Section 3: Return to the Philippines
Blank page
Chapter 10: Prelude to Invasion
At the end of July 1944 MacArthur’s Southwest Pacific Area forces had landed at Noemfoor and Sansapor to end their long campaign in New Guinea while Pacific Ocean Areas troops under Nimitz consolidated positions they had seized in the Marianas. The time had come for a final decision on the interim strategic objective for the Pacific war specified by the CCS in the preceding December at Cairo: seizure of an Allied base in the Formosa-Luzon-China coast area which would permit the establishment of a direct sea route to China and interdiction of Japanese communications with the Netherlands East Indies.
Neither MacArthur nor Nimitz had a definite commitment precisely placing this major strategic base. The JCS on 12 March 1944 had preferred to indicate that, according to the situation on 15 February 1945, either Nimitz would be expected to invade Formosa or MacArthur would be directed to occupy Luzon. Meanwhile, MacArthur would complete operations along the New Guinea coast designed to support a POA invasion of the Palaus on 15 September and a SWPA assault against Mindanao on 15 November.* Nimitz, having occupied the southern Marianas and the Palaus, might be expected to attack Formosa on 15 February 1945; or, if Luzon could not be effectively neutralized by SWPA’s land-based aviation, MacArthur’s forces might be required to move northward from Mindanao to Luzon on 15 February in preparation for a POA assault against Formosa at a delayed target date.1 No other strategic decision of the Pacific war would be discussed at greater length or with more heat.
Invasion Plans
Once the JCS had issued the directive of 12 March 1944, clarification of its tentative strategy necessarily awaited theater action. Busy
* For full discussion, see Vol. IV, pp 549-55, 570-74.
with the Hollandia operation, SWPA undertook no immediate revision of its strategic plans. In answer to a request for information, MacArthur told the JCS on 8 May that he intended to seize an airdrome site on the coast of the Vogelkop about 1 August and then, coordinating his target date with POA’s invasion of the Palaus, to acquire an airdrome site on Halmahera for flank protection and air support of the invasion of southern Mindanao.2 The Joint Planning Staff (JPS) in Washington, who regarded the Palaus as the supporting base for Mindanao, saw little need for another such base in the northern Moluccas, of which Halmahera and Morotai were the chief islands. Moreover, the Joint Strategic Survey Committee (JSSC) on 29 May initiated an investigation of possible short cuts to speed up the Pacific war. Noting that current intelligence indicated the Japanese were building up strength in the Philippines at the expense of Formosa, the JSSC questioned whether it might not be wise to bypass Mindanao and attack Formosa directly. Deterioration of the Allied situation in China seemed to argue that Formosa should be captured at the earliest possible date. Adding this suggestion to their own opinion that invasion of the northern Moluccas was of questionable value, the JPS persuaded the JCS to question MacArthur and Nimitz on 13 June 1944 about their ability to speed operations by omitting steps projected prior to Formosa, by accelerating target dates, or by selecting other objectives, including targets in Japan proper.3
These questions reached the Pacific theaters at an inopportune moment. Only ten days before, Nimitz had issued his GRANITE II plan, which set target dates for POA operations as follows: the southern Marianas (FORAGER), 15 June; the Palaus (STALEMATE), 8 September; Mindanao ( INSURGENT), 15 November; southern Formosa and Amoy ( CAUSEWAY), 15 February 1945, or, if Formosa proved impracticable, Luzon ( INDUCTION), 15 February 1945. Until the results of FORAGER became clear, he could offer no information regarding acceleration of later operations.4
At Brisbane, SWPA planners were just completing the finished draft of RENO V, which would be formally issued on 15 June. This plan, last of the RENO series,* reflected SWPA successes at Wakde and Biak. Subsequent campaigns were phased as follows: 1) Establishment of an air base in the Vogelkop and another in the northern Moluccas, with a contingent operation planned for the Kai and Tanimbar
* For the origin and development of this series, see index to Vol. IV.
The Philippines
islands if a Japanese air concentration on the west of New Guinea demanded additional left-flank protection. This phase would be accomplished between July and October 1944, with the target date for invasion of Morotai, north of Halmahera, timed to coincide with POA’s entry into the Palaus. Simultaneous target dates would
permit the Pacific Fleet to cover both operations at one time. 2) Establishment of bases in Mindanao to support air operations against Luzon and North Borneo, November-December 1944. SWPA forces would seize Sarangani Bay, on the coast of southern Mindanao, on 25 October and establish airfields to support the principal effort on 15 November against northern Mindanao and Leyte. Parts of Samar would be added to the holding, and a major air, naval, and logistic base built. 3) Invasion of Luzon, January-March 1945. During January, a major amphibious movement, supported by airborne troops from Leyte, would seize the Bicol area of southern Luzon, and concurrently another landing operation from POA bases would seize airdrome facilities in the Aparri area of northeastern Luzon. During February the island of Mindoro, lying immediately southwest of Luzon, would be occupied by an airborne invasion from Leyte. With assistance from Filipino troops, SWPA would clean out the Visayas between December 1944 and June 1945, thus ringing Japanese forces remaining on Luzon with Allied air bases. In addition to an intensive bombardment of Luzon, Allied air forces would begin interdiction strikes from Mindanao and Sulu bases over North Borneo and the South China Sea. 4) Reoccupation of Luzon, April-June 1945. A major landing force would seize beachheads in the Lingayen Gulf area of the west coast of Luzon on 1 April, and, with an armored division and strong airborne support, the main attack would penetrate southward to occupy Manila. A secondary shore-to-shore operation from the Bicol Peninsula would seize a beachhead on the eastern coast of Luzon at Baler and Atimonan bays and force its way through the mountains to join the main drive. Reserve forces would be employed in contingent operations to outflank the Japanese on Luzon, and, as rapidly as possible, air bases on the island would be rushed to completion to broaden the strategic air effort against Japan. SWPA presumed that Pacific Fleet support would be made available for each phase of RENO V; by the time of the last phase, it expected to be using an equivalent of twenty-seven divisions.5
This plan was subsequently to be much streamlined, both as to timing and the forces scheduled for employment, but on 18 June MacArthur answered the JCS query with a flat negative:6 his forces would be strained to the utmost to meet target dates already specified. To drop operations intermediate to the landing on Formosa was a radical departure from any previous Pacific plan, and the suggestion that the
attack might be launched from the central Pacific without appreciable land-based air support was most unsound. MacArthur believed it would be necessary at least for SWPA to occupy Luzon and establish air bases there, Similarly, the proposal to bypass all objectives and invade Japan proper was utterly unsound: successes won, MacArthur cautioned, however great, should not lead to suicidal ventures. In addition to purely military reasons for retaking the Philippines, he urged that the United States owed the Filipinos their freedom. If serious consideration were being given to a direct invasion of Japan, he asked permission to come to Washington to express his views in person.
Although Marshall replied that neither of the two propositions was unsound and cautioned MacArthur not to let personal feelings and Philippine political considerations vitiate any plan to shorten the war against Japan,7 it was obvious in Washington that SWPA target dates could not be significantly advanced at the moment. The JPS concluded that deletion of an invasion of Mindanao would hasten the Formosa operation by no more than one month because of a lack of available attack transports and cargo vessels. In addition, weather conditions would prevent an attack on Japan proper prior to October or November 1945. The JPS, however, considered that deletion of a Mindanao operation would avoid the possibility of a long and costly Philippines campaign; the only question was whether Japanese air strength on Luzon could be neutralized prior to Formosa operations without bases in the southern Philippines.8
At this juncture Nimitz, whose forces were being delayed in the Marianas, also confessed an inability to accelerate his campaigns. He had planned to invade the Palaus a week before the JCS target date of 15 September, but recent estimates indicated that the Palaus garrisons were being increased from 9,000 to 40,000 troops, and he was doubtful that even the JCS timing could be met. In order to save time he proposed to limit the operation to seizure of only two islands in the Palaus-Angaur and Peleliu-and to secure a fleet staging point at Kossol Passage. He would take Yap and Ulithi either simultaneously or shortly afterward. Having obtained information about RENO V from a SWPA-POA staff conference at Pearl Harbor on 3 July, Nimitz notified Washington that he considered the plan of campaign to be sound, even if the timing appeared optimistic.9 He stressed the need for SWPA air support from Mindanao and Leyte prior to the Formosa operation. Leyte was of special importance: if this island fell
into SWPA hands, the neutralization of the whole Philippines would be assured and subsequent operations could be expedited.10
These statements ended proposals, at least for a time, to skip operations in the southern Philippines. “We certainly should not take any action now to prevent the Mindanao-Leyte operation,” Maj. Gen. Thomas T. Handy, AC/S, OPD, advised Marshall. “MacArthur’s stand that Luzon must be seized before we go to Formosa may be right. Nimitz is not sure. ... I believe we should await future developments.”11 Nimitz was permitted to reduce the scale of the Palaus operation as he had proposed. Morotai, by implication, gained status as the main supporting base for the invasion of southern Mindanao.12
Later in July questions of strategy again were reviewed at a conference of President Roosevelt, MacArthur, and Nimitz at Pearl Harbor. Remembering MacArthur’s proposal to visit Washington, Marshall had taken advantage of Roosevelt’s inspection of Pacific defenses to direct MacArthur to come to Hawaii on 26 July. On the next evening Roosevelt invited MacArthur, Nimitz, and Halsey to dinner, and after the meal he drew out a map, pointed to Mindanao, and said, “Well, Douglas, where do we go from here?” MacArthur, who had not been told that he would meet Roosevelt in Hawaii or that strategy would be discussed after dinner, nevertheless launched into a discussion of his ideas which lasted all evening. He urged that Luzon be seized (target date 15-25 February 1945) and bases established there from which Japanese shipping in the South China Sea could be interdicted and Formosa neutralized. The Pacific Fleet and POA would then be free to operate against the Japanese fleet and to seize air base areas in the Ryukyus and Bonins. Seizure of Formosa would be a massive operation, extremely costly in men and shipping, logistically precarious, and time-consuming. It would offer to the enemy air and naval opportunities against an overextended Allied supply line which would never otherwise be afforded. He was willing to give a personal guarantee that the Luzon campaign could be completed in six weeks, or thirty days after a landing at Lingayen.* He doubted that Luzon could be adequately neutralized from Leyte-Mindanao bases prior to CAUSEWAY, and he reiterated his conviction that the United States was morally obligated to liberate Luzon. The President agreed about the moral responsibility. Nimitz, presenting his views next morning,
* MacArthur told General Kenney that he would have Manila six weeks after the landing at Lingayen and all of the Philippines within eight months.
Battle For Leyte Gulf , Planes Sink Abukuma …
… But Miss Yamato
Ormoc Bay, FEAF Planes Sink Destroyer
Ormoc Bay, Attack on a Transport
was “clear that the time has not yet arrived for firm decision on moves subsequent to Leyte.”13 Since the conference was only for discussion, no decisions were reached.
Meanwhile, preparation of detailed campaign plans had begun at SWPA headquarters in Brisbane. To move landing forces across the 650 miles of sea between New Guinea and southernmost Mindanao would not be simple, and the augmentation of Japanese air forces under way in the Philippines would further complicate the problem. Even from Morotai it was some 350 miles to Mindanao. After discussing RENO V with Fifth Air Force’s Maj. Gen. Ennis C. Whitehead and Thirteenth Air Force’s Maj. Gen. St. Clair Streett, Lt. Gen. George C. Kenney, commanding the U.S. Far East Air Forces (FEAF)* and the Allied Air Forces, SWPA, informed MacArthur on 11 July that the first two phases of the plan were not in harmony with air capabilities.14 Both the projected seizure of Morotai and the subsequent invasion of Mindanao-Leyte were, in his opinion, based on an over-optimistic expectation of support from Pacific Fleet carriers and contemplated establishment of air bases which would not be mutually supporting. “Carrier units,” he wrote, “are so restricted in their time over targets and radius of action that they cannot be expected to neutralize and maintain neutralization of enemy strong points and air installations which would be within range of our objective.” Direct support by carriers at a beachhead would be unsatisfactory because their planes lacked sufficient strafing and bombardment power. They could furnish fighter cover over a beachhead, but enemy air, surface, and subsurface action, together with the physical limitations of carriers, created constant uncertainty as to its maintenance. Kenney granted that the proposed invasion of the northern Moluccas could be covered by FEAF heavy bombers from Biak, but Japanese air installations threatening Sarangani and Leyte would be outside the range of fighter-escorted heavy bombers from either Morotai or Biak. Distances between Sansapor, Morotai, Sarangani, and Leyte were all too great for mutual air support; the Japanese could select one of the bases and knock it out before SWPA air units could protect it. It seemed to Kenney that these problems could best be met by properly spacing land bases, and he favored scaling down the individual invasions so as to move and build an air base every twenty to thirty days. Specifically, he recommended that if Sarangani were to be delayed until November,
* See Vol. IV, 646-51.
