Chapter 17: Operations on Noemfoor Island
Even before the HURRICANE Task Force had landed on Biak, planners at General MacArthur’s headquarters had considered the possible necessity for occupying additional islands in Geelvink Bay to protect the gains achieved at Biak. With the initial reverses of the HURRICANE Task Force, continuing enemy air attacks against Biak, and growing indications that the Japanese might be attempting to reinforce that island, expansion to near-by enemy bases began to appear an imminent necessity. Moreover, occupation of another island near Biak would provide an additional air base from which the Allied Air Forces could dominate the Vogelkop Peninsula, where another operation was soon to take place. If, as appeared likely during the first week of June, there was to be considerable delay in making the fields on Biak operational, the move to the Vogelkop might have to be postponed. On 4 June, therefore, General Headquarters began preparations for the seizure of Noemfoor Island, which lay about midway between Biak and the Japanese base at Manokwari on the northeast tip of the Vogelkop Peninsula.1
The Noemfoor Plan
Noemfoor was used by the Japanese as a staging area for troops moving to reinforce the Biak Detachment. Japanese barges could make the run from Manokwari to western Noemfoor, a distance of about 60 nautical miles, during the hours of darkness, and another overnight run would take reinforcements the additional 75 nautical miles to western Biak. Two other reasons dictated the choice of Noemfoor. First, the Japanese had constructed or partially completed three airdromes on the island. The Allies could quickly repair those fields for fighters and bombers which could cover the advance to the Vogelkop. Finally, Allied occupation of Noemfoor would deny to the Japanese a base from which the sea lanes west of Biak could be threatened.
The Terrain and the Enemy
Noemfoor, roughly circular in shape, consists of a series of limestone and coral terraces, a topography in many ways similar to that of Biak but not as rough. The highest point on Noemfoor is about 670 feet above sea level, its greatest north-south length is about fifteen miles, and its width approximately twelve and a half miles. Most of the natives, about 5,000 in all, lived in a number of coastal hamlets. The island had experienced little commercial development
before the war, there were no towns of any importance, and port facilities were lacking. Land communications consisted of native tracks skirting the coast line and connecting the many villages. (Map 17)
Inland, the terrain is fairly rugged, although there are some large flat areas on the northern and southwestern sections of the coast. Dense rain forest grows everywhere except along a few strips of the beach, and along the banks of tidal inlets where heavy mangrove swamps are found. The island is completely surrounded by coral reefs, behind which are beaches of many types—sand, forest, mangrove swamp, or rocky shore. Seaward approaches to the island are deep and free of navigational hazards, permitting large ships to draw up to the outer edges of the reefs both day and night. The best landing areas, considering both reef and beach conditions, are on the northwest shore, near Menoekwari West and Kamiri.2
It is not known when the Japanese first occupied Noemfoor, but early in 1944 they began to construct airfields there in accordance with their policy of strengthening their new strategic main line of resistance in western New Guinea. By the end of May the Japanese had completed two fields on Noemfoor: Namber Drome on the southwest coast and Kamiri Drome on the northwest. On the north-central coast, about four miles east of Kamiri Drome, the Japanese had left unfinished a third airfield, designated Kornasoren Drome. The only other construction of note was some widening of native trails to serve as motor roads between the airfields.
In mid-June, realizing that the Japanese were using Noemfoor as a way station for troops moving to reinforce Biak and considering it probable that the Japanese would expect an Allied invasion of the island, the ALAMO Force G-2 Section estimated that the Japanese garrison on Noemfoor comprised 2,850 to 3,250 troops, with a combat strength of 1,600 to 2,000 men. Most of the combat troops were believed to be members of the 3rd Battalion, 219th Infantry, 35th Division. Other combat units of that division and at least a company of the 36th Division were also thought to be stationed on the island. The Japanese were known to have concentrated their strength at the airfields, and the largest single body of Japanese was believed to be stationed at Kamiri Drome.3
Intelligence officers estimated that since the Japanese probably considered Noemfoor only a delaying position, they would make few efforts to reinforce it. No naval interference was expected because, after the fiascos of the KON and A-GO Operations, most of the Combined Fleet had retired to homeland or Philippine waters, leaving only a few destroyers and one light cruiser in the western New Guinea area. Some reaction by Japanese aircraft was expected, but not on the scale attempted at Biak. Allied aircraft could keep Japanese fields within range of Noemfoor out of operation most of the time and, in cooperation with Seventh Fleet PT boats, could stop most overwater
Map 17: Capture of Noemfoor, 2 July–31 August 1944
reinforcing movements of the scale the Japanese could mount.4
Allied estimates of Japanese strength on Noemfoor were too high, for there were not more than 2,000 Japanese on the island. Probably not more than 900 of these could be counted infantry effectives. In addition to the 2,000 Japanese, there were 600-odd Formosan laborers and approximately 500 Javanese slave laborers. The 3rd Battalion, 219th Infantry, contained the bulk of the combat troops, but there were also present about 180 men of the 2nd Battalion, 219 Infantry, and a like number of the 222nd Infantry, 36th Division, troops which had been unable to reach Biak. The Japanese garrison on Noemfoor was commanded by a Colonel Shimizu, who was also the commanding officer of the 219th Infantry. Another unit, commanded by a Major Mori, but apparently under Colonel Shimizu’s operational control, seems to have been a provisional organization containing mostly armed service personnel and numbering about 600 men.5
The ALAMO G-2 Section was correct in assuming that an invasion of Noemfoor would come as no surprise to the Japanese. Colonel Shimizu had estimated as early as 1 June that the Allies would soon invade the island in the vicinity of Kamiri Drome. For reasons unknown, the colonel later changed his guess and began making extensive defensive preparations along the north coast at Kornasoren Drome. During the night of 22–23 June, the colonel’s men discovered an Allied reconnaissance party which, brought to Noemfoor by PTs, was attempting to obtain accurate information concerning reef and tide conditions off Kamiri. When no American landing was immediately forthcoming, Colonel Shimizu re-estimated that the invasion of Noemfoor would occur at the Kamiri Drome area during the first week of July.6 His estimate was soon to prove remarkably accurate.
