by John Prados
But their general sense that the Japanese were facing challenges in moving their fuel remained accurate. The presence and activity of oilers became an indicator of fleet operations. In mid-September a JICPOA weekly estimate noted that large-scale Japanese logistic preparations were apparent. On October 2, the estimate indicated several tankers regularly carrying oil from Sumatra to Lingga and noted the Japanese were training in underway refueling. That report on the observations of tanker activities furnished the best guide to enemy intentions.
• • •
THE QUESTION OF whether the Japanese fleet would emerge to fight a Philippine invasion lay at the heart of the entire Allied intelligence enterprise. If the spooks had the right answer, General MacArthur and Admirals Nimitz, Halsey, and Kinkaid would be ready. At the Zoo, the intel folks published a correct general appreciation as early as September. The JICPOA estimate for September 4 observed that the Japanese were acting to restore their fleet and air strength, an ambitious effort that should culminate in about December. Intelligence analysts were aware of the Imperial Navy’s lack of destroyers and foresaw five joining the fleet in that time. They knew of the three Unryu-class aircraft carriers (but not of the Shinano) and expected the Japanese would be able to assemble the largest air force to date. They would then be able to advance with the surface fleet operating under the umbrella of its own shore-based air force to attempt “an all-out decisive battle.”
Much can be said of this appreciation. General Twitty’s Enemy Situation Section correctly projected the overall Imperial Navy plan—their surface fleet would be used “to complete the destruction of our forces after the air forces have delivered a crippling blow and secured control of the air.” The Zoo analysts were almost correct on Japanese strength (twice as many destroyers would join the fleet by December, along with the the Shinano). Their main shortcoming lay in tying Japanese action to a date (December, which corresponded to the expected availability of the new JNAF air groups) rather than an event (invasion of the Philippines).
On September 15, the intelligence center issued a new estimate. JICPOA hypothesized accurately that the Japanese would be unlikely to use more than a single carrier division of 175 to 200 aircraft. Intelligence also reported the separation of the Japanese surface fleet from its carrier force along with the reason for it—to get the big ships close to their fuel. But the analysts foresaw the Kurita fleet as not acting independently except for an unusual and unexpected opportunity and with strong air support. They also mistakenly saw the force as entirely composed of heavy cruisers. In Washington, both ONI and OP-20-G agreed in their estimates on Japanese fleet dispositions on September 21 that five Japanese heavy cruisers were located in the Philippines. Thus, Halsey’s air intelligence officers during the Luzon carrier raids had briefed their pilots with intel from on high.
Early on JICPOA captured the essence of the Japanese scheme—the use of a surface force under land-based air cover—and it correctly followed tanker activities as an index of Japanese operations. It also scoped out the Japanese concept for decoy forces. The Zoo analysts, however, erred on the specific intentions of the Kurita fleet and on their overall sense of Tokyo’s timing.
More aspects of the Japanese plans had been divined too. A Combined Fleet manual, Striking Force Tactics, which had been captured in the Marianas, was translated and key passages were circulated in a JICPOA pamphlet series called Know Your Enemy. The Japanese manual discussed the use of decoy forces to fool the Allies and illustrated the idea with the example of the hybrid battleship-carriers Ise and Hyuga. In the coming battle, those very ships would sail as part of Admiral Ozawa’s decoy fleet. Also striking, Japanese adoption of the term “diversion attack force,” as tactical titles for the Kurita fleet and the Fifth Fleet were known and reported in the estimates. Allied spooks had even caught wind of the code name Sho, though they did not yet know its meaning.
While Pearl Harbor intelligence did well, it could also be taken by surprise at times. The error on the composition of the Kurita fleet is one example. Late in September, Lieutenant Commander Edward J. Matthews visited the Zoo on his way to join the staff of Task Force 34, a special contingency formation of all the Third Fleet’s fast battleships. Matthews had been assigned as combat intelligence officer to the battleship admiral. Commander Matthews carried data sheets from Washington with intel on Japanese warships that Jasper Holmes had never seen. JICPOA had to apply to Washington for access.