SWPA should install fighters and attack units at Talaud Islands (midway between Morotai and Sarangani) to support it; that SWPA should establish an intermediate fighter and attack aviation base in the Del Monte area of Mindanao prior to Leyte; and that heavy bombers should be operational at Sarangani in time to support the invasion at Leyte.15
This admonition, coupled with the fact that SWPA representatives had been informed in Hawaii that the amount of assault shipping obtainable would be less than expected, led to a recasting of the SWPA plans, It would be necessary to use the same amphibious equipment for both Sarangani and Leyte, making the latter follow Sarangani by thirty-five instead of twenty days as planned in RENO V. Since the troops put ashore at Sarangani would have to remain there for five weeks, they would require additional air support from a base in the Talaud Islands, which would also permit increased air coverage of Japanese targets in the central Philippines and southern Luzon. MacArthur advised the War Department on 23 July that his revised schedule would have to be: Morotai, 15 September; the Talauds, 15 October; Sarangani 15 November; and Leyte, 20 December.16
This schedule, however, was tentative, and it was kept under almost day-to-day study by a series of WIDEAWAKE planning conferences which met intermittently in Brisbane between early July and September 1944. There, representatives of SWPA G-3 and of the subordinate headquarters prepared a series of staff studies covering the planned invasions. Based on a new set of plans called MUSKETEER, the first of these was issued by SWPA on 10 July. MUSKETEER, unlike the more comprehensive RENO V, was concerned solely with operations in the Philippines, and at the initiation of the plan it was assumed that Allied forces would be established in the Marianas, Palaus, and northern Moluccas. The plan of campaign aimed at the establishment of air units in the central region of Luzon in four major phases of operations, KING, LOVE, MIKE, and VICTOR.17
The KING operations were to secure an initial lodgment in the southern and central Philippines and the establishment of bases to support subsequent operations. The preliminary blow, KING I, was to be directed at Sarangani Bay in Mindanao on 15 November 1944; the main effort, or KING II, would come on to December 1944 at Leyte Gulf, where major air, naval, and logistic bases would be established.
The penetration was to continue northward with Operation LOVE, a series of campaigns designed to seize a favorable line of departure and to provide air and naval bases for operations against central Luzon. The main effort of this series, LOVE I (January 1945), was to come in the Bicol provinces of southeastern Luzon. Concurrently, LOVE II would establish air bases at Aparri on the northern coast of Luzon in order to cover convoy movements through the Luzon Strait. LOVE III (February 1945),the occupation and development of airfields in southwestern Mindoro, was designed as a subsidiary airborne operation aimed at securing bases for supplying convoy cover in the San Bernardino Strait-Sibuyan Sea routes, and for mounting airborne and air support operations against central Luzon. MIKE I would take place at Lingayen Gulf in an all-out invasion tentatively scheduled for 1 April 1945;
MIKE II was set for the same month in the Baler-Atimonan area on the eastern coast of Luzon; a concurrent diversion, MIKE III, was projected for the Batangas area of southwestern Luzon; and a supporting operation, MIKE IV, was scheduled for May to strike the west coast of Luzon in Zambales Province in order to forestall a Japanese retreat into Bataan. Consolidation of Luzon, MIKE V, was expected to follow these initial invasions. The final reduction of Japanese garrisons in bypassed portions of Mindanao and the Visayas would comprise the VICTOR series of operations.*
In Washington the JPS who had not been advised of MUSKETEER, believed that early March 1945 was the latest date at which POA could invade Formosa. To allow three months for the preparation of an air base at Leyte and the neutralization of Luzon, it would be imperative that SWPA gain control of Leyte by 1 December. Accordingly, they questioned both MacArthur and Nimitz on 27 July as to 1) the practicability of eliminating the attack on the Palaus and substituting smaller attacks on Woleai, Ulithi, and Yap; 2) whether the Talauds and/or Sarangani Bay could be abandoned in favor of direct movement into northern Mindanao and Leyte; and 3) what specific operations MacArthur contemplated in northern Mindanao.18 MacArthur answered the inquiry with a blistering message, expressing his strongest disagreement with the assumption that the primary purpose of his entry into the Philippines was to establish air bases for support of POA operations against Formosa. With the capture of Luzon,
* See below, pp. 450-63.
the hazardous operation against Formosa would be unnecessary. In answer to the JPS’s specific questions, he stated that the Palaus were essential, that Sarangani Bay and the Talauds could not be eliminated, and that operations in northern Mindanao would follow Leyte as soon as possible so as to relieve the civilian population there.19
To attempt a reconciliation of opinion between MacArthur and the War Department, Maj. Gen. John E. Hull, Chief of the Theater Group, OPD and Col. W. L. Ritchie, head of that office’s Southwest Pacific Theater Section, joined Lt. Gen. Barney M. Giles, Chief of Air Staff, in Brisbane early in August for conferences with MacArthur. The SWPA head insisted that the Palaus would be needed as vital flank protection for the entrance of the southern Philippines. The use of shipping released by canceling the Palaus operation would make it possible to move up the Sarangani landing by about five days, but there would still be an interval of seven weeks between Sarangani and Leyte. Both MacArthur and Kenney were dubious that Luzon bases could be neutralized from Leyte, and MacArthur repeated with customary eloquence his conviction of the necessity for seizing Luzon and the impracticability of the Formosa operation. Both Hull and Giles were tentatively convinced of the correctness of MacArthur’s position, although they reserved final judgment until they had talked to Nimitz and Richardson in Hawaii.20 Giles, reporting his near-conviction to Arnold, observed: “I realize it is very hard to keep from getting ‘localitis’ after having talked to MacArthur for five hours (I mean listen).”21
The visit was not without some effect, however, on SWPA plans. Both Giles and Hull were convinced that SWPA should have two new air commando groups (completely airborne P-51 groups with their own transport aircraft) which were being trained in the United States.” With the expectation of getting these groups, the FEAF staff projected a new airborne invasion (styled KING III) into the Misamis Occidental Province of western Mindanao. Here, in an area controlled by guerrillas, fighter fields would be prepared for cover of air operations into the Visayas and southern Luzon. Given this operation, Kenney was willing to bypass either the Talaud Islands or Sarangani, preferably the latter. Ritchie seems to have instigated further SWPA study looking toward acceleration of operations after Leyte, predicated
* See above, p. 208n.
on MacArthur’s conviction that Luzon would be invaded instead of Formosa.22 A revision of the MUSKETEER plan, drawn up during the last ten days of August, was issued formally on 29 August.23
Primary objectives in MUSKETEER II remained much the same as in the earlier plan, with the significant exception that Formosa and the China coast were no longer mentioned. The initial KING operations still included a landing at Sarangani Bay on 15 November, with the main effort at Leyte Gulf scheduled for 20 December, but KING III, the new airborne operation, was included for 8 December. The LOVE series was reduced to two phases, the seizure of Aparri on 31 January 1945 and an airborne invasion of Mindoro on 15 February. The MIKE operations would begin with the main assault at Lingayen Gulf on 20 February instead of April, and MIKE II, the supporting operation planned for Dingalan Bay on 5-15 March, would be employed if needed.
Meanwhile, the Navy had begun to press for a definite directive about Formosa. On 18 August Nimitz, agreeing to all SWPA operations projected through Leyte and noting that he was prepared to cover them with the Pacific Fleet, asked Admiral King to secure a JCS directive for this SWPA effort and his own attack on Formosa and Amoy. He believed carrier attacks and land-based bombardment from Leyte could neutralize Luzon; if not, SWPA could move on Luzon during the CAUSEWAY operation. Nimitz admitted that he was having trouble reconciling variations in estimates of the required forces submitted by Richardson and Lt. Gen. S. B. Buckner, who would command the military expedition, but he recommended a directive for CAUSEWAY with a target date of 15 February 1945.24 King forwarded the message to Marshall with a request for JCS action.25 Naval planners seemed to have some hope that part of the required forces might be pried loose from MacArthur, and they feared that unless both Leyte and Formosa were coupled in one directive, MacArthur would so plan the Leyte venture as to make Formosa impossible.26 On 23 August, Nimitz reiterated his request for a firm directive on Formosa.27
Opinion in the War Department, meanwhile, had moved toward the SWPA point of view. On his way back to Washington, Hull had stopped in Hawaii long enough to discuss strategy with Richardson
and Lt. Gen. Millard F. Harmon of the Army Air Forces, Pacific Ocean Areas (AAFPOA).* Both had been in agreement on the RENO strategy, and Richardson had already written Marshall on 1 August that he and his staff did not see how Formosa could be supported logistically without Luzon. From observations on Saipan, Richardson did not believe that the Marianas could possibly support an invasion of the magnitude of CAUSEWAY.28 Ritchie had taken back to Washington a copy of the proposed revisions of MUSKETEER, which, with its earlier target date for Luzon, had pleased Marshall. In a teletype conference with Brisbane on 25 August, Ritchie urged Sutherland to send in an official message confirming the new target dates and to hurry some high-level SWPA staff officer to Washington for presentation of MUSKETEER II.29 Two days later, MacArthur confirmed the new target dates.30
On 28 August, Rear Adm. Forrest P. Sherman, Nimitz’ chief planner, explained the CAUSEWAY plan to the JPS. The scheme of operations now contemplated use of three Marine divisions against Amoy, on the south China coast, while two Marine and four Army divisions seized the southwestern part of Formosa instead of the whole island as originally had been intended. This change had been made in order to employ a minimum force, but estimated requirements were still large. Nimitz estimated that 424,000 men would be needed, Richardson 468,000, and Buckner 566,094. Sherman believed that carrier strikes, coupled with Kenney’s air efforts from Leyte and Seventh Air Force attacks from the Palaus, could keep Japanese air units in the Philippines “pounded down.” He considered SWPA operational plans up through Leyte to be “necessary and well coordinated,” but when asked his opinion of the feasibility of a direct attack on Kyushu instead of Formosa following Leyte, he doubted its practicability since it could be supported only by carrier and VHB aircraft. Even supposing the destruction of the Japanese fleet, Sherman did not believe a Kyushu operation would be sound “without shore-based aircraft which must be counted on to support the troops continuously over a sustained period.”31
At a JCS meeting on 1 September, Sherman urged the immediacy of Nimitz’ need for a firm directive, if the Formosa operation were to be undertaken on 1 March, when the weather would be most favorable. Marshall observed that an immediate decision would have to
* See below, pp. 507-12.
favor Luzon in view of available resources, and he proposed that it would be better to await future developments. King, on the other hand, urged at length that a Luzon operation would delay the war against Japan. Leahy suggested as a compromise that it might be possible to assemble supplies for either Luzon or Formosa and that immediate attention should be given to the directive for Leyte. The JCS directed the JPS immediately to prepare a directive on that basis.32 A draft directive submitted the next day proposed to inform MacArthur and Nimitz that a firm decision regarding Luzon or Formosa would be postponed. Meanwhile, SWPA, after conducting necessary preliminaries, was to occupy the Leyte-Surigao area on 20 December and establish air bases to support either a POA attack on Formosa on 1 March 1945 or its own invasion of Luzon on 20 February 1945. POA was to support the Leyte operation, submit plans for the invasion of Formosa and Amoy, and be prepared for assistance to a Luzon operation.33
King, not satisfied with this solution, presented his own draft directive to the JCS.34 He argued that seizure of Formosa promised psychological and material assistance to China, interdiction of Japanese sea traffic to the Indies, and the establishment of air, naval, and logistic bases for an attack on Japan proper. If the JCS were unable to issue a firm directive for Formosa, he recommended that they direct MacArthur to proceed with his operations through the invasion of Leyte and Samar and to develop bases on the former for containment of Japanese forces in the northern Philippines and for support of further SWPA and POA advances. Nimitz would be directed to provide fleet cover and support for MacArthur’s advances through Leyte and to prepare for a move into Formosa-Amoy on I March 1945 if the JCS directed it.