Organization of the Forces
On 5 June 1944 General MacArthur informed General Krueger that the Noemfoor operation was to take place under ALAMO Force direction. At the same time General Krueger was warned that operations on the Vogelkop Peninsula were to follow close upon the occupation of Noemfoor. General MacArthur’s planners wanted to use a regimental combat team of the 6th Infantry Division at Noemfoor,7 but General Krueger planned to employ that division for the Vogelkop invasion, and therefore secured General MacArthur’s approval to send the 158th Regimental Combat Team against Noemfoor. The latter unit was then at Wakde–Sarmi, and provision had to be made to speed shipment of the 6th Division
there so that the 158th could prepare for the Noemfoor operation.8
As in previous operations along the New Guinea coast, General Krueger was made responsible for coordinating air, naval, and ground planning for Noemfoor. He scheduled an interservice planning conference for 16 June, a date to which Allied Naval Forces planners objected. Most of the important naval planners were at least a day away from ALAMO Force headquarters, and, moreover, the Allied Naval Forces did not believe that enough intelligence data was on hand for planning purposes. General Krueger, on the other hand, considered available data sufficient, and held the conference as scheduled. The Allied Naval Forces was represented only by one officer from the staff of the VII Amphibious Force, Seventh Fleet.9
At the conference it immediately became apparent that, as usual, the provision of air support for the landing was the major problem. Allied Air Forces held that since carrier-based aircraft were not available, a minimum of two fighter groups would have to be based at Biak before the Noemfoor operation could be initiated. General Kenney, the air commander, estimated that Mokmer Drome could be used by 20 June without further interference from the Biak Detachment and that the airfield then being constructed on Owi Island, off southeastern Biak, would be ready on 25 June. Upon completion, each field would be able to accommodate one group of fighters. These aircraft would need a few days to become acquainted with the target areas on Noemfoor. Finally, while the concentration of troops and supplies for ground elements scheduled to land on Noemfoor could be accomplished by 26 June, it would be necessary to allow these forces time for rehearsal, and more time would be needed for loading. The combination of these factors led General Krueger to believe that 30 June would be the earliest practicable date upon which the Noemfoor landing could take place. This date was approved by General MacArthur.10
On 20 June a second planning conference was held at General Krueger’s headquarters, this time attended by the commanders of all the major air, ground, and naval units participating in the Noemfoor operation. It was discovered that the 30 June target date was optimistic. First, additional time was needed for rehearsing the amphibious phases of the operation. Next, General Kenney wanted the landing delayed until a second runway could be completed on Owi Island. It still appeared possible that the Biak Detachment might delay full use of Mokmer Drome for some time, and the expansion of the Owi strip was believed necessary to provide a safe airfield for fighters supporting
the Noemfoor landing. The additional construction on Owi could not be completed until 30 June. Finally, more time was needed to move forward to the staging area (Wakde–Sarmi) a number of LCMs and LCTs which were to participate in the landing. Therefore, General Krueger secured approval from General MacArthur to postpone the landing until 2 July. With the date settled, other points of air, ground, and naval coordination were discussed by the conferees, and agreement on all important issues was soon reached. The commanders returned to their respective headquarters which, within a few days, produced the necessary field orders, operations instructions, or other final plans.11
For the Noemfoor operation the 158th Regimental Combat Team, reinforced, was designated the CYCLONE Task Force. The task force and combat team commander was Brig. Gen. Edwin D. Patrick, who had commanded the same combat team and the TORNADO Task Force at Wakde–Sarmi. The CYCLONE Task Force staff was formed by members of ALAMO Force headquarters, the 158th Infantry, No. 62 Works Wing of the Royal Australian Air Force, and other units attached to ALAMO Force for the operation. The total combat strength of the task force was approximately 8,000 men, of whom over 7,000 were to land on D Day.12
The principal mission of the CYCLONE Task Force was to seize airdrome sites which were to be quickly developed so that Allied aircraft could support operations west of Noemfoor. The task force was initially to prepare facilities for two groups of fighters and half a squadron of night fighters and later to expand these facilities for an additional fighter group, two squadrons of medium bombers, and two squadrons of light bombers. To accomplish the necessary airfield construction, service units of the CYCLONE Task Force were to include two American engineer aviation battalions and No. 62 Works Wing. With an excellent record of achievement at Aitape as engineer of the PERSECUTION Task Force, the commander of the Australian works wing, Group Capt. W.A.C. Dale (RAAF), was appointed engineer of the CYCLONE Task Force.
The service units assigned to the task force numbered about 5,500 men, among whom approximately 3,000 were to be engaged in airdrome construction. Other service troops included the usual medical, quartermaster, ordnance, and signal units needed for a small amphibious operation. Air force units scheduled to arrive on Biak after the airfields were completed totaled some 10,000 men. Air operations from Noemfoor were to be controlled initially by No. 10 Operational Group, Royal Australian Air Force, and later by the Fifth Air Force’s 309th Bombardment Wing (H). The first aircraft scheduled to operate from Noemfoor were to be Australian, and Fifth Air Force planes were to follow on a date
governed by the extent of airfield construction.13
The amphibious phase of the Noemfoor operation was to be directed by Rear Adm. William M. Fechteler as the Commander, Naval Attack Force.14 Admiral Fechteler divided his Attack Force into three groups. The Covering Force, under the command of Rear Adm. Russell S. Berkey (USN), contained 1 heavy cruiser, 2 light cruisers, and 10 destroyers. Admiral Fechteler retained command over the Main Body, consisting of 15 destroyers, 8 LSTs, 8 LCTs, 4 patrol craft (PCs), 1 tug, and 14 LCIs (including 3 rocket-equipped LCIs and 2 LCIs transporting demolition experts and their equipment for blasting approaches for landing craft through the reefs). An LCT-LCM Unit, commanded by Lt. Comdr. James S. Munroe (USNR), contained 3 PCs, 5 LCTs, and 40 LCMs, the latter manned by Company A, 543rd Engineer Boat and Shore Regiment. A Beach Party and two Reinforcement Groups (the latter arriving after D Day) completed the naval organization.15
The bulk of the air missions in support of the landing were to be flown by the U.S. Fifth Air Force, now under Maj. Gen. Ennis C. Whitehead. The newly arrived Thirteenth Air Force, under Maj. Gen. St. Clair Streett, was to have a share in the support, as were Australian and Dutch planes. Also available was Task Force 73, the Seventh Fleet’s land-based aircraft, which flew under the operational control of the Allied Air Forces.16
Initially, ALAMO Force Reserve for the Noemfoor operation was the 503rd Parachute Infantry Regiment. This unit, stationed at Hollandia, was to be prepared for airborne movement to Noemfoor on C-47’s of the 54th Troop Carrier Wing, Fifth Air Force. Since there were not enough C-47’s available to move the entire regiment at one time, provision was made to fly it forward in battalion groups.17 Late during the planning for Noemfoor, General Krueger, concerned lest the ground situation on Noemfoor on D Day require overwater reinforcements as well, alerted the 34th Infantry, then on Biak, to prepare for waterborne movement to Noemfoor on twenty-four hours notice. He requested and
obtained from the VII Amphibious Force the use of ten LCIs, which were to stand by off Biak at Mios Woendi pending a decision on the necessity for moving the 34th Infantry to Noemfoor.18
The CYCLONE Task Force had no specific reserve set aside as such for the landing. The Task Force Shore Party, however, consisting of engineer and quartermaster units, was to assemble the equivalent of three rifle companies for combat missions upon call from General Patrick. The reserve for the 158th Infantry was to be Company K, reinforced by a heavy machine gun platoon from Company M. The 1st Platoon, 603rd Tank Company, once ashore, was to assemble to assist the attack of the 158th Infantry and, until needed for this task, could be considered as a mobile reserve.19
Logistics and Tactics
The logistic plan for Noemfoor was similar to that for earlier operations along the New Guinea coast. As usual, the transportation of men and supplies to the forward area was a responsibility of the Allied Naval Forces until relieved by the Services of Supply. The date for the transfer of this responsibility was not set prior to D Day.20 The assault elements of the CYCLONE Task Force were to carry with them ten days’ supply of rations, clothing, unit equipment, fuels, lubricants, medical supplies, and motor maintenance matériel. All weapons except the 4.2-inch mortars were to be supplied with two units of fire; the mortars were to have four units. Provision for resupply was similar to that of previous operations. Ultimately, thirty days of supply of all matériel (except engineer construction equipment) and three units of fire for all weapons were to be built up at Noemfoor. Engineer construction supplies were to be brought forward as necessary.21
There were no critical shortages and no supply problems other than relatively minor difficulties concerned with loading and unloading. The Naval Attack Force did not want any bulk supplies loaded on LSTs of the D Day echelon, but planned that all supplies would be carried packed on vehicles transported by the LSTs. This plan was prompted by Admiral Fechteler’s wish to pull the LSTs away from Noemfoor as quickly as possible, since he felt there might be a strong enemy air reaction to the landing, a reaction which, because of possibly adverse weather conditions, the Fifth Air Force might not be able to counter. After conferences with the task force and ALAMO Force supply sections, however, Admiral Fechteler agreed to bulk-load 200 tons of cargo on each LST of the D Day echelon, provided that a 100-man unloading detail for each ship were made available. The 6th
Infantry Division, at Wakde–Sarmi, was called upon to provide 800 men for the unloading. These men were to return to Wakde–Sarmi on the LSTs they unloaded and were not to be committed to combat operations on Noemfoor. Admiral Fechteler also believed it necessary to make roller conveyors available to unload the bulk cargo at Noemfoor, and ALAMO Force procured adequate lengths of these conveyors from stocks in eastern New Guinea bases.22
In many essentials, the landing plans for Noemfoor were very similar to those used at Biak. Like the latter island, Noemfoor was surrounded by coral reefs which were barely covered by water even at high tide. Therefore, as at Biak, LVTs and DUKWs were to make up the assault landing waves for Noemfoor. Again, LCMs and LCTs were to be run up on the reef and over it if possible, probable damage to these craft having to be accepted because of the importance of their cargoes of tanks, trucks, bulldozers, and engineer equipment. LCIs and LSTs were to beach at the outer edge of the reef, their troops and vehicles to go ashore over the reef. DUKWs and LVTs were to aid in the unloading of the LSTs.
