted their hunting grounds: “certainly less than 300 aircraft, and probably less than 200.” Just forcing the U-boats to stay submerged as much as possible while transiting the bay cut down their operational time at sea significantly. And Blackett noted that even if the U-boats adopted maximum submergence tactics while crossing the bay, they would still have to cross half the distance on the surface, where they were vulnerable to detection by radar and to attack: that was because their submerged speed was so much slower. There was thus a limit to how much the U-boats could ever successfully evade aircraft. Saturating a band about 120 miles wide with constant patrols day and night for twenty-four hours would in theory catch every U-boat passing through that zone during the period.

Even when sightings did not offer a chance to carry out an attack, forcing
the U-boats to dive when air patrols passed over produced
increasing
returns on each additional investment in flying time. Each time a boat dived it had to spend an extra quarter of an hour on the surface recharging its batteries just to replace the energy expended diving and resurfacing; that extra time on the surface in turn increased its chances of being spotted by another air patrol. More flying hours thus increased the number of U-boats sighted per flying hour. In March and April, when Coastal Command aircraft began to be equipped in substantial numbers with 10-centimeter radar, Allied aircraft were already spotting 60 percent of the U-boats that made the transit across the bay—which worked out to one for every 30 flying hours—and attacked 40 percent of them.
⁵²

Those numbers shot up in June and July. Blackett and Williams and others in the Admiralty ORS produced a series of studies on the most effective balance of day and night flying, the advantages of regular versus irregular patrols, the best area and proper width of the patrol patterns, the expected German countermeasures and how best to deal with them. From March to August, aircraft sighted 350 U-boats in the bay and attacked 200 of them. In June, the U-boats began transiting the bay in groups hoping to fight off air attacks with massed antiaircraft fire from their deck guns; the Allies countered by sending surface vessels into the bay to add an additional threat the U-boats had to counter.
⁵³
At the peak of the bay air offensive, every operational U-boat was being sighted on average nearly twice each month by Allied aircraft. From July 2 to August 3, the Allies destroyed 41 more U-boats, 16 of those by air attacks in the bay.
⁵⁴
The U-boat war, as Admiral Noble had said, was now indeed an air problem: in June through September naval and shore-based aircraft destroyed four times as many U-boats as surface vessels did. So it would remain throughout the rest of the war, with aircraft consistently accounting for one and a half to two times as many kills as surface vessels.
⁵⁵

Churchill would reflect in his memoirs on this turning point of the war in the summer of 1943, when “the decisive battle with the U-boats was now fought and won,” and the fear and caution yet remained:

But hope was in the ascendant even as vigilance remained. One incredible fact stood out: for the month of June 1943, the number of merchant ships lost to U-boats was for the first time no greater than the number of U-boats sunk.
⁵⁷

IN JUNE
the first of the American four-rotor bombes were completed and running, finally turning the daily U-boat Enigma solution into an almost routine assembly-line process, which would continue to the end of the war as dozens more of the machines were shipped in the ensuing months to Washington from the National Cash Register factory in Dayton.
⁵⁸
Bletchley Park would teleprint bombe menus to Washington; the U.S. Navy WAVES who operated the bombes at the Naval Communications Annex, located in a former girls finishing school that the navy had seized at the start of the war in the northwest of the city, set up the machines; and a few hours later the results were on their way back across the Atlantic, and decoded Enigma messages flowed into the Operational Intelligence Centre at the Admiralty and the “Secret Room” within the submarine tracking center at Main Navy in downtown Washington.
⁵⁹