At a meeting of the JCS on 5 September, Marshall noted the considerable change between the original plan to take all of Formosa and the new plan to take only part of that island together with Amoy. Until a clearer picture was available, he could only agree to postpone the decision. Leahy, taking a broad view of the Pacific war, saw three possible courses of action: occupation of the Philippines to include Luzon, occupation of southwestern Formosa and Amoy, or occupation of southern Kyushu. Necessary forces were available only for the first, which also promised the smallest number of casualties; he therefore favored a strategy based on reoccupation of the Philippines,
together with an intensive bombardment of Japan by VHB’s and a rigorous air and naval blockade of the enemy homeland.35 Meeting again on 8 September, the JCS approved the directive which the JPS had furnished on 2 September, substantially as it had been drafted.36 MacArthur and Nimitz had won approval of their objectives in the northern Moluccas and the Palaus, and Leyte was scheduled for December. With this action, the JCS had finally evolved a directive for the initial invasion of the Philippines, but once more they had cautiously postponed a decision as to the main strategic objective.
Netherlands New Guinea Bases
Concurrently with the debate on strategy, SWPA’s Allied Air Forces had been engaged in moving its units into Netherlands New Guinea, where they would be in position to support the invasions of the northern Moluccas, the Palaus, and then the southern Philippines. Anticipated missions would not be simple: great distances – all over water – separated hostile concentrations which would have to be neutralized, and the Japanese air forces were showing signs of recovery from defeats at Wewak, Hollandia, and in the Marianas. Estimates on 31 July placed 860 operational aircraft (over half of them in the Philippines) within striking distance of the Allied invasion areas. It also seemed probable that the Japanese might bring naval aircraft equivalent to a carrier division to the Philippines by 15 September, making a grand total of 1,220 planes available for the defense of Morotai or the Palaus. If the enemy followed his usual procedure, he would divide his force and try to defend each area, but he could be expected to concentrate his main striking force in the southern Philippines, especially on bases around Davao.37
The Allied Air Forces, even without the expected help from Pacific Fleet carriers, could muster a twofold numerical superiority: on 4 August, FEAF had 2,306 serviceable aircraft, and RAAF Command, the other large component of the Allied Air Forces, had 413 serviceable aircraft – making a total of 2,719 planes (including 460 transports) available for operations. But this superiority was theoretical at best, because Allied air units were strung out all the way from Biak and Noemfoor islands to Guadalcanal and Australia, with the real center of gravity in the Nadzab-Manus Island areas, over a thousand miles from Morotai and even farther from the Palaus. Before the air support for Morotai and the Palaus could approach effective strength, the
combined striking power of the Fifth and Thirteenth Air Forces and of RAAF Command would have to be moved to Netherlands New Guinea.38
The general plan for this forward deployment assumed that the Fifth Air Force would serve as the assault force in initial Philippine operations, while the Thirteenth would concentrate at Sansapor, Morotai, and the Talauds. On leaving the Solomons and Admiralties, the Thirteenth would turn over its part of the commitment against bypassed areas to Aircraft Northern Solomons (AIRNORSOLS), a composite organization of U.S. Marine, New Zealand, and Australian air units commanded by a Marine airman.* Initially, RAAF Command would furnish an air garrison at Noemfoor and eventually displace Thirteenth Air Force units at Morotai as the latter moved into the Philippines. Forward progress of the Far East Air Service Command (FEASC) would closely follow the tactical situation. Utilizing its IV and V Air Service Area Commands independently, FEASC was to plan for complete evacuation of its service organizations from Australia by the second quarter of 1945. The IV ASAC, through December 1944, would be expanded to include all FEASC activity in New Guinea; the V ASAC would implant its units in the southern Philippines, moving its depots from Darwin to Sarangani Bay beginning in December 1944 and from Townsville to Leyte beginning in January 1945. Starting in June 1945, the IV ASAC would move its depots from Finschhafen and Biak to Manila, thus liquidating FEASC activity in New Guinea.39
Although SWPA forces had landed at Cape Sansapor on 30 July and would have Middelburg and Mar airdromes ready for tactical air garrisons within a few weeks, the most advanced Allied airfields on 1 August were on Noemfoor Island. There the engineers were rehabilitating, virtually rebuilding two airdromes captured from the Japanese. The RAAF 10 Operational Group, controlling the 78 Wing, was occupying the Kamiri airstrip. Kornasoren, scheduled for use as a heavy bomber airdrome, would be used by a U.S. air garrison under control of the 309th Bombardment Wing (H), but as yet only an advanced detachment of the wing and a detachment of the 419th Night Fighter Squadron had arrived. The site of a third Japanese strip – Namber – was to be used for unloading surface vessels during the winter season.40
* See Vol. IV, p. 647.
More extensive airfield development was under way in the Schouten and Padaido islands, approximately 100 miles east of Noemfoor. Facilities available on Biak early in August were Mokmer airstrip, being extended into a modified heavy bomber base; Borokoe strip, under development as a fighter-medium bomber field; and Sorido, to be built to serve transports and as an air depot. Because of prolonged ground fighting on Biak, engineer units had begun construction of a heavy bomber base on nearby Owi Island. Operational air units at Mokmer on 1 August were the 49th Fighter Group, the 475th Fighter Group, the 345th Bombardment Group (M), the 17th and 82nd Reconnaissance Squadrons, the 25th Photo Reconnaissance Squadron, and a part of the 2nd Emergency Rescue Squadron. On Owi were the 8th Fighter Group, the 43rd Bombardment Group (H), and the 421st Night Fighter Squadron. Patrols were being flown by PBY’s and PB4Y’s of Navy squadrons VB-115, VP-34, and VP-52. All units on Owi and Biak were under operational control of the 308th Bombardment Wing (H)41.
Farther down the coast of New Guinea-r 80 miles east of Biak – the 348th Fighter Group, a part of the 90th Bombardment Group (H),* and the 418th Night Fighter Squadron (B-25H) were based on the heavy bomber strip and dispersals which completely covered the small island of Wakde. These units were under the local operational control of the 85th Fighter Wing, which was itself directly subordinate to the 310th Bombardment Wing (M) at Hollandia. Only 275 miles east of Biak, but rapidly becoming a rear area, were the Hollandia, Sentani, and Cyclops strips at Hollandia. Here on 1 August were located a small headquarters detachment of FEAF and the headquarters of the 310th Bombardment Wing (M), which controlled the 3rd and 312th Bombardment Groups (L), a detachment of VP-33, and the 317th Troop Carrier Group reinforced by the 67th and 70th Squadrons of the 433rd Troop Carrier group.42
Movement to the forward bases of these tactical units with their supporting services had been fraught with extreme difficulties. SWPA had been forced to depend upon its own insufficient shipping, delays in the ground campaign on Biak had complicated projected shipping schedules, and a combination of both of these had delayed airfield
* The 90th Bombardment Group, although based at Nadzab, was flying missions through Wakde.
construction. Noemfoor was a reef-surrounded island, which made the unloading of cargo vessels there a tortuous task. By extreme effort, including maximum use of troop carrier C-47’s and bombers withdrawn from tactical operations, air echelons of units needed to support the Netherlands New Guinea campaign had been ferried forward; despite a loss of almost half of their combat efficiency because of poor maintenance and supply, they completed their assigned combat missions. Living conditions at advanced bases during the deployment were, without exception, bad. On Owi scrub typhus threatened to reach epidemic proportions until checked by impregnation of all clothing and blankets in a dimethyl phthalate soap solution. Quartermaster rations received were poor in quality and variety: C-rations, corned beef, canned salmon, dehydrated potatoes, canned carrots, canned chili, and canned sauerkraut was the diet of the 8th Fighter Group during July. In one shipment of 600,000 rations to Biak, two-thirds of the meat component was corned-beef hash. The water supply at Owi, a small coral island, depended on shallow wells and such supply of rain water as could be trapped from deluges so great that it seemed “to rain horizontally.” The brackish water, heavily chlorinated, was distasteful to drink and unavailable for bathing.43 Japanese night air raids plagued Biak and Owi, the latter island being so small that it created “the impression that a bomb hit anywhere was a near miss for everyone.” But, as reported by a fighter group on Biak, “Nomadic life had achieved the status of normalcy,” and cynics often asked, even as they cleared a new camp site in virgin jungle, “When do we move?”44
Difficulties with water transportation improved with time, and hard and ingenious labor bettered living conditions. During August and early September the Allied Air Forces built up to their planned strength in Netherlands New Guinea as quickly as shipping and facilities permitted. The fighter strip on Middelburg Island and the Cape Sansapor bomber strip at Mar were completed with nominal delays, and between 18 and 26 August, a detachment of the 419th Night Fighter Squadron and the 347th Fighter Group began operations on Middelburg. Ground echelons of the 18th Fighter Group arrived on 23 August, but Mar airdrome was not ready for its aircrews until 6-7 September. Ground personnel of the 42nd Bombardment Group (M) arrived on 24 August, but the aircrews of the group, held at Hollandia
to fly missions coordinated with the invasion of Morotai, did not arrive until 14-18 September. Arrival of the Catalina patrol squadron VP-33 completed the 13th Air Task Force garrison at Sansapor.45 On Noemfoor, both the Kamiri and Kornasoren air garrisons were augmented. Aircrews of the 35th Fighter Group landed at Kornasoren on 7 August to operate without a ground echelon pending movement to Morotai. The 348th Fighter Group (less its 342nd Squadron, left behind to fly cover for Wakde and Hollandia) with its new P-47D-23 type aircraft arrived on 26 August; on 3 September the group was joined by the newly activated 460th Squadron, making it the only four-squadron fighter group in SWPA. On 29 August the 868th Bombardment Squadron, XIII Bomber Command’s radar-equipped night-bombing B-24 unit, moved to Noemfoor to continue night attacks against the Palaus already inaugurated from Manus. The 58th Fighter Group unloaded on 30 August and received its flight crews on 6 September, and two days later the A-20’s of the 417th Bombardment Group began to move in from Saidor. By 15 September, 10 Operational Group at Kamiri airdrome had in place its 77 and 81 wings.46 Biak and Owi were built up as befitted their role as the major Allied air base in Netherlands New Guinea. During August and early September, Fifth Air Force Headquarters moved into Owi, allowing the 308th Bombardment Wing (H) a period of badly needed rest at Hollandia. Whitehead also pressed forward the two heavy bombardment groups remaining at Nadzab (the 90th and 22nd), moving them squadron by squadron as hardstands were completed, the official change of station being on 10-11 August. The 38th Bombardment Group (M) was operating from Borokoe strip on 31 August, and by 15 September all of the squadrons of the 6th Photo Reconnaissance and the 71st Reconnaissance Groups were in place on Biak. Upon completion of Sorido strip, the 3rd Emergency Rescue Squadron, just arriving from the United States for assignment to the Thirteenth Air Force, joined the 2nd Emergency Rescue Squadron there, and on 27 September the 375th Troop Carrier Group also took station at Sorido.47 Until he had located this tactical air garrison, Whitehead staunchly opposed any diversion of construction and shipping effort to air depot installations, despite FEASC’s objections that this was vitiating nearly one-half of its depot repair capacity. Beginnings were made on Depot No, 3 at Sorido during August. The 60th Air Depot Group arrived on 1 September and the 13th Air Depot Group on 20 October, and by November
the depot was nearly complete. By mid September it had begun to operate, substantially improving the logistic situation forward.48
As quickly as Fifth Air Force units cleared Wakde, the Allied Air Forces secured permission, effective 15 August, to cease Thirteenth Air Force raids on Truk, Woleai, and other Carolines targets which had been supporting POA’s Marianas operations,* and began movement of the XIII Bomber Command, together with its 5th and 307th Bombardment Groups (H) into the island. Using their B-24’s and such C-47’s as could be obtained, the groups moved most of their personnel into Wakde on 12-24 August, With the XIII Bomber Command officially established at Wakde on 22 August, the detachment of the 85th Fighter Wing returned to Hollandia; similarly, the 418th Night Fighter Squadron, relieved of B-25 night intruder missions an 18 August, moved back to Hollandia to train with P-61’s. The air echelon of the 342nd Fighter Squadron, however, continued to fly cover at Wakde until 21 September, when it was permitted to join its parent group at Noemfoor,49
During the first week of September GHQ, SWPA, and the Allied Air Forces moved from Brisbane to Hollandia, the Allied Air Forces, without many of its RAAF representatives, becoming now almost completely identified with its American component-FEAF. During the middle of September an advanced echelon of Thirteenth Air Force Headquarters moved from the Admiralties to Hollandia, effecting the closer coordination desired by FEAF. Control of the residual tactical air garrison at Hollandia passed from the 310th Bombardment Wing (M), which was preparing to move to Morotai, to the 308th Bombardment Wing (H) on 3 September.50
Morotai and the Palaus
On 15 June, Admiral Halsey had been relieved of his South Pacific command and transferred to Pearl Harbor to undertake detailed planning for the invasions of the Palaus. By 14 July Nimitz had agreed to a new plan ( STALEMATE II), divided into two phases.51 First, POA forces would take Peleliu, Angaur, and possibly Ngesebus islands, all at the south end of the Palau chain and believed to be defended by about 9,700 Japanese. Fields would be rehabilitated on Peleliu for Marine fighter units by D plus 10, and a heavy bomber base would be built on Angaur by D plus 35. Air units would then neutralize the
* See Vol. IV, pp. 676-90.