In one major essential the Noemfoor landing plan differed radically from that employed at Biak. At the latter island the HURRICANE Task Force had used a beach which, while within easy marching distance of the principal objectives and the main concentration of enemy troops, was relatively undefended. But at Noemfoor, the landing was to be made in the face of the enemy’s strongest defenses, known to be located in the Kamiri Drome area. YELLOW Beach, as the landing area was designated, extended approximately 800 yards along the western end of the airfield, which was situated almost at the high water mark. The reef presented fewer hazards there than elsewhere, since it was somewhat narrower than at most other points along the island’s coast. The relative narrowness of the reef at Kamiri would also permit LCIs, LCTs, LCMs, and LSTs to approach to within 450 yards of the beach, which, at the airfield, was believed to be firm. Moreover, landing at YELLOW Beach had the advantage of placing the assault troops immediately on their objective, permitting a rapid seizure of Kamiri Drome before the Japanese could recover from the shock of the naval and air bombardments. Enemy forces on the island would be split, and those stationed at Namber and Kornasoren Dromes would be isolated.
To insure that the assault troops would get ashore with minimum casualties, the landing plan demanded the heaviest naval bombardment yet delivered in the Southwest Pacific area. Two and one-half times the amount of ammunition normally thought necessary to neutralize the landing area was to be expended against YELLOW Beach and its environs. Any miscarriage of plans, it was realized, would allow the Japanese to recover from this bombardment and inflict serious losses on the landing waves. Admiral Fechteler recognized that the landing plan called for ideal conditions of wind and sea, and he planned to postpone the assault if unfavorable weather conditions prevailed on the morning of 2 July. The landing itself was scheduled for 0800, fifty-seven minutes after sunrise. This hour was later than customary for landings along the New Guinea coast, but it had the double
purpose of allowing more naval gunfire to be laid on the beach with accuracy and insuring proper beach identification. The plan of naval support fire, except for its volume, was similar to that of previous operations, as was the composition and timing of assault waves.23
The Landing
The CYCLONE Task Force was formally organized on 21 June, when General Patrick, in order to coordinate his final planning with that of other organizations concerned with the Noemfoor operation, set up a temporary command post near ALAMO Force’s rear headquarters at Finschhafen. General Patrick’s plan of operations was approved by General Krueger on 22 June and published as CYCLONE Task Force Field Order No. 1 the next day. General Patrick and his staff then returned to the Wakde–Sarmi area to complete final preparations for the Noemfoor operation. A rehearsal for amphibian vehicles was held on 28 June, and final loading of the assault ships was finished soon thereafter.24
Approach and Bombardment
The LCT-LCM Unit, escorted by three PCs, left Toem on 29 June and sailed for Biak, where the landing craft had a twenty-four hour layover. Two LCIs, carrying most of the troops who were to move to Noemfoor aboard LCMs, accompanied the unit as far as Biak. The Main Body of the Attack Force departed Toem at 1800 on 30 June, arriving off southeastern Biak about 1740 hours on 1 July. The LCM troops then left the LCIs for their own craft, and 8 of the 13 LCTs were taken in tow by LSTs of the Main Body, which immediately sailed for Noemfoor. The 8 LCTs were towed to Noemfoor in order that they might be available to unload equipment from the LSTs during the first phases of the assault. The rest of the LCT-LCM Unit, consisting now of 5 LCTs, 40 LCMs, and 3 PCs, proceeded behind the Main Body at best speed, arriving off Noemfoor not much more than half an hour after the Main Body.25
The Main Body began deploying off YELLOW Beach about 0500 on D Day, 2 July. Various sections of the Main Body, including accompanying destroyers, were released from formation during the predawn hours to take up assigned control or fire support stations. About 0700 the assault ships stopped while the LSTs cast off their LCT tows. As a result of this halt, the Main Body arrived at the transport area, about 3,000 yards offshore, approximately ten minutes late, and the LSTs were delayed an additional five minutes in completing their deployment. However, the LSTs then steamed on to their stations with bow doors open and ramps half lowered, thus speeding launching of LVTs and DUKWs of the assault waves and making up all the lost time.
Four control craft had already arrived on station, two at the line of departure about 1,000 yards beyond the outer edge of the reef and two more at the reef’s edge. Heavy haze, caused by the smoke and dust of the prelanding air and naval bombardments, obscured YELLOW Beach and the control
boats, which could not be seen from a distance of more than 500 yards. The control boats therefore turned on white flood lights, enabling the assault craft to obtain correct bearings for the run to the beach.
The naval bombardment was carried out by the Covering Force, augmented by destroyers of the Main Body. The Covering Force had formed at the Admiralty Islands, whence it had sailed westward to reach a point about thirty miles southeast of Biak at 1900 hours on 1 July. Proceeding toward Noemfoor about ten miles ahead of the Main Body, the Covering Force made a few radar contacts with Japanese aircraft, but its trip was otherwise without incident. The force moved into assigned fire support areas in time to begin its bombardment on schedule at H minus 80 minutes.
One heavy cruiser (HMAS Australia) and four destroyers bombarded YELLOW Beach and its flanks from H minus 80 to H minus 30 minutes, while two American light cruisers and six destroyers threw the weight of their fires on target areas east of the beach, including Kornasoren Drome. Four additional destroyers hit YELLOW Beach and its right flank and four more fired on the left flank of the beach. Among the most important targets were low coral ridges immediately behind Kamiri Drome. Three destroyers stood by for call fire, and two others set course south along the west coast of Noemfoor to place harassing fire on Namber Drome and its environs.26
The Allied Air Forces had been bombing Noemfoor and enemy air bases on the Vogelkop Peninsula for some days in preparation for the landing of the CYCLONE Task Force. On 1 July, 84 B-24’s, 36 A-20’s, and 12 B-25’s were over the island, dropping 195 tons of bombs, while 22 P-38’s glide-bombed Kamiri and Kornasoren Drome installations with 11 tons of 1,000-pound bombs. On D Day 33 B-24’s, 6 B-25’s, and 15 A-20’s expended 108 tons of bombs and 32,000 rounds of ammunition in strafing runways. In addition, two squadrons of fighters were on air alert over the landing area to protect the assault shipping and drive off any Japanese planes which might appear. Immediately before the landing, air bombardment, like much of the naval gunfire, was directed against the low coral ridges and hills behind Kamiri Drome. It was believed that the most determined enemy opposition would come from positions in these ridges, and to neutralize these possible defenses the 33 B-24’s, at approximately H minus 15 minutes, dropped 500-pound bombs along the ridge lines.