There were other drawbacks too. Holmes agonized over getting the Zoo’s intelligence where it was needed “across this wide Pacific.” The admirals would hear something that sounded interesting, and then demands for the data would pour in. With its limited numbers of copies of the intelligence studies, JICPOA could never meet demand. More appeals to Washington. “Plans change so rapidly and new commands spring up so quickly that we are having the devil’s own time trying to keep the proper people informed.” General Twitty asked Admiral Nimitz if JICPOA could set up forward distribution centers. The CINCPAC agreed. An office on Guam became the first—and it turned into the nucleus for a full-service “JICPOA Forward.” But the advance echelon unit had yet to open its doors when the Philippine battle started.
The perspective from MacArthur’s intelligence people was equally interesting. Captain Arthur McCollum headed the Seventh Fleet Intelligence Center (SEFIC) and had largely the same problem as Holmes. Though not quite so far-flung as Nimitz’s Pacific Ocean area, the Southwest Pacific still represented a huge region. Distributing publications by pouch often seemed impracticable. Instead, McCollum adopted a hybrid system. Materials like ATIS translations or technical manuals were still hand-delivered. Intelligence estimates for strategic planning appeared on paper initially but would then be modified by radio updates. This system led to confusion regarding SEFIC’s view of Japanese intentions.
A basic appreciation of the Japanese, published after the Philippine Sea battle, agreed with JICPOA’s initial assessment that the Imperial Navy would avoid action for six to eight months while training new carrier air groups. But from late September onward, Captain McCollum began to anticipate that the Imperial Navy would counter a Philippine invasion with surface forces. A few months later, the Seventh Fleet commander sent General MacArthur a report crediting McCollum’s staff memo of September 24 as foreseeing that the enemy would use the bulk of their fleet in defense of the archipelago.
That view, however, remained contentious. The SEFIC administrative officer disputed McCollum’s suspicions, and General George Kenney, MacArthur’s air commander, also felt Japanese fleet action unlikely. According to Lieutenant Lawrence F. Ebb, of the Enemy Information Section of the fleet intelligence center, the line intelligence experts took McCollum’s side. Ebb and his colleagues knew that the Japanese did not expect any threats from the Indian Ocean, so the presence of the Japanese surface fleet at Singapore had to be for its favorable location in relation to the Indonesian oil fields. The SEFIC experts also knew the Japanese did anticipate an Allied move against the Philippines.
Several times Captain McCollum petitioned SOWESPAC to ask the U.S. command in the China-Burma-India theater to use a few of its very-long-range B-29 bombers to reconnoiter the Singapore-Lingga area, but General MacArthur and his top spyman, Major General Charles Willoughby, never agreed. The fleet intelligence staff then moved from Brisbane, to Hollandia, to the headquarters ship Wasatch, in two groups. The ones who stayed behind when the first wave set off were led by the officer who believed the Japanese would be a negligible quantity in the campaign.
British observers with SOWESPAC, misled or not, gathered that authorities were skeptical the Japanese would risk it all. A British official monograph reconstructing the campaign quotes the observers’ report: “It was not believed that the major elements of the Japanese Fleet would be involved in the present operation, but that fast Task Forces might strike at our supply lines taking full advantage of darkness, surprise, and land-based air.�
��
As for the Australians, their Central Bureau functioned in the radio intelligence field, and its intercept organization included nearly 3,400 operators. But they had focused on land warfare. MacArthur asked the Australians to provide him an element to use for the invasion and they responded, forming the Royal Australian Air Force 6th Wireless Unit. Working as a mobile intercept unit in the invasion convoys, the Australians were able to report that their ships had been spotted, but they discovered nothing about the Japanese fleet.