approximately 27,000 Japanese troops remaining on Babelthuap and adjacent islands. Second, the POA forces would seize Ulithi Atoll (300 to 600 Japanese) and would take Yap Island (10,500 Japanese) for development into a fighter base to cover the fleet anchorage at Ulithi. This plan was incorporated into a formal CINCPAC operations order on 21 July, with 15 September as the tentative target date for the first phase.52
Fearing that Nimitz might postpone the Palaus invasions, Kenney and Whitehead had worked out plans to get SWPA to Mindanao without Pacific Fleet Support,53 but these plans were laid aside when Nimitz’ order confirming fleet and carrier support for the northern Moluccas invasion was flashed to SWPA on 7 July.54 Selection of a target area in the Moluccas had not been difficult. Since it was desirable to have the Allied air base as far north as possible and to avoid most of the 30,000 Japanese defense troops in the islands, Morotai, an island just north of Halmahera, lightly held and with a seemingly abandoned airstrip site on its southeastern end, was the logical objective. Actually, as was usual in SWPA, no adequate terrain intelligence was available; a scouting party put ashore during June had never reported. With the equivalent of six aviation engineer battalions the GHQ engineer believed that air facilities could be constructed in southeastern Morotai within twenty-five days. GHQ issued a warning order for Operation INTERLUDE on 21 July, assigning it the target date of 15 September.55
Matters requiring intertheater coordination were resolved when a POA delegation met with SWPA planners in Brisbane on 27 July. It was agreed that the Third Fleet, organized around Task Force 38 under Vice Adm. Marc A. Mitscher, would move upon the Palausfrom the direction of Emirau, commence long-range fighter sweeps against Yap and the Palaus on the afternoon of D minus 9, and hit these targets with sustained attacks on D minus 8. Leaving one fast carrier group and three escort carrier divisions to continue neutralization of the Palaus, Task Force 38 with its other three fast carrier groups would attack Mindanao airfields from D minus 7 to D minus 5, concentrate against southern Mindanao airfields from D minus 3 to D minus 1, and detach one fast carrier group to give support at Morotai on D-day. SWPA agreed that the Allied Air Forces would 1) support POA operations by an intensive bombardment of the Palaus from D minus 40 to D minus 10 and with nightly attacks until D minus 1;
2) cover the approach of Task Force 38 by attacks on southern Mindanao airfields between D minus 12 and D minus 7; and 3) intensify attacks on enemy airfields west of New Guinea subsequent to D minus 7 for protection of the fast carrier group in its approach to Morotai. Allied Air Forces long-range patrol planes would also fly search missions to cover all approaches by the Third Fleet.56
Halsey issued his operations plan on 1 August, incorporating the fleet maneuver agreed on at Brisbane and outlining the amphibious and ground campaigns.57 Rear Adm. Theodore S. Wilkinson would command the Joint Expeditionary Force which would transport, land, and support the ground troops; Maj. Gen. J. C. Smith, USMC, would be in over-all command of the ground forces. Maj. Gen. R. S. Geiger, commander of the III Marine Amphibious Corps, would use the 1st Marine and 81st Infantry Divisions to seize Peleliu, Ngesebus, and Angaur, beginning on 15 September. Maj. Gen. J. R. Hodge, commander of the Army XXIV Corps, would use the 7th and 96th Infantry Divisions to seize Ulithi and Yap, beginning on 4 October. MacArthur’s operations instructions, issued on 29 July,58 directed Lt. Gen. Walter Krueger, commander of the Alamo Force (Sixth Army), to seize Morotai, beginning on 15 September. For this mission Krueger organized the TRADEWIND task force, consisting principally of Maj. Gen. C. P. Hall’s XI Corps, the 31st Infantry Division, and the 126th Regiment of the 32nd Division.59 Acting under Vice Adm. T. C. Kinkaid, commander of the Seventh Fleet and Allied Naval Forces, Rear Adm. D. E. Barbey, commander of the Seventh Amphibious Force, would transport, land, and cover theTRADEWIND force. For naval support Barbey would have two escort carrier divisions (six CVEs)borrowed from CINCPAC and the Seventh Fleet’s cruisers. The fast carrier group of Task Force 38 which would be at Morotai on D-day would cooperate with Barbey, but it would not be within SWPA command channels.60
To the Seventh Air Force, operating from central Pacific bases, fell the responsibility for the neutralization of Truk, Yap, Ulithi, other Carolines, Marshalls, Marianas, and Bonins. The burden of land-based preliminary bombardment fell to SWPA’s Allied Air Forces, which delegated most of the air mission incident to STALEMATE and INTERLUDE to its Fifth Air Force. The XIII Bomber Command, flying from Wakde under operational control of the Fifth Air Force, would neutralize the Palaus with intensive strikes until D minus 10,
and the 868th Bombardment Squadron, flying from Noemfoor, would continue to harass the Palaus nightly until D minus 1. More directly in support of INTERLUDE, XIII Bomber Command, taking over a task already under way, would seek to render Galela, Lolobata, and Miti airfields on Halmahera completely unoperational between D minus 8 and D minus 1. Aircraft Seventh Fleet search planes, under operational control of the Fifth Air Force, would extend 1,000-mile-long search sectors out of Owi to blanket the waters between the Philippines and the Palaus. The 13th Air Task Force At Sansapor and the RAAF 10 Operational Group at Noemfoot (both under Fifth Air Force operational control) would add weight to attacks on the northern Moluccas, neutralize each Japanese airfield remaining on the Vogelkop, and provide cover for the convoys and direct support to the ground operations on Morotai. The RAAF Command, using the 380th Bombardment Group (H) and other shorter-range units based in northwestern Australia, would continue neutralization of Japanese airfields on the Ambon-Boeroe-Ceram islands and on other islands in the Timor and Arafura seas; from D minus 2 through D plus 2 it would hold its forces in readiness to support the Fifth Air Force as ordered.61
In the Palaus, POA planned to develop airfields on Peleliu and Ngesebus suitable for short-range Marine units which would neutralize Japanese strength remaining in the islands, and on Angaur a 6,000-foot heavy bomber strip for use by the Seventh Air Force’s 494th Bombardment Group (H). Initial construction would be supervised by the ground task force commanders, but at the conclusion’ of the combat phase, island commanders – an Air Corps officer, Col. Ray A. Dunn, had been named Island Comrtiander, Angaur – would assume responsibility for the completion of outlined heavy bomber base facilities.62 SWPA specified that air facilities for Morotai would include a rehabilitated 5,000-foot fighter strip by D plus 2, construction of a 7,000-foot bomber strip by D plus 25, and completion of a third 6,000-foot strip (oriented or extension to 7,000 feet) by D plus 45. As usual in SWPA, the ground task force received control of all engineering effort during the combat phase, but Brig. Gen. Donald R. Hutchinson, commander of the 310th Bombardment Wing (M) and thus senior air commander, would be permitted to designate the airstrip sites, a procedure which had been permitted at Gusap and Nadzab in the fall of 1943 and much later at Sansapor.63
As soon as heavy and medium bombers could operate from Biak and
Owi bases and returning P-38’s could be sure of minimum facilities on Noemfoor, the Fifth Air Force began attacks on the northern Moluccas to cover the Allied landings at Cape Sansapor on 31 July.* Aerial photographs of the northern Moluccas, taken by high-flying F-5’s on 21-23 July, had revealed a substantial enemy air garrison and diligent enemy efforts to build new air bases.64 Although Whitehead suspected that the air garrison was defensive, he sent a combined heavy-medium bomber raid, covered by P-38’s, against Lolobata, Miti, and Galela airdromes on 27 July; the prize was some sixty Japanese planes. One P-38, with a mechanical failure, made a water landing en route home and its pilot was saved by a Catalina; otherwise, there were no Allied losses to a lethargic Japanese defense. Four P-38’s of the 433rd Fighter Squadron, making a long-range sweep of the waters northwest of Halmahera on 1 August, destroyed two Rufes and a Val, but after this date the Japanese evidently preferred to hoard their remaining aircraft. Against almost daily air attacks the aim of the Japanese gunners improved slightly, but as August wore on even the AA crews slacked up, firing usually at the first planes of a formation and then, evidently having saved face, taking to cover. Only three aircraft were lost to hostile action during August, all B-25’s, one of which was shot down by AA on 13 August and the other two planes collided during evasive action over Dodinga Bay on 15 August. Only weather gave the Japanese surcease from Fifth Air Force attacks. The weight of attack amounted to only about one-fourth of the total of 3,631 tons of bombs expended by the Fifth Air Force during August, but by 2 September there were no enemy aircraft operational on the Molucca airfields.65
Late in July, his supposition about the northern Moluccas confirmed by the lack of air opposition to the first Allied raid, Whitehead began to send missions against the airfields on the Ambon-Boeroe-Ceram islands. Although the area was assigned to the RAAF Command, Whitehead considered it a dangerous flanking threat to the line of the SWPA advance. On days when weather prevented missions to the Halmahera area, Fifth Air Force planes were turned into the Ambon-Boeroe-Ceram triangle, often after the missions were airborne. One harassed intelligence officer, after briefing a mission on six different objectives, complained that the requisite maps and target photos “looked like the first half of Cook’s travel pamphlet of the Dutch East Indies.66 The Boela oil fields and oil storage tanks on Ceram, first attacked
* See Vol. IV, pp. 661-70.