As this bombing ended, the first wave of troop-carrying LVTs began approaching the outer edge of the reef. About the same time, rocket-equipped LCIs launched almost 800 rockets into the immediate beachhead area, adding final touches to the bombardment. Automatic weapons aboard four LVT(A)’s of the Support Battery, 2nd Engineer Special Brigade, accompanying the leading assault wave, kept up a steady fire on YELLOW Beach as the troop-carrying LVTs, also manned by the Support Battery, clambered over the reef toward the beach.27
The Assault
There was no opposition to the landing. The LVTs of the first wave, ashore on
schedule at 0800, continued from the beach line across Kamiri Drome to the base of a coral ledge overlooking the field. There the assault riflemen of the 1st and 2nd Battalions, 158th Infantry, dismounted, the 1st Battalion on the west and the 2nd on the east. Subsequent waves of LVTs and DUKWs (the latter manned by the 464th Amphibian Truck Company) brought the rest of the two battalions ashore quickly. The troops assembled at the coral ridge and rapidly began advancing west, east, and south to extend the beachhead.
First resistance was encountered by the 2nd Battalion about 500 yards from the eastern end of Kamiri Drome, when without warning about forty Japanese ran out of a cave in the ledge and began milling around in a rather aimless fashion. Showing no tendency either to surrender or to flee, the Japanese were killed by rifle fire or the automatic weapons of the Support Battery’s LVT(A)’s. Beyond the area of this encounter, a number of lightly manned caves and prepared defenses were found. There was little organized resistance from these positions, but the 2nd Battalion, advancing slowly, halted its attack to mop up each cave, dugout, and foxhole. These methodical operations were supported by the Support Battery LVT(A)’s, later joined by the 1st Platoon, 603rd Tank Company. On the west flank the 1st Battalion, 158th Infantry, encountered less opposition and had no difficulty securing a low hill off the southwest end of Kamiri Drome. While part of the battalion cleared that hill, the rest of the unit pushed south from the airfield about 1,000 yards to the north bank of the Kamiri River.
Meanwhile the 3rd Battalion, 158th Infantry, had come ashore. It marched rapidly westward after the 2nd and joined the latter unit in the mopping-up operations along the low, jungled, coral ridges at the east end of the airstrip. Upon the arrival of the 3rd Battalion, most of the 2nd turned south from the field toward the Kamiri River, meeting no opposition on its way over more ridges and through dense jungle to the north bank of the tidal stream. By 1600 hours the three infantry battalions had secured a rectangular area about 3,000 yards wide and some 800 deep, extending south to the banks of the Kamiri. It had been expected that the 158th Infantry would have moved some 1,800 yards farther east toward Kornasoren Drome on D Day, but progress had been delayed because the advance elements of the 2nd and 3rd Battalions had lost momentum as they stopped to mop up minor points of enemy resistance instead of leaving such defenses to follow-up forces.28
While the assault battalions had been seizing the beachhead, the rest of the CYCLONE Task Force had been pouring ashore. The first seven landing waves, consisting almost entirely of LVTs and DUKWs, moved to the beach in good order, slowed only by the difficulty experienced by some DUKWs in negotiating the reef. The eighth wave, comprising four LCTs with the 1st Platoon, 603rd Tank Company, aboard, followed the last DUKW wave to the outer edge of the reef where the tanks unloaded and lumbered ashore, reaching the beach by 0850. LCIs carrying the 3rd Battalion and other nonassault elements of the 158th Infantry, closed on the reef about 0825. Some of the men debarked into water about waist-deep, but most of them waded ashore in less than a foot of water. Small rubber boats, inflated aboard the LCIs, were used by some
troops to drag ashore heavy equipment such as ammunition and mortars. The debarkation of all elements of the 158th Infantry was completed by noon.
LCMs, carrying engineer equipment, trucks, and bulldozers, began coming up to the reef about 0815 and three hit the reef at full throttle to see if it could be jumped. This proving impossible, the LCMs unloaded their cargo at the reef edge. It had been planned to keep the LCMs clear of the approach lane until the LCTs had finished unloading. Somehow, orders to this effect had either gone astray or been misunderstood, and within fifteen minutes after the first LCMs hit the reef, the approach lane began to be clogged with these craft, jockeying for position with LCTs and LCIs. Crowded out, the LCMs moved west of the lane to an unmarked section of the reef. This proved a happy circumstance, for at the new area wheeled vehicles found a smoother reef crossing.
Nevertheless, during the first stages of the landing, practically all wheeled, nonamphibian vehicles, whether waterproofed or not, had to be towed over the reef by DUKWs, LVTs, tanks, or bulldozers. After H plus 2 hours, ebbing tide permitted most of the vehicles to move ashore under their own power without drowning out their engines. During the course of the landing, 6 trucks, 2 jeeps, and 5 small trailers were lost in pits in the reef. All but one truck and one trailer were later salvaged.
As soon as the assault waves were on the beach, naval demolition personnel began blasting operations along the outer edge of the reef so that LSTs could move closer inshore, and the Shore Battalion, 593rd Engineer Boat and Shore Regiment, began constructing temporary causeways over the reef. Some vehicles aboard the LSTs were set directly on the reef, but most of them were transshipped to the reef edge from the landing ships by LCMs. Bulk cargo aboard the LSTs was manhandled into LVTs and DUKWs which returned from the beach for the express purpose of aiding unloading. The 27th Engineer Battalion was responsible for unloading three LSTs and getting the cargo thereof ashore, and troops of the 6th Division handled most of the bulk cargo aboard the others. By a combination of all the foregoing methods, four LSTs were completely unloaded on D Day and most of the cargo of the others had been sent ashore. The latter moved out to sea at dusk and returned on D plus 1 to complete unloading.
The 105-mm. howitzers of the 147th Field Artillery Battalion were brought ashore from LSTs by DUKWs and were dropped into position on land by a few DUKWs which were especially equipped with A-frame cranes. The battalion was ashore and in position by 1100, ready to provide hasty support to the 158th Infantry; registration for more accurate fire was completed by 1145. Antiaircraft artillery units began coming ashore about 0810, and all were landed and set up to defend the beachhead by 1600.