The intelligence impasse continued until the eve of the invasion. On September 24, an update estimated that two major “diversion attack forces” would sally against the Allied invasion, and it tabulated overall Imperial Navy strength that included virtually all of the available Japanese warships. That update went out by radio, as did subsequent versions. The same day, General Kenney issued an appreciation that doubted major Japanese naval activity in the Philippines—an assessment air intelligence repeated on October 4. Captain Raymond D. Tarbuck, the top operations planner at General MacArthur’s headquarters, on October 4 expected the Kurita fleet to maneuver in the South China Sea as a preliminary move to attacking via Surigao Strait. Ray Tarbuck later decided Art McCollum was right. Tarbuck wrote a new memorandum some deemed prescient. This time he took SEFIC’s point of view completely that all remaining Imperial Navy surface power would operate from west of the Philippines, with a specific threat through the Surigao Strait.
Following the war, both Navy historian Samuel Eliot Morison and the Naval War College concluded that the Sho operation took Seventh Fleet and MacArthur’s command by surprise. Arthur McCollum went to his grave in 1976 insisting there had been no surprise. McCollum observed that in the wake of the Great Marianas Turkey Shoot it had been impolitic to put on paper a judgment that the Japanese fleet was about to fight again, so there had been an initial mistake, but that this had swiftly been corrected by updates in dispatches. Lieutenant Ebb in the trenches at SEFIC agreed. No matter the paper trail, Ebb recalled, SEFIC briefed the top brass on flagship Wasatch that a naval battle impended. How could it be otherwise? Into the summer of 1944, over a period of five months the Japanese fleet had been driven out of three operating bases, creamed in a major battle, then forced to split up. The Imperial Navy had their backs to the wall, and the Allies were going to seize the position that linked the parts of the Combined Fleet. The Japanese were going to fight. Many were sure of it.
Thus, JICPOA noted in September 1944, “The overall tanker movement picture is somewhat confusing insofar as determining fleet intentions goes, but it is apparent that large-scale logistic preparations are in the making.” Japanese messages were intercepted that ordered a couple of oilers to the Pescadore Islands, near Taiwan. It was early October. This was important to Admiral Bill Halsey and his Third Fleet, for just then Task Force 38 had gone wild in the Japanese backfield again, slashing at the aerial pipeline in Okinawa, the Ryukyu Islands, and especially Taiwan. Those orders to the Japanese oilers told Halsey that the enemy might be coming after him.
Over the next weeks, decoded messages concerning tankers revealed a succession of key indicators. One early October message, complaining that even with the help of the Japanese Army the Sumatran oil could not be handled efficiently, sent four tankers to Balikpapan, a port on the north coast of Borneo. That place would be ideal for the Kurita fleet to refuel on its way to the Philippines. The command designated one particular tanker to handle all the Palembang-Singapore fuel hauling. Later, that tanker, too, would be sent on to Balikpapan. Another dispatch directed oilers responsible to Admiral Ozawa’s main body to take on full loads of fuel from the Imperial Navy’s central storage facility at Tokuyama. More messages for this group were intercepted too. Other decoded messages indicated that tankers had been ordered to fuel the Fifth Fleet, and—very suggestively—that merchant ships had been ordered to give up their own oil if the Kurita fleet required it. On October 16, Combined Fleet chief of staff Kusaka assigned half a dozen oilers directly to the Kurita fleet and instructed others to follow the admiral’s orders. The OP-20-G comment was “This is the first indication of a possible sortie by [the Kurita fleet] which may comprise the bulk of the . . . units now in the Singapore area.”
The stage had been set for a great battle.
CHAPTER 5
DESTROY THE INVADING ENEMY
Admiral William F. Halsey liked to say he had a destroyerman’s nose. By that, Halsey meant he could stand in the wind and foretell the weather. Pete Mitscher possessed a similar gift. The carrier combat commander would often outshine his meteorological staff, making precise predictions, whereas the scientific observations and calculations of his experts led to uncertain forecasts. Both men’s skills would be in great demand in early October. Both admirals, with the bulk of the Third Fleet, were at Ulithi atoll, a typhoon bearing down on them. Halsey had wanted his warships to get a week’s rest. Instead, the Bull decided he had better order them to sea to be safe from the storm. Fortunately the typhoon, its center several hundred miles to the north and west, dealt only a glancing blow to the Allied anchorage.