during July, continued to serve as secondary targets, and during August the Ambon-Boeroe-Ceram islands absorbed some 789 tons of Fifth Air Force bombs – an effort which was augmented by the 380th Bombardment Group from northwestern Australia. These missions were contested by enemy AA, and hostile fighters presented at least an incipient hazard, although the Japanese seemed to wish to save their planes, probably for night raids through Vogelkop fields against Allied bases. On 17 August fifteen to twenty Japanese fighters were flushed off Haroekoe airdrome on Ceram by a Liberator mission, and twelve P-38’s of the 80th Fighter Squadron shot down three of them with the loss of one P-38 which crashed while pursuing an enemy plane. At almost the same hour, eleven P-38’s of the 35th Fighter Squadron, escorting B-24’s to Liang airdrome on Ambon, contacted about eleven enemy planes and, despite the efforts of the enemy to avoid combat, shot down four Oscars and a Sally bomber. One of these P-38’s, having prolonged its flying time in combat, ran out of fuel on the return trip and ditched near Japen Island, but its pilot was saved by a PBY. Despite the success of this bombing effort in clearing enemy airplanes from the fields on Ceram, the Japanese were thought on 2 September to have forty-eight operational aircraft on Ambon and Boeroe.67
Fifth Air Force heavies and mediums, weathered out of both the Moluccas and Ambon-Boeroe-Ceram targets, commonly dumped their bombs on the Japanese airfields in the Vogelkop, which were the primary targets for shorter-range Fifth Air Force and RAAF planes flying from Noemfoor, Biak, and Hollandia. This largely unspectacular effort, which carried the heaviest tonnage dropped by the Fifth during August, was designed to prevent Japanese night raiders from staging through the Vogelkop fields, but a few picayune night raids during the month showed that it was nearly impossible to interdict all night attacks by such tactics as long as the Japanese had any airplanes within striking distance. Fifth Air Force B-25’s by day and B-24’s by night enforced a rigid antishipping blockade in the waters of the Vogelkop-Ceram-Halmahera triangle, but the fact that the Japanese were keeping their larger vessels out of the area denied the bombers notable success.68
During August planes of XIII Bomber Command undertook the strikes designed to soften the Palaus for invasion. Targets in those islands
had been previously attacked by Pacific Fleet carriers and Fifth Air Force B-24’s, but photos taken by F-5’s on 4-5 August showed that the Japanese were still maintaining thirty-six planes at the rough strip on Peleliu. The strips on Ngesebus and Babelthuap appeared serviceable but untenanted; nine Jake floatplanes were parked on a ramp at Arakabesan Island. Headquarters and supply buildings on Malakal and Koror islands (the latter the site of both the local military headquarters and that of the South Seas Bureau, the Japanese civil administration for the mandated islands) appeared battered but impressive, and each of the main islands was heavily defended by AA. Beginning on the night of 8 August, the 868th Bombardment Squadron began a series of nightly attacks from Los Negros against either Malakal or Koror targets and, prevented only by weather on 16 August, continued through 28 August. Moving to Noemfoor, the squadron flew nightly strikes against the Palaus during 7-14 September. With the loss of only 1 plane and 5 crewmen in an operational accident, the squadron sent 57 B-24’s and 91.2 tons of frags, demos, and incendiary bombs against the Palaus during these raids. Results were unobserved.69
By 23 August the 5th and 307th Bombardment Groups were getting into place at Wakde. That day the crew of a single B-24 photographing targets in the Palaus reported few enemy airplanes visible but observed so much small shipping in Malakal Harbor that XIII Bomber Command, scheduling its first raid for 25 August, devoted the 307th Group to shipping and ordered the 5th Group to bomb nearby Koror town. Two squadrons of the 307th placed a good pattern of 250-pound bombs across the harbor, but with few vessels there, only a barge and a small cargo ship were hit. The 372nd Squadron, seeing no targets, bombed the piers at Koror town, drawing seven Japanese fighters which badly damaged one of the B-24’s. The plane successfully limped back to Wakde, but two other B-24’s in the squadron collided and crashed while penetrating a weather front en route home. Planes of the 5th Group weathered heavy, moderate, and generally inaccurate flak to drop their 100-pound bombs in an excellent pattern over Koror town. Withdrawing from the target, the B-24’s were attacked by six Zekes and a Hamp. In a twenty-minute engagement two enemy planes were shot down, but Lt. Grant M. Rea’s B-24 caught fire and was getting out of control. To avoid a collision with others in
his flight, Rea feathered his propellers and dropped out of the formation. Two Zekes strafed the five crewmen seen to parachute from the stricken bomber.70
Thereafter, Japanese naval airmen in the Palaus proved no more anxious for combat than their fellow army pilots in the northern Moluccas. The 5th and 307th Groups, hindered only by weather and attacked only by AA, returned to the Palaus daily (except 27 August, when weather prevented) through 5 September. In a total of 11 missions, the 2 groups sent out 394 sorties, only 23 of which failed to reach the Palaus, and dropped 793.6 tons of bombs. AA remained dangerous, shooting down a B-24 on 28 August, destroying a second on 1 September, and damaging a third so badly on 2 September that it had to be ditched with only four survivors. These missions resulted in the destruction of most of the major installations and building areas in the Palaus, especially Koror town, where 507 buildings were demolished. While SWPA intelligence estimated that the Japanese still had twelve fighters, twelve floatplanes, and three observation aircraft in the Palaus on 5 September, the local airstrips had been so badly cratered that they could be made operational only with extensive repairs.71
Meanwhile, the Fifth Air Force had begun attacks on the southern Philippines to prepare the way for the Third Fleet carrier strikes. Late in July Kenney had informed Whitehead that attacks against Davao airfields and port installations should begin as soon as possible.72 Fearing that Japanese opposition to continued day raids on the area would become costly, Kenney had advised Whitehead to use his “snooper” force until he could stage a day attack in force. The 63rd Squadron initiated such night raids on 5/6 August with a rather ineffectual single bomber strike on Sasa airdrome, six miles north of Davao. Planes from the 63rd continued to harass the airdromes and harbor area around Davao during August, assisted by PB4Y’s on reconnaissance missions. Captured documents indicate that these heckling raids frequently killed small numbers of military personnel, sometimes destroyed aircraft, and often wrecked buildings. Japanese resistance was ineffectual. Even on 20 August, when the 345th’s B-25’s raided Beo and Rainis villages in the Talaud Islands to cover low-level photography by the 17th Reconnaissance Squadron, only two inquisitive Japanese fighters appeared and they seemed reluctant to attack. This daylight
raid was in easy fighter range of Japanese bases in southern Mindanao, but there was no reaction.73
Larger-scale attacks required more target information, and on 20-23 August F-5’s covered southern Mindanao. Surprisingly enough, 163 Japanese aircraft were revealed on the to August photos, while similar coverage on 22 August revealed only 108 planes. The Japanese, evidently reasoning that bombers would follow the photo planes, had withdrawn northward. Most of this strength was concentrated at Likanan airdrome (a four-runway base twelve miles northeast of Davao), at Sasa airdrome (north of Davao), and Matina airdrome (a bomber base two and one-half miles southwest of Davao).
Nine other airfields were in the area varying from operational to probably abandoned, the most important being Padada (thirty miles south-southwest of Davao), Daliao (under construction six miles southwest of Davao) , and Buayan (at the head of Sarangani Bay). There was a seaplane base at Bassa Point. Davao, the second largest city in Mindanao, sheltered many Japanese troop concentrations, and nearby Santa Ana contained the docks and waterfront warehouses for Davao.74
MacArthur questioned Kenney again on 21 August regarding the state of preparation for a raid on Davao, reiterating his interest in having the “big wallop” take place as soon as possible to prepare for the carrier strikes and to stimulate sabotage and guerrilla resistance.75 Whitehead had been hoping to use the two XIII Bomber Command heavy groups for an initial five-group raid. He had also been holding up the heavy attack until he could stage B-25’s through Sansapor for a simultaneous photo and strafing mission against the Sarangani Bay area. Because the disappearing Japanese air forces made target selection difficult, Whitehead asked for permission to bomb Davao, but MacArthur, having received word that the Japanese hoped to exploit such attacks for purposes of propaganda, limited attacks in the Philippines to airfields, hostile installations, and shipping.76 Harbor installations which might be of use to the Allies were to be spared as much as possible, and villages and cities were not to be bombed except with the express permission of GHQ.77
The tactical situation and state of the airfields at Sansapor limited the Fifth Air Force to its own heavy bomber resources, but on 1 September fifty-five B-24’s of the 22nd, 43rd, and 90th Groups bombed dispersal areas at Matina, Likanan, and Sasa dromes. Over Matina AA
was intense, heavy, and accurate, shooting down two 22nd Group bombers; one burst killed a pilot and a gunner in another of the group’s planes. Some ten enemy fighters attacked the 43rd Group over Likanan, holing several of the bombers at a probable loss of two fighters. The three groups, with each plane loaded with 20-pound frags, had attempted to knock out dispersed aircraft; strike photos showed that they had destroyed twenty-two. Three squadrons of P-38’s, staging through Sansapor, had accompanied the B-24’s to within sixty miles of the target, but they had been turned back by a weather front. The next day the same groups, this time with effective fighter escort, flew back to clean up stores and personnel areas around the airdromes with 100-pound bombs. Twelve B-24’s of the 43rd Group bombed stores at Bunawan, near Likanan; 22 B-24’s of the 90th Group dropped 196 bombs on the barracks at Likanan with good success; and 24 B-24’s of the t2nd Group dropped 216 bombs upon supply and personnel areas in the vicinity of Lasang, northwest of Likanan. Both AA and the several phosphorous bombs dropped in the vicinity of the 2td Group by high-flying enemy fighters were ineffective. Twenty-one P-38’s of the 9th’ 35th, and 36th Fighter Squadrons prevented any closer interceptions, and the 35th Squadron pursued and shot down a Zeke and a Lily. The V Bomber Command attempted a B-25 shipping strike at Davao harbor on the night of 2/3 September, using six volunteer crews of the 345th Bombardment Group; they staged through Middelburg, but only one of the B-25’s managed to reach the harbor and it scored no more than a near miss on a merchant vessel.78
The coordinated medium and heavy bomber raid on southern Mindanao took place on 6 September. Forty-five B-24’s of the 22nd, 43rd, and 90th Groups bombed the Santa Ana dock area with 1,000-pound bombs, while 11 B-25’s from the 345th Group, which had staged to Middelburg the day before, made low-level strikes on Buayan airdrome. Although they found few planes on the field, the B-25’s destroyed 4 of them and dropped their 100-pound parademos on barracks, warehouses, and hangars. The whole area was thoroughly strafed, and numerous Japanese soldiers were chased to cover with .50-caliber bullets. Twenty-nine P-38’s of the 431st and 433rd Squadrons, covering the Liberators, had an uneventful trip, but thirty-five P-38’s of the 68th and 399th Squadrons, escorting the Mitchells, strafed Buayan as the bombers departed, and the 399th Squadron shot down a
Topsy transport which tried to land at Buayan. After these 6 September strikes Whitehead, at the request of the Third Fleet, suspended all missions against southern Mindanao. The bomb tonnage dropped had not been as large as that desired by MacArthur; counting “snooper” and PB4Y tonnages as well as the daylight raids, only 366 tons had been dropped. But when the Third Fleet’s carrier planes appeared over the area, they would find that the Japanese air forces had deserted their bases, evidently because of the Fifth Air Force action.79
Having completed the missions against Mindanao, the Fifth Air Force turned its medium and heavy bombers to the Celebes. This strangely formed island, roughly the shape of a “K’ with the vertical stroke looped over the whole letter, lies between the Moluccas and Borneo. Although a single island it had been given a plural designation by early explorers who were puzzled by its peculiar conformation. To reach its western extremities would tax the range of B-24’s from either Biak or Darwin, but its most important installations were located in the northeastern and southeastern peninsulas where even B-25’s, staged at Noemfoor, could attack them. On the long northeastern peninsula which curls over the whole island, centering around Menado, the Japanese had built Langoan, Mapanget, and Sidate airfields. Japanese garrisons and some industrial activity had been noted in the towns of Menado (also headquarters of the Second Area Army), Gorontalo, and Tomohon. At the extremity of the northeastern peninsula, Lembeh and Bangka straits provided shelter for shipping, and Amoerang Bay, on the north coast of the northeast peninsula, was a shipping center. Less was known about the southeastern peninsula, but in addition to the old airfields at Kendari and Pomelaa the Japanese had developed five new airfields in the area-Baroe, Boroboro, Tiworo, Ambesia, and Witicola. As of 1 September the Japanese were believed to have 177 planes in the Celebes.80
A few B-24’s and PB4Y’s had bombed the Celebes earlier in August, but the first large-scale effort against the area was flown on 24 August by thirty-six B-25’s of the 345th Bombardment Group, staging through Noemfoor. This mission successfully attacked merchant shipping in the Bangka and Lembeh straits, damaged the mine-layer Itsukushirna with near misses, and strafed and bombed storage areas at Lembeh.81 Except for reconnaissance planes and night-flying B-q’s, the Celebes went free until 2 September chiefly because weather held off scheduled 345th Group strikes. On that date thirteen B-25’s from
the group tried to attack Langoan airfield, but when their fighter cover did not appear on schedule, the B-25’s once again attacked shipping in the Lembeh Strait. On this mission the AA positions along the straits, aided by gunners on the damaged mine-layer, put up a curtain of flak which veteran pilots said was the most intense seen since Rabaul; two B-25’s were shot down and two others so badly damaged that they were forced to land at Middelburg. This mission showed that the Celebes were too well defended for medium bombers. On the next day thirty-seven B-24’s of the 22nd and 43rd Groups were sent to bomb the targets adjoining the Lembeh Strait: twenty-two 90th Group B-24’s bombed dispersal areas at Langoan, destroying thirteen Japanese planes on the ground. Fighter cover was again delayed by weather, and the Japanese intercepted each of the three groups. Over the strait the interceptions were not closely pressed, but the 22nd Group shot down two Tonys. The 90th Group was hotly contested over Langoan by some ten Zekes, Tojos, and Hamps, losing one B-24 but claiming two Japanese fighters in exchange.82