The first sign of enemy countermeasures came about 0905, when Japanese mortar or 70-mm. artillery shells began falling in the beachhead area and on the coral reef beyond. Few casualties were suffered from this fire which, far from slowing the landing, probably did much to speed unloading. The enemy’s shells set one DUKW afire and destroyed a truckload of ammunition. The fire continued intermittently for about two hours, despite efforts of naval support vessels
DUKW burning on the beach at Noemfoor
and planes of the air umbrella to locate and destroy the Japanese weapons.29
The Shore Battalion of the 593rd Engineer Boat and Shore Regiment, the Naval Beach Party, and other elements of the Shore Party quickly organized the landing beach. Matting was laid on the beach so that wheeled vehicles could find traction, dispersal areas were located and cleared on the south side of Kamiri Drome, supplies were sent off the beach to these dumps, shell holes were filled in, and Japanese supplies piled up out of the way. Some of the 27th Engineers participated in these Shore Party activities while other elements of the battalion moved inland with the infantrymen to supply flame thrower teams for mopping up operations. The rest of the battalion, working under the direction of Headquarters, No. 62 Works Wing, began repairs on Kamiri Drome, pending the arrival of the rest of the Australian engineer unit and American engineer aviation battalions.30
Despite extensive enemy defensive preparations in the Kamiri Drome area, the CYCLONE Task Force’s losses on D Day were only 3 men killed (1 accidentally), 19 wounded, and 2 injured. This is a tribute to the heavy air and naval bombardment, which succeeded in driving most of the Japanese away from the beach or keeping those that remained pinned down as the assault waves moved ashore. The Japanese suffered much more heavily. About 115 were killed or found dead and 3 were captured.31
The CYCLONE Task Force had expected to meet about 3,000 of the enemy, most of whom were considered combat troops. By evening on D Day, the task force had put ashore over 7,000 men, nearly all of them, including the 3,300 of the 158th Infantry, classed as combat troops.32 So far, there had been no evidence of organized resistance and few Japanese had been located or observed. In the evening, however, General Patrick concluded that 3,500 to 4,500 Japanese combat troops were on Noemfoor and that the enemy garrison totaled about 5,000 men. This new estimate was apparently based upon the evidence of a Japanese prisoner who had heard that 3,000 Japanese infantry reinforcements had arrived on Noemfoor about 25 June. The prisoner admittedly had not seen any of these troops and his information was either uncorroborated or flatly contradicted by two other prisoners and a recovered Javanese slave laborer.33
Since little organized resistance had been encountered, plans for 3 July were to institute patrolling designed to locate the main body of Colonel Shimizu’s Noemfoor Detachment. The 2nd and 3rd Battalions, 158th Infantry, were to continue eastward toward Kornasoren Drome. These units began moving at 0900 on the 3rd. The 3rd Battalion was temporarily delayed at a mine field which the 27th Engineers cleared, but advanced over 1,800 yards by late afternoon. A number of well-prepared defensive positions, situated both to defend the beach and prevent lateral movement between Kamiri and Kornasoren Dromes, were found, but none was manned by the Japanese. On the west flank the 1st Battalion patrolled south of the Kamiri River but located only a few enemy stragglers. By the day’s end the 158th Infantry had lost only 2 men wounded and 1 injured, while 14 Japanese had been killed.34
While the 158th Infantry had been expanding the beachhead, American paratroopers began to drop on Kamiri Drome
to reinforce the CYCLONE Task Force. This reinforcing operation was to have tragic results for the airborne soldiers.
Ordering the 503rd Parachute Infantry Forward
One of the missions assigned to the CYCLONE Task Force had been to secure a suitable dropping ground for paratroopers. From study of maps and aerial photographs before the landing, General Patrick had decided that he would use Kamiri Drome if that field were found reasonably free of shell holes and other obstacles. Examination of the ground on the morning of the landing confirmed General Patrick’s decision and at 1028, three minutes after he assumed command ashore, he radioed to General Krueger that Kamiri Drome was a satisfactory dropping ground.35
At 1115 General Patrick dispatched another radio to General Krueger recommending that the 503rd Parachute Infantry be sent forward to Noemfoor and dropped on Kamiri Drome. General Patrick gave the following reasons for requesting reinforcements: “. . . to guard against unknown situation as to enemy strength and to speed up operation on island.”36 This radio was decoded at ALAMO Force headquarters at 1410.37 Twenty minutes later the 503rd Parachute Infantry received verbal instructions from ALAMO Force’s advance headquarters at Hollandia to the effect that one battalion of the regiment was to drop on Noemfoor on 3 July and that the other two battalions would drop on the succeeding two days.38
This drop plan meant that it would be midmorning of 5 July before all the 2,000 odd men of the 503rd Parachute Infantry could reach Noemfoor. The other ALAMO Force Reserve for Noemfoor—the 34th Infantry, 24th Division, at Biak—was assembled on the beach at Biak Island on the afternoon of 30 June and could load for overwater movement to Noemfoor on less than twenty-four hours’ notice. LCIs taking the 34th Infantry to Noemfoor could cover the 75–80 nautical miles from the beach at Mokmer Drome, Biak, to the reef off Kamiri Drome, Noemfoor, in not more than nine hours.39 Figured from 1115 on 2 July, when General Patrick first asked for reinforcements, the 2,700 men of the 34th Infantry could have reached Noemfoor not later than 2000 hours on the 3rd. By that time, in accordance with the air reinforcement plan, only one battalion of less than 750 men of the 503rd Parachute Infantry would be on the island.
General Patrick’s request for paratroop reinforcements and ALAMO Force’s speedy acquiescence and subsequent orders to the 503rd Parachute Infantry probably surprised no one. Apparently, the parachute
unit was to have been sent to Noemfoor unless General Patrick was strongly opposed to its movement; only the 1115 radio from the CYCLONE Task Force was awaited before the machinery for the movement started. In fact, the 503rd Parachute Infantry had known at least as early as 1615 on 1 July—the day before D Day—that one battalion of the regiment was to drop on Noemfoor on 3 July. Orders to that effect had been sent to the regimental command post at Hollandia during the afternoon of 1 July by the ALAMO Force G-3 section.40 The only alteration in plans made necessary by the receipt of jump orders at 1430 on 2 July was to change the dropping ground. The 503rd had expected to jump at Kornasoren Drome, but now it had to restudy available information in preparation for the drop on Kamiri Drome.41
The 503rd Parachute Infantry Drops at Noemfoor
About 0515 on 3 July regimental headquarters and the 1st Battalion, 503rd Parachute Infantry, began loading at Cyclops Drome, Hollandia, on thirty-eight C-47’s of the 54th Troop Carrier Wing, Fifth Air Force. At the same time, three B-17’s, from which supplies and ammunition were to be dropped on Kamiri Drome, were loaded. The first C-47 took off from Cyclops Drome at 0630, and by 0747 all forty-one planes were in the air. The commanding officer of the parachute regiment, Col. George M. Jones, and most of his staff were in the lead C-47. The planes were to fly over Kamiri Drome in flights of two each, the first plane at a height of 400 feet and the second echeloned slightly to the right rear at 450 feet. Subsequent flights were to follow at a distance of 300 yards.42
About 0600 on the 3rd, almost twenty hours after he had been advised on the point by a paratroop officer, General Patrick radioed to ALAMO Force that it would be wise if the C-47’s flew over Kamiri Drome in single file.43 He made this recommendation because he feared that the falling paratroopers might suffer casualties if they landed on obstacles along the sides of the narrow airfield, which comprised a 250 by 5,500-foot cleared area and a 100-foot-wide runway. The radio was received at Headquarters, ALAMO Force, about 0740 but apparently was not delivered to the G-3 Section until 0915. Sometime between 0740 and 0915 the radio was passed to Headquarters, Fifth Air Force, by the ALAMO Force message center.44 By then, the troop-carrying
Paratroopers landing on Noemfoor. Note equipment along the airstrip
planes were airborne and well on their way to Noemfoor.
No attempt seems to have been made to establish radio contact with the 54th Troop Carrier Wing’s C-47’s to effect the desired change in formation. Whether such an eleventh-hour alteration could have been made is a difficult question. Last-minute attempts to change plans might have created confusion which could have delayed or postponed the parachute drop. Moreover, the radio traffic necessary to effect the change might have brought every Japanese plane within range of Noemfoor over that island. In any case, no change in formation was made, and the thirty-eight C-47’s flew into sight of Kamiri Drome about 1000. Ten minutes later, the ‘troopers from the leading C-47 were on the ground, followed closely by the men in the neighboring plane.