Task Force 38 returned to Ulithi and did manage a forty-eight-hour stand-down before going back to sea. Weather remained rough, though. Aboard light carrier Cowpens in Admiral McCain’s Task Group 38.1, the ship’s pitch in the running seas became so violent that six fighter planes and a torpedo bomber broke loose from their moorings and smashed around the hangar deck. The torpedo plane and several fighters were so badly damaged they had to be pushed overboard. In Group 38.3, the Essex took water over her flight deck. Airman Jack Miller of the squadron Bombing 15 heard that carriers and battleships both were breaking waves so high the spray splattered their bridges. Rear Admiral Ralph E. Davison’s Task Group 38.4 caught even more storm coming north from Palau. On the Enterprise, part of that group, sailors found their mess trays sliding off the tables. The vessel’s usual slow gyration and stately progress became a sharp lurch and a buck, like a small boat tossed about in heavy seas. In one day alone, the wind force increased fivefold, to near fifty knots, settling back to twenty-five—still a stiff wind—for the next couple of days.
Davison’s group rendezvoused with the others on October 7. Radar operators were pleased to see the electronic “pips” of all those friendly warships appear on their screens one, or a few, at a time. All the task groups then sailed in company. This armada comprised seventeen aircraft carriers (eight light), six battleships, seven heavy and ten light cruisers, and sixty-four destroyers.
The fleet followed the storm. Mick Carney, the fleet chief of staff, exulted that the storm lingered in the Tokyo area. Halsey’s sailors began to call the typhoon “Task Force Zero” because it advanced ahead of them, smiting the enemy. They imagined the storm and their task force as a one-two punch.
TEMPOS, CUSTOMS, AND COMBAT
The myriad combinations of strategic vision were now going onto the chart table for execution. The plan was to tear up the Japanese rear areas with a carrier raid on the Ryukyu chain and Taiwan, the islands along the Asiatic coast that linked the Philippines with Japan. Any Japanese reinforcement from the Empire would have to traverse this region; in particular ships and planes would need to refuel there. There were too many bases to be knocked out, but at a minimum, Halsey’s fleet might cut deeply into Japan’s air strength and transportation capacity.
Halsey’s plan, simple and daring, provided that Task Force 38 would appear at the northern end of the strike zone, off Okinawa, and hit hard across a 300-mile arc of the Ryukyus. Then Task Force 38 would make its way south, off the Taiwanese coast, finally doubling back along the eastern littoral of the Philippines to provide the heavy cover as MacArthur’s invasion convoys entered Leyte Gulf. A surface attack group would divert the Japanese, appearing off Marcus Island to bombard them, making Japanese commanders think “invasion” there. Newly arrived long-range patrol bombers would interfere with the Imperial Navy’s search planes and si
nk their picket boats, clearing the way for the carrier sweep.
This time, however, the Allies did not achieve the surprise they coveted. Imperial Navy intelligence, while not specifically informed, provided a certain strategic warning. On October 2, the Owada Group, the fleet’s radio intelligence unit, noted that message traffic from Pearl Harbor and the Big Blue Fleet indicated that the American striking force would be large, and predicted it might target the Ryukyus or Taiwan. On October 3, the NGS ordered local commands in the Ryukyus to check radars and effect necessary repairs. That same day, the Southwest Area Fleet at Manila reported that its radio traffic analysis indicated a forthcoming attack on either the Philippines or Taiwan. On October 5, Manila ordered an alert for the northern end of Luzon and a withdrawal of shipping from the Taiwan Strait. The next day, the Owada Group noted the departure of a U.S. naval force, which must have been Davison’s Task Group 38.4. The Kure Naval District and the Combined Fleet itself issued similar reports. Chatter on the Japanese side attained such levels that the Office of Naval Intelligence, from Washington, informed commanders that Japanese radio traffic reflected expectations for an American attack against the Ryukyu, Taiwan, or Philippine areas. On October 9, the JNAF scout plane on the 140-degree vector from Kanoya stopped transmitting midmessage at 450 nautical miles from base. It had been downed by a U.S. patrol plane from Tinian, though the Japanese assumed the scout had fallen to a carrier plane. The Sasebo Naval District called an alert. The fat had gone in the fire.