After these initial raids Japanese resistance wilted and Allied fighter cover began to function properly. On 4 September, when all strikes from Biak and Owi against the Celebes were canceled because of weather, twenty-three B-24’s of the 380th Group from Darwin made a night attack on Kendari airdrome, setting a number of fires. On 5 September, fifty-eight B-24’s from Biak and Owi, in a mission described as no more exciting than “taking a nine and a half hours ... bus trip,” returned to blast Langoan’s dispersals, destroying or badly damaging seventeen Japanese aircraft on the ground. On 7 September fifty-three B-24’s hit warehouses, factories, and army headquarters at Menado, and the next day forty-five B-24’s bombed Langoan town and airdrome. A series of heavy bomber missions was flown against Mapanget, Langoan, and Sidate on 7-14 September, designed to knock out their strips with 1,000-pound and 2,000-pound bombs. Reconnaissance photos taken on 14-15 September showed that each of the three runways was so badly cratered that the few remaining Japanese planes in the northeastern Celebes could not be flown from them. At each airfield, however, the Japanese were busily filling up the craters, and the fields would require continuing attacks to insure the safety of Morotai. In all, this neutralization effort had required some 1,389 tons of bombs prior to 15 September.83
While FEAF long-range planes had been committed to attacks on
the Palaus, southern Mindanao, and the Celebes, other shorter-range Allied units had been attempting to destroy the enemy airfields on the Vogelkop and on the left flank. Fifth Air Force P-38’s, P-47’s, A-20’s, and B-25’s – unopposed except for AA which shot down a B-25 over Namlea and two A-20’s at Amahai on 10 September – repeatedly raided the enemy airfields on Ambon, Boeroe, and Ceram, dropping a total of 303 tons of bombs on these targets during the first 2 weeks of September. The 13th Air Task Force P-38’s from Sansapor flew 172 P-38 sorties and dropped 119 x 1,000-pound and 53 x 500-pound bombs on the 2 Namlea fields during the same period. Missing only two days during 5-15 September because of weather, RAAF 22 and 30 Squadrons’ Bostons and Beaufighters, flying from Noemfoor, kept Boela’s airstrips cratered. Limiting their load to two 200-pound bombs, RAAF P-40’s from Noemfoor stretched their radius to the Kai Islands for half-hour attacks on shipping and targets of opportunity. The 380th Group from northwest Australia continued its campaign, concentrating after 8 September on Ambon airfields. Following the 4 September night strike on Kendari, it hit Lautem town, Timor, on 2 and 12 September, Kai Islands dromes on 8 September, and Laha airdrome on Ambon on 10 September. At the same time, the short-range air units dropped some 676 tons of bombs on the airfields and hostile installations bypassed in the Vogelkop, displaying a great deal of ingenuity in running up the tonnage. A-20’s from Hollandia, for example, were accompanied to Utarom by B-25’s and bombed from medium altitudes on the B-25 lead ship. Other than two P-38’s lost over Jefman to AA on 4 September (one of the pilots was rescued) there were no casualties. The importance of this effort was emphasized by the continued Japanese ability to sneak night raiders into Netherlands New Guinea. Taking advantage of moonlight, two raiders killed one man, wounded seven, and damaged five planes at Mokmer airdrome on 2 September; five days later, two raiders wounded one man and damaged five bombers on Owi. On 9 September about ten hostile planes raided Biak and Owi, killing three men and wounding twelve. Allied AA shot down two of these planes, a P-61 destroyed a “bogie” over Biak on 1 September, and other Allied night fighters drove off hostile planes over Sansapor on 10/11 September; but it was evident that the Japanese were still able to harass the Allied line of advance toward Morotai.84 During the first two weeks of September FEAF planes unleashed a
stepped-up aerial attack on the northern Moluccas which sought to immobilize the Japanese garrison, destroy their logistic establishments, and knock out their airfields. The 5th and 307th Bombardment Groups, beginning on 7 September and continuing through D-day, sent 297 B-24 sorties to drop 881 tons of bombs almost exclusively on the enemy airfields. Concurrently, the Fifth Air Force used every type of combat aircraft available to soften the northern Moluccas and to isolate the island of Morotai. P-47’s and P-38’s dive-bombed villages, supply dumps, and airfields; B-25’s made minimum-altitude sweeps, bombing and strafing airfields and other targets of opportunity up and down the coasts of Wasile Bay; and Noemfoor-based A-20’s of the 417th Bombardment Group attacked Kaoe airdrome dumps on 11 September, losing one plane to AA-the only casualty of the two-week period of attacks on the Moluccas. To provide full measure of airdrome neutralization, Fifth Air Force heavy bombers attacked Kaoe airdrome on D-day, followed by thirty-one A-20’s of the 417th Group which swept the field and the adjacent town. Altogether, Fifth Air Force units had put 362 tons of bombs into the Moluccas during 1-15 September. Morotai was hardly touched, partly to avoid giving away the target and partly because there were no Japanese installations worth much air effort on the island. The 38th Group sent their B-25’s for minimum-altitude attacks against Morotai on 2 and 6 September, strafed and bombed a supposed radar installation on the northeast coast on 12 September, and swept the invasion area for the last time on 13 September. On D-day, shortly after the landings began, 2 B-25’s sprayed the area, not with the bombs and bullets usual at landings in enemy territory but with 460 gallons of DDT insecticide.85
Meanwhile, the magnificent success of the Third Fleet had guaranteed Allied victories in the Palaus and Moluccas. Halsey had sent Task Group 38.4 to strike the Bonins between 31 August and 2 September; it had attacked Yap on 7-8 September; and it had arrived in the Palaus in time to relieve the other three groups of Task Force 38, which had been operating there on 6-8 September. These three groups, beginning attacks against Mindanao on 9 September, met next to no opposition in areas which had been bombed by the Fifth Air Force, and Carney, Halsey’s chief of staff, wrote Sutherland that the “damned 13th Air Force has just about spoiled the war for our carriers, particularly at Yap. ... “86
Continued success against limited opposition
next day caused Halsey to order Task Force 38 against the Visayas and Luzon; achieving tactical surprise over the Visayas on 12-13 September, it destroyed more than 300 Japanese planes. As agreed with SWPA, Task Group 38.1 was detached and moved southward for support at Morotai, striking Zamboanga, the Talauds, and Menado en route. But Halsey still was not through: having replenished Task Force 38 (less Task Group 38.4), he sent it against Luzon on 21-22 September. Before withdrawing, these groups repeated attacks against the Visayas and staged a long flight to catch Japanese shipping which had fled from Manila to Coron Bay. The total damages claimed against the enemy between 31 August and 24 September were phenomenal: 1,000 Japanese planes destroyed and over 150 ships sunk, at a cost of only 114 U.S. planes.87
These successes, added to information brought back by a carrier pilot rescued from Leyte that there were no Japanese on the island led Halsey to make a startling proposition on 13 September. He suggested to Nimitz that he be allowed to cancel all of STALEMATE II except the capture of a fleet anchorage at Ulithi. He wished to turn over the forces so released to MacArthur for an immediate assault on Leyte which he would cover with carrier aircraft until airfields could be built ashore. This proposal, based on an over quick estimate of damages and erroneous intelligence from Leyte, might have succeeded, but, as events would show next month, only at extremely hazardous risks. Neither Nimitz, MacArthur, nor the JCS proved willing to eliminate the Palaus, but all were willing to release the forces required for the Yap phase of STALEMATE II so that they could be used by SWPA for the occupation of Leyte – with a target date set at 20 October instead of 20 December, and with all MacArthur’s planned operations between Morotai and Leyte canceled.88
Preparatory operations for the landings in the Palaus had already begun when Halsey made his proposition to cancel STALEMATE. During the first week in August the 81stDivision had been shipped from Hawaii to Guadalcanal, and, joining the 1st Marine Division in the Russells, it had been integrated into the III Amphibious Corps. On 4-8 September this task force had sortied from the Solomons; the Fire Support Group and the Escort Carrier Group (ten CVE’s) had departed in time to arrive at the target area by 12 September, when, promptly at dawn, they had commenced bombarding Peleliu. Because of excellent Japanese camouflage (the Japanese persistently refused to
Palau Islands
return fire and reveal their positions), most of the naval shells had to be fired blindly into the island. Four torpedo bombers and four fighter squadrons from the CVE’s were in constant use. Covered by Allied Air Forces search-planes in its movement northward, the Western Task Force was standing off the Palaus on 15 September, ready for landings which proceeded as scheduled; ashore without unusual incident, the 1st Marine Division quickly captured the Japanese airfield.
On 17 September, as soon as it could be determined that it would not be needed on Peleliu, the 81st Division, less one regimental combat team (RCT) which Halsey ordered held for the capture of Ulithi, attacked and easily overran the Japanese defending Angaur, ending all organized resistance on the island by 20 September.89
he ground advance on Peleliu, however, slowed up as the Japanese were pushed back into a series of fortifications honeycombing the rocky ridge – the Marines called it “Bloody-nose Ridge” – forming the backbone of the island. Against such cave fortifications, air support missions, flown by carrier pilots who had been primarily trained for attacks against naval units, offered only limited assistance. They tried napalm incendiary bombs to burn the Japanese out of their positions with small success, partly because many pilots dropped the napalm containers from too high altitudes and partly because the inflammable mixture was in too thin a solution. An AAF observer at Peleliu also noted that the carrier pilots began to strafe from 4,000 to 5,000 feet and pulled out of their diving angle at about 2,000 feet; dive bombers recovered at 3,000 feet as a safety measure from ground fire, although such fire usually ceased during an air attack. Relative ineffectiveness of such air support meant that the Marines had to root out and destroy Japanese positions, usually in expensive hand-to-hand combat. On 23 September one regiment of the 81st Division had to be brought to Peleliu to relieve the weary marines.90
Elsewhere the troops met little difficulty. Aforce of Marines seized Ngesebus on 25 September, while the 323rd Regiment of the 81st Division, supported by the escort carrier group, occupied Ulithi Atoll on 23 September without opposition other than extensive minefields. Halsey also seized Kossol Passage, and had it swept for possible use as a fleet regulating point. Although there would still be some mopping up of isolated enemy pockets, Halsey on 13 October turned the area over to Vice Adm. John H. Hoover, commander of Forward Area Central Pacific. Casualties as of that date had been very high: against 10,500 Japanese on Peleliu-Ngesebus and 1,500 on Angaur, the 1st Marine Division had lost 5,031 men killed, wounded, and missing, and the 8rst Division had sustained 1,911 casualties.91
Because of the separation of the two objectives in the Palaus, engineer troops were attached directly to the two combat divisions for the execution of CINCPAC base-development plans until the function could pass to the garrison or island commanders. The III Amphibious
Corps engineer found this procedure “most objectionable” since the combat divisions had little time or personnel for construction. The 1st Marine Division, with two naval construction battalions attached, was assigned the task of rehabilitating the Japanese fields on Peleliu and Ngesebus. The once-impressive airdrome on Peleliu was found in shambles, but despite initial delays in beaching the LST’s containing most of their heavy equipment, the Seabees had one runway ready to receive Marine fighters on D plus 8 and for Marine bombers by D plus 20. Ngesebus airfield was so poorly built that it was not worth diversion of effort from Peleliu.92
The 1884th and 1887th Engineer Aviation Battalions, initially attached to the 81st Division, had the more difficult assignment of building facilities on Angaur for a heavy bombardment group within thirty days. On 17 September, the day of the invasion, the 1884th sent a surveying party with a combat bulldozer to begin clearing a center line for the strip, even though the area was still under fire. The whole area was covered with dense jungle growth, and a six-inch to two-foot deep ground surface of humus had to be stripped and replaced with coral. Nevertheless, a steel-mat runway, service apron, and warm-up area-minimum facilities for flying-were completed by 20 October. According to construction directives, a 32,000-barrel gasoline tank-farm should have been completed by D plus 30,but a critical tie-up in unloading over the beaches delayed this work, and not until 8 November did the storage capacity reach 12,000 barrels, an estimated week’s supply for a heavy bombardment group.93
Deployment of the air garrisons to the Palaus, after the area was transferred to Forward Area Central Pacific, became the responsibility of Maj. Gen. Willis H. Hale, commander of the Shore-Based Air Force Forward Area. He supervised Maj. Gen. J. T. Moore, USMC, who was immediately responsible for land-based air operations in the Palaus. Prior to the arrival of land-based squadrons, three tender-based patrol squadrons and one air evacuation squadron were moved to Peleliu on 17 September. As soon as the strip on Peleliu could support them, three fighter squadrons, one night fighter squadron, and one torpedo bomber squadron, all Marine units, were flown in to undertake the local defense and neutralization of Japanese forces remaining in the Palaus. The long-range striking force scheduled for Angaur – the Seventh Air Force’s 494th Bombardment Group – was unable to operate there in force until late November, although a Marine transport
Japanese Airfields on Morotai Island
squadron had begun to use the field somewhat earlier. Thirteen B-24’s of the 864th and 866th Squadrons, all the heavy bombers which could be maintained from limited gasoline stores, reached Angaur on 21-22 October. The other two squadrons would not arrive until November, and since the 494th was a new group never before in operation, its full effectiveness had to await shakedown missions and local orientation.94
The Palaus would thus be of no value in the aerial preparations for the invasion of Leyte nor for the fleet action which followed, but the field on Angaur did prove most useful as a staging and heavy bombardment base before the completion of the Philippines campaign.