Contrary to plans, the first two C-47’s flew over the strip at a height of about 175 feet, and the next eight planes all flew below 400 feet. Dropping from this low altitude caused the paratroopers in the first ten C-47’s to suffer many casualties; more casualties resulted because the planes flew over the strip two abreast. The broad formation caused many ‘troopers to land off the southern edge of the 100-foot-wide runway in an area where Allied vehicles, bulldozers, supply dumps, and wrecked Japanese aircraft were located. Additional hazards beyond the cleared area were jagged tree stumps, trees partially destroyed by preassault air and naval bombardments, and a number of antiaircraft gun emplacements. Altogether, there were 72 casualties among the 739 men who dropped on 3 July. Included in this number—a rate of almost 10 percent—were 31 severe fracture cases, most of whom would never again be able to make a parachute jump.45
The first reports of the 3 July jump received by ALAMO Force stated that there had been only 1 percent casualties during the drop.46 Later information, received at Hollandia about 0200 on the 4th, raised that rate to 6.7 percent.47 But even before these reports began to filter back to Hollandia, the 503rd Parachute Infantry was informed that another battalion was to drop on 4 July.48 In preparation for this second jump, General Krueger instructed General Patrick to make sure that the edges of Kamiri Drome were clear of vehicles and the 54th Troop Carrier Wing was ordered to fly its C-47’s in single file over the airfield.49
At 0955 on the 4th the 3rd Battalion, 503rd Parachute Infantry, and the rest of regimental headquarters began dropping on Kamiri Drome; by 1025 the 685 men of this echelon were on the ground. This time all the C-47’s flew at a height of at least 400 feet in single file formation, and, although the flight pattern of five to seven planes was not entirely satisfactory, nearly all the troopers landed on the airstrip.50
Even with the new precautions there were 56 jump casualties, a rate of over 8 percent. Most of the injuries on the second drop were attributed to the hard coral surface of Kamiri Drome, on which considerable grading, rolling, and packing had been accomplished since the morning of 3 July. So far, 1,424 officers and men of the 503rd Parachute Infantry had dropped at Noemfoor. There had been 128 jump casualties, a final rate of 8.98 percent, among them 59 serious fracture cases. There had been no casualties from enemy action. The parachute regiment had lost the services of one battalion commander, three company commanders, the regimental communications officer, and a number of key noncommissioned officers.51
Colonel Jones, the regimental commander, considered that injuries had been excessive on both the 3rd and 4th of July, and he therefore requested General Patrick to arrange for water shipment of the remaining battalion.52 The task force commander agreed that no more drops should be attempted, but he suggested to ALAMO Force that the remainder of the regiment be brought forward by air as soon as Kamiri Drome was sufficiently repaired to receive C-47’s. With these recommendations, General Krueger agreed.53 However, torrential rains and a shortage of heavy equipment at Noemfoor combined to keep the airfield inoperational longer than had been expected. Finally the 2nd Battalion, 503rd Parachute Infantry, was flown from Hollandia to Mokmer Drome on Biak. Disembarking from C-47’s at Mokmer, the troopers moved aboard LCIs for the trip to Noemfoor, which they reached on 11 July.54
The Occupation of Noemfoor Island
Possibly the only valuable result of the parachute drop was that mopping-up operations on Noemfoor could begin sooner than might otherwise have been possible. The 1st Battalion, 503rd Parachute Infantry, upon its arrival on 3 July, assumed responsibility for about 2,000 yards in the center of the defenses around Kamiri Drome, thus permitting the 2nd and 3rd Battalions, 158th Infantry, to concentrate at the eastern end of the field and extend the perimeter. The 3rd Battalion, 503rd Parachute Infantry, when it landed on 4 July, relieved elements of the 3rd Battalion, 158th Infantry, as the latter unit pushed east to Kornasoren Drome.55
Hill 201
The 3rd Battalion, 158th Infantry, moving eastward along the coastal road, encountered no opposition on 4 July.56 The battalion found a number of well-prepared, but abandoned, enemy defensive positions along both sides of the road, and the entire area around Kornasoren Drome and the village of Kornasoren was discovered to be mined in a haphazard fashion, principally with 200-pound aerial bombs, many of which were incompletely buried. The infantry unit had little difficulty picking its way through the mine fields, which the engineers rapidly cleared. At evening on the 4th, the battalion dug in at the eastern end of Kornasoren Drome. Meanwhile the 1st Battalion (less Company A) had crossed the Kamiri River by LVT and LCM and occupied Kamiri village without opposition. Out of Kamiri, the battalion followed a road leading southeast to a large Japanese garden area 1,700 yards distant. The battalion’s leading elements began approaching a terrain feature designated Hill 201, in the western section of the garden area, about 1330. So far, only scattered rifle fire had opposed the march from Kamiri village.
The Japanese garden area was about 600 yards long, east and west, and 350 yards across, north to south. The ground was devoid of large trees except for a few atop Hill 201, but thick, secondary jungle growth covered the eastern and southern slopes of the hill, while the rest of the garden area was overgrown with partially cultivated papaya, taro, and cassava, all averaging about eight feet in height. The trail from Kamiri village passed over the southern slope of Hill 201, and 300 yards away, near the eastern edge of the gardens, joined the main road from Kamiri Drome to Namber Drome, located about six miles to the south. Another trail ran along the eastern side of the hill, branching to the north and northwest at the northeastern corner of the low terrain feature.
As the 1st Battalion approached the western side of Hill 201, a small group of Japanese, heavily armed and carrying packs, was seen hurrying south along the main road below the hill. Fearing that Hill 201 might be occupied in some force, the battalion commander halted the advance and prepared a double envelopment, to be supported by heavy machine guns from positions west of the hill and south of the Kamiri-garden trail. First, Company C seized a knoll called Hill 180, lying 300 yards north of the gardens. Then Company B surprised a Japanese platoon in a mangrove swamp at the southwest corner of the garden area and pushed the Japanese southeast over Mission Hill, a low terrain feature lying east of the main road and southeast of Hill 201. With the ground on both flanks of Hill 201 secured, the battalion, at 1515, began concentrating around the hilltop for the night, setting up a tight defensive perimeter.
Company C protected the northern and eastern slopes; Company B was on the south and southeast; and Company D, together with battalion headquarters and headquarters company, dug in along the western slope. One heavy machine gun was sighted to cover a trail leading up the southern slope of the hill from the mangrove swamp, and a section of heavies was emplaced to fire on the main road and Hill 170, 600 yards east
of Hill 201. Company D’s 81-mm. mortars registered on the edge of the mangrove swamp and on the trail to Kamiri at the point at which that trail entered the garden area. Three guns of an attached platoon of 4.2-inch mortars (641st Tank Destroyer Battalion) registered on the eastern slopes of Hill 180, on a trail leading north past that hill, and on a target area on the main road to Kamiri Drome at the point where the road entered the north edge of the garden. A battery of the 147th Field Artillery Battalion’s 105-mm. howitzers, emplaced near Kamiri Drome, registered along the main road through the gardens, on the west side of Hill 170, and on Mission Hill, where the main road left the garden area. By 1800, with the accompanying hazard of occasional rifle fire from hidden Japanese, all defensive preparations were completed. Moonset was at 0500 on the 5th, and the battalion commander called his men’s attention to the danger of a Japanese attack between that hour and first light. The warning was well taken.