Meanwhile, land-based air operations in support of the landings at Leyte Gulf depended on Morotai. Loading of the amphibious troops for the invasion of that island had begun during lace August at Aitape, Hollandia, Maffin Bay, Biak, and Sansapor. Despite the strain on slender SWPA resources caused by loadings at somany places, the two
attack groups had arrived at Aitape and Maffin Bay on 2-3 September for final rehearsals. The group at Aitape weighed anchor on I1 September, joined the other group off Wakde next day, and, keeping out of sight of enemy-occupied areas, moved up the coast of New Guinea. Cruiser and escort carrier groups joined on 14 September, and the combined force proceeded toward Morotai with four CVE aircraft on constant antisubmarine patrol. Using units from Biak, Noemfoor, and Sansapor progressively, FEAF maintained four P-61’s on dawn and dusk patrols and sixteen P-38’s on continuous day patrols over the convoy; P-61 night fighters, working in pairs out of Sansapor, covered the convoy all night on 14/15 September. All patrols were without incident. After excellent gun and air barrages, the 31st Division and the 126th RCT began unopposed landings at 0830 on 15 September, at once encountering the worst reef conditions ever met in any SWPA landing. Defying fissures which trapped their vehicles, the assault troops waded ashore, and by 1300 hours on D plus 1 they had established their perimeter defense line. Patrols were sent out to scatter small bands of Japanese, other parties seized offshore islands needed for radar sites, and the original perimeter was extended to include additional dump areas. By 4 October the Sixth Army terminated the ground campaign, reporting casualties of 30 killed, 80 wounded, and 1 missing against a Japanese loss of 104 killed and 13 captured.95
Direct air support for the ground operation was not needed. The fast carriers of Task Group 38.1 were released on D plus 1 without having been used, and the planes from the CVE’s, aided after D plus 2 by torpedo boat patrols, had only to maintain air patrols over Morotai and to break up any possible Japanese efforts to slip troops across the narrow channel separating Morotai from Halmahera. To extend the SWPA search pattern and provide air-sea rescue, the Allied Air Forces moved the Catalina squadron VP-33, based on the seaplane tender Tangier, to Morotai on 19 September. Kenney agreed to release four of the escort carriers on 25 September in order to give them time to prepare for Leyte, and after their departure P-38 patrols from Sansapor supplemented the air defense.96
High-level decisions projecting an invasion of Leyte on 20 October demanded the utmost speed in the development of air facilities on Morotai. With little knowledge of the terrain on the island, SWPA and FEAF had been vague about timing and inexact about specifications, but they permitted Hutchinson wide latitude in selecting airdrome
sites, in determining priorities for construction, and in calling forward air force units as facilities permitted. Engineer units coming into Morotai between D-day and D plus 15 were concentrated on the air facilities, except for one battalion which worked on roads and dumps. Survey of the abandoned Japanese field began on D-day, but the site, after some clearing, was less practicable than the good site at Wama and was set aside for crash landings. The strip at Wama, its completion delayed by torrential rains which prohibited work for 72 hours, had 4,000 feet of usable steel-mat surface on 4 October. On 21 September a second site had been located north of and parallel to Wama for a dual-runway bomber airfield, later called Pitoe. At Kenney’s request, GHQ approved extension of Wama to 5,500 feet and authorized a regulation heavy bomber airfield at Pitoe; by 17 October 7,000 feet of the south runway at Pitoe, 1 taxiway, and 36 heavy bomber hardstands were open. Effort was increased on 18 October because of news that the Third Fleet had withdrawn support at Leyte,* and by 20 October Wama was nearing completion, with runway, taxiway, thirty-two hardstands, and six service aprons serviceable. At Pitoe seven new hardstands and four new service aprons had been added in the three days following its opening.†97
The first air force units ashore on Morotai had been the signal air warning (SAW) support aircraft parties and fighter control units which accompanied the assault forces. The 310th Bombardment Wing (M) Headquarters arrived on D plus 3, and between D plus 4 and D plus 16 the ground echelons of the following organizations debarked: 8th Fighter Group, 38th Bombardment Group, 418th Night Fighter Squadron, 35th Fighter Group, 82nd Reconnaissance Squadron, and VB-101 and VP-146. Since FEAF had canceled the movement of the 17th Reconnaissance Squadron to prepare it for Leyte, these units completed the Fifth Air Force garrison. Of the Thirteenth Air Force units scheduled to take over at Morotai as the Fifth went north, only an advanced echelon of the air force headquarters had arrived on D plus 12. The 17th Photo Reconnaissance Squadron and the 5th and 307th Bombardment Groups (H) would not be brought into Morotai until after 20 October. Movement of these ground echelons into Morotai
* See below, p. 354.
† Men of the 931st Engineer Aviation Regiment, the 836th, 841st, and 1876th Engineer Aviation Battalions, would be surprised to learn that, by Navy account, the Morotai fields were “Seabee built.” See Capt. Walter Karig, et al., Battle Report, The End of an Empire (New York, 1948), p. 289.
was on the whole well ordered, although in the shipping shortage the Air Force had been forced to accept Liberty and transport ships for the movement, and unloading at the beachheads was difficult. The ground echelon of the 35th Fighter Group, for example, arrived on 27 September but could not begin to unload until 5 October. Cargo damage ran high; the advanced echelon of Thirteenth Air Force estimated that 15 per cent of its cargo was damaged beyond repair in unloading. After 15 October, with one floating Liberty-ship pier and two coral jetties completed, the unloading of the vessels became easier.98
These delays were not especially significant because all were ashore before facilities permitted aircraft to be brought forward. When Wama was opened to fighters on 4 October, Hutchinson immediately called up the P-38’s of the 8th Fighter Group; enough were on hand the first day to permit him to relieve the remaining CVE’s. The air echelon of the 418th Night Fighter Squadron, newly equipped with P-61’s, arrived next day. Headquarters and two of the flights of the 2nd Emergency Rescue Squadron moved to Morotai on 4-10 October. Air echelons of the 38th Bombardment Group, called forward by Col. John T. Murtha, Jr., who had relieved Hutchinson as 310th Wing commander on the 16th, arrived at Pitoe on 17 October. Two squadrons of the 35th Fighter Group brought their P-47’s to Wama the same day, and next day the aircrews of the 82nd Reconnaissance Squadron, flying P-40N’s instead oftheir old P-39’s, flew to Wama. By 19 October Murtha had located the Venturas of VP-106 and the PB4Y’s of VB-101 at Pitoe.99 Such was the air deployment at Morotai on the eve of the invasion at Leyte Gulf. Even though they correctly anticipated both attacks, the Japanese opposition to the twin invasions-with the exception of a tenacious ground defense on Peleliu-was meager. As early as 6 August the Fourteenth Area Army in Manila, charged with the defense of the Philippines, had predicted that Allied strategy would aim at the recapture of Mindanao via the Palaus and the northern Moluccas. The Japanese Thirty-fifth Army, defending the southern Philippines, had estimated on 9 September that Morotai and Talaud would be invaded within the month. Imperial Japanese Naval Headquarters in Tokyo on 7 September warned against an attack on the Palaus and the Moluccas, and a Japanese plane observed the invasion convoys about Noemfoor on 11 September. But there was little that the Japanese could do, and, according to postwar interviews, they virtually wrote off both objectives
in favor of a vigorous defense of the Philippines. Loss of carrier aircraft in the Marianas made defense of the Palaus impracticable, although the ground garrisons there were expected to put up a good fight. Japanese orders captured on Morotai revealed that during August and September the enemy had hoped to deceive the Allies into thinking the island was heavily defended; at another date they had proposed to “decoy the enemy to Morotai Island and destroy them.” The Fourth Air Army planned to defend the northern Moluccas by shuttling aircraft between Davao and Menado, attacking Morotai on each trip. Similarly, planes from Kendari and Makassar were to move to Ambon, operate against the invasion area, and land at Menado. According to a postwar interrogation, planes were actually being concentrated at Davao to effect the plan, when on 8 September a coast watcher erroneously reported an Allied landing in Davao Gulf and caused cancellation of the plan. At any rate, Third Fleet neutralization of Japanese air strength in the Philippines made such tactics impossible.100
Thus by necessity the Japanese were limited to sporadic air attacks on the Palaus, which, with the exception of the severe strafing of a destroyer on the night of 1 October, did little damage. Night attacks on Morotai were more vigorous, but they were never of sufficient force to endanger the success of the operation or use of the base. Since the island was mountainous to the north and surrounded by landmasses to the south, Allied radars could not operate effectively, and the Japanese raiders could sneak in and bomb the concentrated airfield area with little warning. Between 15 September and 1 February the Japanese sent 179 sorties, in 82 raids, over Morotai, mostly on moonlight nights between 0300 – 0500, a timing indicating that they staged from Ceram or the Celebes, landed on Halmahera fields (which, despite repeated bombings, the Japanese persistently repaired), and then took off about midnight for Morotai. Fifty-four raids caused no damage, but one notable raid on 22 November resulted in two men killed, fifteen injured, fifteen planes destroyed, and eight damaged. Altogether, nineteen men were killed, ninety-nine were injured, forty-two aircraft were destroyed, and thirty-three damaged. P-61’s and P-38’s, the latter with searchlight cooperation, were employed against the raiders, but the cramped maneuver area (the night fighters had to make their interceptions in the short time before the Japanese planes reached a gun-defended area) gave most of the twenty-six definite
kills to the Allied AA. After the end of January, the Japanese suddenly gave up their night attacks against Morotai, and there was only one more raid on the night of 22 March.101 By this time Morotai, obviously a poor base to defend, had fulfilled its purpose to the Allies.
The Balikpapan Raids
SWPA airmen had long coveted the Japanese their uninterrupted use of the refining and oil center at Balikpapan in Borneo, second in production only to Sumatra’s Palembang in the entire NEI. By September 1944 the Netherlands Military Oil Intelligence Service estimated that Balikpapan refineries were processing some 5,240,000 barrels of crude oil annually and were turning out diesel fuel, motor fuel, aviation gasoline, and lubricating oil. True, the Japanese had an estimated two years of fuel stores in the homeland, but a reduction in aviation fuel and “black oils” produced at Balikpapan would disrupt their military operations in the forward areas rather severely.102 After a few 380th Bombardment Group strikes on Balikpapan and Soerabaja during the late summer of 19/13, Kenney had noted that within two weeks the “Japs were short of aviation fuel at all of their fields from Ambon to Wewak and even at Palau and Truk.”103
Judging NEI oil installations to be “the finest and most decisive set of targets for bombing anywhere in the world,” Kenney had tried diligently to get some B-29’s assigned to him for operation from the Darwin area.104 Armed with the SEXTANT planning paper, he had built a VHB base at that place, and had urgently requested that the AAF initiate VHB attacks on the NEI from northwest Australia, a request which both Nimitz and MacArthur had supported.105 Giles, while in Australia, had been persuaded to propose that FEAF be permitted to borrow two XX Bomber Command groups from the CBI and to employ two XXI Bomber Command groups while they were awaiting movement to the Marianas,106 but this proposition, like Kenney’s request a month later for just two B-29 groups to bomb Balikpapan, had been refused by the AAF.107
Failing to get B-29’s, Kenney, Streett, and Whitehead made plans during August 1944 to employ B-24’s from Netherlands New Guinea bases against Balikpapan. They had originally intended to wait until Mar airdrome at Sansapor could be lengthened to take B-24’s. Streett, however, mistrusting the plan to place Fifth Air Force heavies in such
an exposed place, had suggested that he be permitted to employ the two Thirteenth Air Force heavy groups from Noemfoor in a series of strikes at Balikpapan. Kenney agreed to the proposal, and Whitehead offered to furnish the Fifth Air Force’s three heavy groups when Streett needed them.108 Streett accordingly moved an advanced echelon of his headquarters to Noemfoor and opened his command post there on 23 September; on 1 October he assumed command of the air garrison at Sansapor, using the XIII Fighter Command as the local headquarters there vice the 13th Air Task Force, simultaneously dissolved. XIII Bomber Command with its 5th and 307th Bombardment Groups moved to Kornasoren drome, Noemfoor, on 18-28 September.109
Planning the Balikpapan raids was complicated by the extreme distance of the target – 1,080 nautical miles – from Noemfoor. Considered by itself the distance was not an obstacle until a bomb load of 2,500 pounds and conservative amounts of ammunition and gasoline were included in the plans. After careful tests, the Thirteenth Air Force prepared rigid cruise charts for fuel conservation and allowed each aircraft 3,590 gallons of gasoline; 40 per cent of the normal ammunition load was permitted. The Fifth Air Force, to be called on for support on the first, third, and fourth raids, preferred to remove all excess weight from its planes – including armor plate and the lower turret – rather than sacrifice ammunition. Fighter cover, in view of the long distance from Sansapor (936 miles) and Morotai (845 miles) to Balikpapan, would not be immediately available, but the Thirteenth Air Force, long used to unescorted missions, was not particularly apprehensive at first. Take-offs at Kornasoren airdrome, which had only a single runway, would have to be managed with exceptional care if the groups were to be able to get to their rendezvous points on schedule. Finally, rescue services had to be coordinated: a submarine was to lie five miles off the coast of Borneo, and Catalinas were to orbit along the mission route to pick up downed crews.110
Selection of targets and bombs was carefully considered. The Netherlanders, who had built many of the installations, picked as primary targets: 1) the Pandansari refinery, a new and modern plant essential to distillation of aviation gasoline; 2) the cracking units, a central plant for the area upon which all the refineries depended for gasoline refining; and 3) the Edeleanu plants, required to produce sulphuric acid for the solvent treatment of aviation gasoline. Target folders, based on photos of these installations taken on 3 August, were prepared. On the
basis of MAAF experience at Ploesti and in Italy, Thirteenth Air Force operations analysts decided that 250-pound bombs would be the proper weapon, reasoning that they would provide wide coverage, fracture oil containers, and fire the installations. For the first two raids the Thirteenth Air Force would use these bombs, but thereafter would turn to the heavier ordnance carried by the Fifth Air Force from the beginning.111
Fifth Air Force units struck Celebes targets, twice raiding Kendari, while the Thirteenth was setting up its establishment on Noemfoor. While awaiting movement from Wakde to Noemfoor, the 5th Group sent several raids to the Ambon-Boeroe-Ceram area. Both efforts sought to beat down opposition along the route to Borneo. By 29 September the 5th had completed its movement to Kornasoren, and that afternoon the 90th Group staged its planes up from Biak; all was ready for the first strike to Balikpapan. At 0040 hours on 30 September the first B-24 of the 5th Group roared down the runway at Kornasoren and disappeared into the darkness. Planes of the 307th and 90th Groups followed at one and one-half minute intervals until seventy-two of the heavily loaded bombers were airborne. Nine hours later, between 0933 and 0940 hours, the twenty-three planes of the 5th Group which had made the group rendezvous were over the Pandansari refinery at Balikpapan. After a last-minute change in heading to avoid cumulus drifting over the target, they placed 60 per cent of their bombs on the Pandansari refinery. When the twenty-three B-24’s of the 307th Group arrived five minutes later, the refinery was almost completely covered, and seven planes bombed it by radar, five bombed the paraffin point, and eleven dropped their bombs through the undercast without aiming. The 90th Group’s twenty-three planes, last to arrive, found solid cloud cover over Balikpapan, and only one of the squadrons attempted to bomb the target, on ETA. Bomber crews had been briefed to expect a strong AA reaction, but no one had foreseen that the Japanese had been hoarding one of their best naval air units for the defense of Balikpapan. These pilots were both calm and experienced. Two of them had picked up the bombers at the coast of the Celebes and had flown with the Liberators to Balikpapan, keeping discreetly out of range, obviously spotting for the AA and perhaps co-ordinating fighter interceptions. In all, the 5th Group lost three B-24’s shot down by fighters or AA, while the 90th Group lost one B-24. The three groups claimed to have destroyed nine of the hostile fighters.
Despite the losses, this strike had demonstrated that long-range missions to Balikpapan were practicable; there had been few turnbacks, most of the planes having made their return to Noemfoor with a little gasoline, and the cripples had been able to land at Sansapor and Morotai.112
On 3 October the 5th and 307th Groups went back to Balikpapan. According to the operations plan, the two groups were to bomb separately, each divided into two twin-squadron sections. Each of the sections was supposed to go over the target in javelin-down formation from 13,000 to 15,000 feet. The 307th Group, leading for the day, managed to get twenty of its planes over Balikpapan, and most of them scored hits on the Pandansari plant, Some forty Japanese fighters intercepted, however, beginning five minutes before the B-24’s bombed and continuing all the way back to the coast of the Celebes. Assisted by AA, these fighters shot down seven B-24’s’ at a cost of twenty-four of their own number. In the hour-and-ten-minute interception, most of the B-24’s ran short of ammunition, and when the Japanese broke off, two planes were entirely out. Meanwhile, the 5th Bombardment Group’s planes, contrary to briefing, joined their sections closely at the rendezvous, whence they flew abreast, each section javelin-down and as close together as possible. The Japanese largely avoided this formation, and it encountered no losses. Eighteen B-24’s bombed the Edeleanu plant and one bombed the Pandansari refinery. As on the preceding mission most of the surviving planes returned to Noemfoor and crippIes to Morotai and Sansapor.113
Some change in tactics was evidently necessary, for never before had the 307th Group had such damages and losses. Streett, taking his cue from the success of the 5th Group, introduced each of his B-24 groups to the combat box formation. On 5-7 October he had each group fly two simulated bombing missions against P-47 interceptors, and he sold the crews on the new formation. Streett also borrowed the three Fifth Air Force groups for the next raid: they were to fly over the target at medium altitude and draw the AA while the Thirteenth Air Force went over at high altitude and took the fighters. Fifty P-47 pilots, perceiving the need for cover over such a hotly contested target, approached Kenney with a proposition that they be permitted to escort the heavies to Balikpapan, fight it out over the target and then fly back to a predetermined spot in the Celebes, bail out, and trust that they would be picked up by Catalinas. Fortunately, however, there
would not be need for such heroic measures, for experimental longrange fighter flights under way in the SWPA during the spring and summer would now pay off. By attaching a 310-gallon tank under one wing and a 165-gallon tank under the other, new model P-47’s from Morotai and P-38’s from Sansapor or Morotai could make the flight to Balikpapan with a little gasoline to spare. Such a long flight would be hard on the pilots, but, by rigid adherence to a cruise-control chart and by certain stimulants, they could both get to the target and maintain a sufficient degree of alertness to engage fresh enemy pilots in combat. It was planned that the 40th and 41st Squadrons, with their P-47’s, sweep the sky above Balikpapan clean of enemy fighters shortly before the bombers arrived; the 9th Squadron, with P-38’s, was to escort the bombers. In addition, 43rd Squadron “snoopers” were to harass the Balikpapan defenses and keep the Japanese pilots awake at night, and, shortly before the bombers were due to arrive, the 868th Bombardment Squadron was to have a “snooper” in the area dispensing “window” to dislocate the radar defenses and to get the Japanese fighters up on a false mission.114
The first phases of this mission took place on 8-10 October, approximately as planned. Seven 43rd Squadron B-24’s harassed Balikpapan on the night of 8/9 October and five on the night of 9/10 October. Between 0832 and 0900 on 10 October, an 868th Bombardment Squadron B-24 dropped 1,000 pounds of window on a course leading to within sixty miles of Balikpapan, and between 1010 and 1045 sixteen P-47’s appeared over Balikpapan at about 20,000 feet. Pouncing down on twenty-five to thirty-five hostile fighters, they shot down twelve of them at a loss of one P-47. According to schedule, the two Thirteenth Air Force groups should have gone over the target immediately prior to the Fifth Air Force groups; within ten minutes all of them were supposed to have completed their runs. Arriving at their rendezvous first, however, the Fifth Air Force wing delayed as long as possible and then proceeded to the target. The 90th Group with twenty-one B-24’s divided its bombs between the Pandansari and Edeleanu refineries, at the same time attempting to beat off twenty-five to thirty Japanese fighters. In this engagement one of the B-24’s was exploded by a phosphorous air-to-air bomb, but the group claimed sixteen Japanese fighters destroyed. The 22nd Group, with eighteen B-24’s reaching Balikpapan, placed most of its bombs on the Pandansari target, but a vigorous interception caused inaccuracies. One 33rd Squadron B-24 was so badly damaged by enemy 20-mm. fire that it crash-landed on
Batoedaka Island, but before going down it destroyed six of the swarm of Japanese fighters which were attempting to hasten its demise. A Zeke rammed another B-24 and caused it to explode, and one other was crippled by fighters, which then ganged up and shot it down. Altogether, the 22nd Group claimed nineteen hostile fighters definitely destroyed. By this time the Japanese venom was somewhat expended, and the nineteen B-24’s of the 43rd Group bombed the paraffin refinery and surrounding storage area without loss, claiming thirteen enemy planes destroyed. When the 5th Group’s 24 planes came over at about 20,000 feet, the Japanese were not eager and the gunners managed to shoot down only 1 of 15 making halfhearted passes. The 5th Group managed a good strike on the paraffin and lubricating oil plant. A few minutes later, twenty-five of the 307th Group’s B-24’s bombed the cracking plants, and neither the B-24’s nor seven to ten hostile fighters sustained any serious damage in the aerial engagement which followed. At the same time that the Thirteenth’s planes came over, the eleven P-38’s of the 9th Fighter Squadron were over Balikpapan. Of sixteen fighters intercepted, they shot down six with no losses; Maj. Richard I. Bong, leading AAF “ace” and supposedly in retirement as a combat pilot, had flown the mission and shot down his twenty-ninth and thirtieth victims.115
The next mission scheduled was virtually a repetition of this raid, with the addition of XIII Fighter Command P-38’s from Sansapor. Single “snoopers” attacked Balikpapan on the nights of 12/13 and 13/14 October, and the major attack was made on 14 October. Promptly at 1020 the high-altitude fighters, fifteen P-47’s of the 40th and 41st Squadrons, appeared over Balikpapan, shooting down nineteen Japanese fighters; the 41st Squadron lost two planes over the target, one of which was shot down by P-38’s. Two of the 40th Squadron planes failed to make it back to Morotai because of mechanical failures, but their pilots, as well as the pilot of the plane shot down by P-38’s, were rescued by Catalinas. The Fifth Air Force wing – a “V” formation made up of group diamonds and led by the 90th Group took forty-nine B-24’s, each loaded with one 500-pound and two 1,000-pound bombs, over the Edeleanu refinery precisely on schedule; a few planes had difficulty releasing their heavy bombs, but most of them were successful. Interceptions were not pressed vigorously, but the bombers claimed nine victories at the cost of one B-24. Another B-24 was lost from unknown causes somewhere en route to Balikpapan. Thirty-one 838’s of the 9th, 50th, and 432nd Squadrons, escorting
the heavies, accounted for sixteen hostile fighters. Forty-nine 5th and 307th Group B-24’s, flying in two-group boxes, passed over the paraffin and lubricating oil works a little late, but they plastered the target with 500-pound bombs. Interceptions being nominal, no Liberators were lost, but two hostile fighters were claimed as destroyed. The XIII Fighter Command had sent forty-mo qzqz P-38’s to cover these bombers, but only six-those from the 68th Squadron – got through to destroy two hostile fighters which were bothering the Liberators. This strike was the first really successful blow of the series, and MacArthur sent his commendation for a magnificent strike.116
This, however, was also to be the last effective strike against Balikpapan, for the last strike of the series, made by the 5th and 307th Groups on 18 October, was largely thwarted by weather. Fifty-two of their bombers found Balikpapan completely covered by clouds and dropped their bombs by ETA. Seventy-five XIII Fighter Command escorts started from Sansapor but only eight reached Balikpapan; since no Japanese planes appeared, however, there was nothing for them to do.117 This anticlimactic mission ended the Balikpapan raids, for the full strength of the Fifth and Thirteenth Air Forces was now required forward in the Philippines.
In five raids the Fifth and Thirteenth Air Forces had put 321 B-24’s over Balikpapan, had dropped 433.3 tons of bombs, and had provided 66 P-38’s and 30 P-47’s over Balikpapan to cover the heavies. In the process, they had lost twenty-two B-24’s, three P-38’s, and six P-47’s. Not all of the personnel on these planes, however, had been killed, for the Navy and AAF Catalinas and the submarine USS Mingo had been spectacularly successful in picking up the downed flyers: they had saved sixty combat crewmen. Possibly the greatest gain for the Allies was the resultant experience in long-range missions, experience which would be of value in the Philippines campaign. Intelligence officers recognized, however, that the bombing campaign had only scratched the oil targets at Balikpapan. The Pandansari refinery was out of action but could be repaired; the Edeleanu plants would require complete rebuilding; installations producing diesel oil and lubricants were damaged but could be repaired in a short time. The Japanese could, and would, be able to get the refineries operating again in a short time and without too large a reduction in their annual output.118 But by this time the Allies would have seized the bases in the Philippines needed to blockade the whole Netherlands East Indies.