About 0430 a few Japanese were heard moving around at the northeast edge of the perimeter. Near the same time, more enemy approached Hill 201 from Mission Hill and through the mangrove swamp south of the perimeter. At 0520, without the warning of preparatory fire, Japanese infantry began pressing forward all along the southern and southeastern sides of the perimeter. As the attack started, some enemy light mortars began firing, but all the shells from these weapons landed about 200 yards west of Hill 201. As soon as the enemy attack began, the 1st Battalion requested that the prearranged artillery and mortar concentrations be fired. The battalion’s 81-mm. mortars immediately began throwing shells into the mangrove swamp and along the trail approaching Hill 201 from the south. Cries of wounded or dying Japanese attested to the effectiveness of these fires, and the enemy was driven off the trail into the second-growth jungle and overgrown gardens. Now, two Japanese light machine guns opened up, one from the western slope of Hill 170 and another from Mission Hill. Both these weapons were firing from positions on which the 147th Field Artillery Battalion had previously registered. They were quickly put out of action.
Some Japanese managed to get through the mortar and artillery barrages and continued up the sides of the trail from the south through fire from Company D’s machine guns. The enemy found cover behind a low, 150-yard-long log fence which led from the southeast toward the center of the 1st Battalion’s defenses. While the fence afforded some protection, the attacking infantrymen were silhouetted as they tried to clamber over the top. The main body of the attacking force therefore kept down behind the fence, trying to crawl along it to the top of the hill. But the fence did not extend beyond the 1st Battalion’s outer defenses and the attackers soon found themselves debouching into steady machine gun and rifle fire from the defenders atop the hill. The attack soon degenerated into a series of small suicide charges by groups of three to six Japanese, all of whom were cut down as they tried to climb the fence or move beyond its end. By 0630 the last enemy efforts had ceased and all firing had stopped.
The 1st Battalion now sent patrols out over the battlefield. From prisoners it was determined that the attacking force had consisted of 350 to 400 men—the 10th and 12th Companies, 219th Infantry, reinforced by approximately 150 armed Formosan laborers. During the morning over 200 dead
Japanese were counted around the 1st Battalion’s perimeter, and the number of enemy dead found or enemy wounded captured on subsequent days along trails leading south from the hill indicated that virtually the entire original attacking force had been annihilated.
Mopping Up
At Hill 201, the Noemfoor Detachment made its only significant offensive effort, and after that affair operations on Noemfoor evolved into a series of patrol actions as the Allied forces extended their control over the island and rapidly expanded the airdrome facilities.57 On 5 July, the 1st Battalion, 158th Infantry, mopped up the garden area, while the 3rd Battalion patrolled in northeastern Noemfoor, finding no Japanese. The 2nd Battalion, relieved at Kamiri Drome by the 3rd Battalion, 503rd Parachute Infantry, started preparing for an amphibious landing at Namber Drome, on the southwest coast.
This operation got under way about 0900 on 6 July. LCMs of Company A, 543rd Engineer Boat and Shore Regiment, moved the 2nd Battalion, 158th Infantry, from Kamiri Drome to the north side of Roemboi Bay, at the lower end of Namber Drome. The Support Battery, 2nd Engineer Special Brigade, covered the landing with antiaircraft boats, and three destroyers and a rocket-equipped LCI were also on hand. Following a short naval bombardment, which was accompanied by bombing and strafing by six B-25’s, the 2nd Battalion poured ashore without opposition. Namber Drome was secured by 1240, not a shot having been fired by the 2nd Battalion and not a single casualty having been suffered. Fifteen minutes after Namber Drome was taken, a liaison plane of the 147th Field Artillery Battalion landed on the strip.
From 7 through 10 July vigorous patrolling by all elements of the CYCLONE Task Force was continued, but only small Japanese parties were encountered. The conclusion was therefore reached that no organized large enemy force remained on Noemfoor and that future operations would consist of hunting down small enemy groups. For the latter purpose, it was decided to divide the island into two parts, making the 503rd Parachute Infantry responsible for the southern section and the 158th Infantry for the northern.
Regrouping of units according to this plan started on 11 July, when the 2nd Battalion, 503rd Parachute Infantry, reached Namber Drome from Biak, via LCIs. During the next day or so the 158th Infantry reconcentrated in northern Noemfoor. The regiment’s activity to 31 August was confined to extensive patrolling, which produced contacts only with small groups of enemy stragglers. To the end of August the regiment killed 611 Japanese, captured 179, and liberated 209 Javanese slave laborers. During its operations on Noemfoor, the 158th Infantry lost 6 men killed and 41 wounded.
To the 503rd Parachute Infantry fell the task of mopping up the remnants of the Noemfoor Detachment, which, after the abortive attack at Hill 201, concentrated in southern Noemfoor. The largest organized group of Japanese (400–500 strong)
gathered under Colonel Shimizu’s command at Hill 670, in the west-central part of the island about three miles northeast of Namber Drome. The 1st Battalion, 503rd Parachute Infantry, made contact with this enemy group on 13 July. Finally reaching the top of Hill 670 on the morning of the 16th, the ‘troopers found that the Japanese had evacuated the hill the previous night. Contact with the main body of Colonel Shimizu’s force was lost until 23 July, when patrols of the 2nd Battalion, 503rd Parachute Infantry, located the group about four miles northwest of Inasi, a native village on the shore of the lagoon which cuts into Noemfoor’s eastern coast.58 Contact was again lost on the 25th and not regained until 10 August, when the main Japanese force was found near Hill 380, two and a half miles south-southwest of Inasi. From the 10th through the 15th, the 3rd Battalion, 503rd Parachute Infantry, supported by the 147th Field Artillery Battalion and a few B-25’s of the 309th Bomb Wing, converged on Hill 380. At the same time, elements of the 1st Battalion moved toward the hill from Menoekwari South to prevent enemy escape. Though boxed in by the bulk of five companies of the 503rd Parachute Infantry, Colonel Shimizu, during the night of 15–16 August, slipped through the cordon around Hill 380 and withdrew his remaining force, now not much over 200 men strong, south-southwest toward Pakriki, on the south-central coast.
Colonel Shimizu’s party was not again located until 17 August, when two brisk engagements resulted in the capture or destruction of the last machine guns left to the Noemfoor Detachment and broke the last organized resistance. At least 20 Japanese were killed, bringing to 342 the total of Japanese killed in the series of actions southwest from Inasi to Pakriki during the period 10 through 17 August. In the same week, 43 Japanese were captured. Colonel Shimizu was not among those killed or captured, and was at large at the end of the month.
On 23 August the 503rd Parachute Infantry, less three companies, began concentrating at a new camp near Kamiri Drome. The three companies, which had been left in the southern part of the island to continue patrolling, were finally relieved on the 27th by elements of the 1st Battalion, 158th Infantry. According to one report, the last paratroopers left the southern section of Noemfoor with some regrets:–
As the troops left the area in which they had chased SHIMIZU so relentlessly, their disappointment was not so much at their failure to apprehend SHIMIZU the man, nor even to capture the regimental colors of the 219th Infantry. It was rather that they missed their chance of retrieving the colonel’s 300 year old saber which prisoners said he still carried when last seen near PAKRIKI. When the operation officially closed on 31 August this same saber was still inducing the most vigorous patrolling by the 1st Battalion, 158th Infantry.59
By 31 August, when General Krueger declared the Noemfoor operation over, the CYCLONE Task Force had lost 63 men killed,
343 wounded, and 3 missing. Approximately 1,730 Japanese had been killed and 186 were captured. Most of the Allied casualties were suffered by the 503rd Parachute Infantry during its clashes with the Noemfoor Detachment in southern Noemfoor after 11 July, and the regiment can claim credit for killing about 1,000 Japanese. In addition to the Japanese losses, 1 Korean, 1 Chinese, and 552 Formosan prisoners of war were taken. Finally, 403 Javanese slave laborers were recovered on the island.60
Base Development on Noemfoor
Civil Affairs and Atrocities
For the purpose of supervising the civil population of Noemfoor, a Netherlands Indies Civil Administration (NICA) Detachment was attached to the CYCLONE Task Force, just as NICA parties had been attached to previous Allied task forces operations in Dutch New Guinea.61 Initially, the Noemfoor NICA Detachment consisted of 4 officers and 35 enlisted men, but it was later augmented by 10 local policemen recovered on the island. The detachment assisted in obtaining intelligence information, recruited and supervised native labor, and administered the native population. Only 400 of the total native population of some 5,000 were ever used as laborers, the rest being either women or children or men too old or too ill to work.
The Japanese had never brought the Melanesians of Noemfoor entirely under their control, for the natives had either offered a passive resistance or had faded into the interior to live off the land. A few were impressed into service by the Japanese, while others who were captured but still refused to co-operate were executed. The natives greeted the Allied landings with great enthusiasm and came out of hideaways in the hills carrying Dutch flags which they had concealed from the Japanese. Under the direction of NICA, the natives were gradually resettled in their old villages, where they were protected by CYCLONE Task Force outposts. Late in July the village chiefs gathered in formal council and officially declared war on the Japanese. Thereafter, native cooperation increased. Up to 31 August the natives had captured and brought to Allied outposts more than fifty Japanese and had killed an equal number.
One tale of horror concerns the Javanese on Noemfoor. According to information gathered by the NICA Detachment, over 3,000 Indonesians were shipped to Noemfoor in late 1943, mostly from Soerabaja and other large cities on Java. The shipment included many women, children, and teenaged boys. The Japanese, without regard to age or sex, put the Javanese to work constructing roads and airfields almost entirely by hand. Little or no clothing, shoes, bedding, or shelter was provided, and the Javanese had to supplement their very inadequate allowance of rations by shifting for themselves. Driven by hunger, many attempted to steal Japanese rations but for their pains were beheaded or hung by their hands or feet until dead. Starvation and disease (the Japanese provided no medical
care) took a steadily increasing toll. The dead were periodically collected for mass burial, and survivors alleged that many of the sick were buried alive. It was considered probable that not more than ten or fifteen Javanese were killed accidentally by Allied forces. Yet only 403 of the 3,000-odd brought from Java were found alive on Noemfoor by 31 August. The physical condition of these survivors almost defied description—most of the others had succumbed to Japanese brutality within a period of eight months.
The story of Formosan labor troops brought to Noemfoor, ostensibly as part of the Japanese armed forces, is also tragic. Originally the Formosans had numbered about 900 men. They had been worked for months on airfield and road construction, on half the ration of rice issued to the regular Japanese troops. When they collapsed from exhaustion or hunger, or became victims of tropical diseases, they were herded into what the Japanese euphemistically called a labor convalescent camp, actually a sort of prison stockade into which the Formosans were placed to die. There, their rations were again cut in half, and the shelter and blankets provided covered but a fraction of the inmates. Medical care was given only to the worse cases, and then was inadequate. Upon the arrival of Allied forces on Noemfoor, most of the remaining Formosans were rounded up, armed, and forced to fight. But these attempts at compulsion were fruitless. Over 550 Formosans voluntarily surrendered to the Allies, more than half of them suffering from starvation and tropical diseases. Not more than twenty had been killed by Allied action, but about 300 had died before 2 July.
To complete the story of the dire straits to which the Japanese on Noemfoor had been reduced, it is necessary to tell of cannibalism. About 1 August CYCLONE Task Force patrols began to discover Japanese bodies from which portions of flesh had been cut. Initially, this was not considered direct evidence of cannibalism, although increasing numbers of bodies with fleshy portions removed were later found. Finally, some American dead, left outside defensive positions overnight, were discovered in like condition. The prisoners began to report that cannibalism had been generally practiced since 1 August, principally on freshly-killed Formosans. Some prisoners admitted eating human flesh themselves. In at least one instance a two-day-old cadaver had been used for food.
Construction and Unloading
Airfield construction on Noemfoor Island began on D Day, 2 July, when elements of the 27th Engineers used improvised drags and rollers rigged behind six-by-six trucks to start grading Kamiri Drome. This work continued until the afternoon of 4 July when, in the opinion of the task force engineer, Group Captain Dale, the strip was ready to receive a few aircraft. But because of bad weather and existing plans for parachute drops, no planes used the strip until the afternoon of 6 July, when an Australian P-40 squadron landed to remain for future operations.62
It was not until 16 July that sufficient facilities were completed at Kamiri Drome to accommodate an entire fighter group. Ultimately, the 1874th Engineer Aviation Battalion and No. 5 Mobile Works Squadron
of No. 62 Works Wing extended the Kamiri runway to 5,400 feet. Sufficient taxiways and aircraft dispersal facilities for the two groups of fighters were also completed. All except maintenance work on the field was finished on 9 September.63
At first, it had been planned to improve Namber Drome, and some work started there immediately after the field was secured. But Group Captain Dale, finding the site rough and badly graded, recommended that Namber be abandoned in favor of Kornasoren, although he realized that this would increase construction needs at the latter field. General Krueger approved the new plan and preliminary surveys were immediately begun at Kornasoren. Plans were made to construct there two parallel runways 7,000 feet long, with correspondingly large taxiways and dispersal areas.64
On 14 July, General MacArthur directed that by 25 July minimum facilities would have to be prepared at Kornasoren Drome to accommodate fifty P-38’s in order to provide additional air cover for the impending invasion of the Vogelkop Peninsula. Except for continuing work at Kamiri Drome, all remaining engineer units, all available heavy equipment, all service troops who could possibly be spared from ship unloading, all available native labor, and large numbers of combat troops were concentrated at Kornasoren Drome to complete the necessary new construction on schedule. For ten days all hands worked around the clock and by 1200 on 25 July had completed a 6,000 foot-long strip which, with associated dispersal facilities, could accommodate one fighter group. Two days later this strip was extended to 7,000 feet, and the second 7,000-foot runway was completed on 2 September.65
Ultimately, Allied aircraft based on Noemfoor supported not only operations on that island but also the invasions of the Vogelkop Peninsula and Morotai Island. Soon after Kornasoren Drome was completed, B-24’s began flying from the field to conduct the first large-scale bombing attacks on Japanese sources of petroleum products at Balikpapan, Borneo.66
Ship unloading operations at Noemfoor were hampered by the surrounding reefs, but gradually, by the construction of jetties or temporary ramps and by extensive demolitions at the outer edges of the reefs, the difficulties attending supply operations were reduced. Service troops and native labor were both inadequate at Noemfoor, and throughout the operation a daily average of some 600 combat troops had to be used for ship unloading, while others were used on various construction projects.67
The reefs at Noemfoor created another hazard concerning which a classic remark is attributed to Group Captain Dale. The story goes that on D Day Group Captain Dale, after striding up and down Kamiri Drome to ascertain the extent of repairs necessary to that field, returned to his temporary command post, looked out to sea over the reef, and said, “Bad show this—no bloody place for a swim!”68