Chapter 5: Phase I Continued
Progress of Logistical Support1
Shortly after the L-Day landings, Radio Tokyo predicted that the beachhead on Okinawa would be wiped out,2 From L-Day on, the impressive flow of troops and supplies ashore gave little support to this optimistic enemy forecast, however, as the Tenth Army hold on the island rapidly tightened. While the assault units fanned out to gain assigned initial objectives, battalion shore party commanders assumed control of their beach sectors. During L-Day, successively higher command echelons landed, and, by nightfall, divisions had assumed control of shore party operations.
A coral reef extending the length of the beaches was the only real obstacle to early unloading operations. During flood-tide, a steady procession of DUKWs and LVTs shuttled cargo across the reef, and only within this 4-to-5 hour period of high tide could ships’ landing craft make runs directly to a few scattered places on the beaches. Low tide, however, exposed the coral outcropping, and necessitated the establishment of offshore transfer points to maintain the flow of supplies to the beach. Barge cranes required at the transfer points to transship cargo were not available in appreciable numbers until L plus 2.
Increasingly intense Kamikaze raids posed a threat to the transport groups and caused delays in the buildup of supplies ashore. Additionally, the unexpected rapid infantry advances disrupted the unloading schedule. Meanwhile, shore party officers faced such other problems as the lack of suitable beach exits and the scarcity of engineering equipment to prepare them. Another critical matter of note was the shortage of transportation to clear the beaches of supplies. As the volume of cargo being landed increased, the number of trucks available for hauling to inland dumps decreased. According to the operation plans, organic assault division motor transport was to have supported the efforts of the shore parties initially. When frontline troops began to outdistance their support elements, the divisions were forced to withdraw their trucks
from the beaches to resupply forward assault units.3
The effort beginning on L-Day to bridge the reef barrier off the Hagushi beaches bore fruit by 4 April. In place opposite Yontan airfield on Red Beach 1 were ponton causeways that had been side-lifted to the target by LSTs. Earth fill ramps were constructed across the reef to Purple Beach 1 and the Orange Beaches near Kadena. Within the mouth of the Bishi Gawa, close to Yellow Beach 3, a small sand bar had been cleared of surface obstructions and enlarged. A loop access road was then cut through the beach cliff to the bar by engineers with Seabee assistance. As soon as these facilities were ready, cargo from landing craft as large as and including LCTs, could be unloaded directly over the two causeways and the improved sand bar.4
A total of 80 self-propelled barges, also side-carried to Okinawa, was in use constantly from the beginning of unloading operations. The barges were employed in various ways, essentially at the discretion of the division commanders. One barge was assigned to each LST(H) as a landing float onto which the bow ramps of the landing ships were dropped to ease the transfer of casualties from small boats or amphibians. The majority of the barges served as floating supply dumps. These were particularly valuable for supplying critical items to the units ashore at night when cargo ships carrying needed supplies retired from the transport areas.
IIIAC mounted cranes on 12 of these self-propelled units and positioned them at the reef where netted cargo was transferred from boats to LVTs or DUKWs for the final run to inshore dumps. Referring to the demonstrated success of this method, one Marine shore party commander commented:–
This was the [1st Marine Division’s] innovation, first practiced successfully at Peleliu. Two of these barge-mounted cranes were loaned to [the] 6th MarDiv on [L plus 1] to facilitate their cargo handling, and XXIV Corps took up the method. That method accounted for the comparative lack of clutter on the 1st MarDiv beaches. That [the 1st Marine] Division had no beach dumps is a fact of prophetic import for future operations, for I believe establishment of such will invite their destruction in an assault landing.5
Encouraged by the satisfactory tactical picture, Admiral Turner authorized the use of floodlights and night unloading on all beaches starting 2 April, and directed that ships’ holds be cleared of all assault cargo immediately. On the same day, he ordered that the personnel
and equipment of the aviation engineer battalions and the MAGs be expeditiously unloaded. On 3 April, General Geiger recommended to Turner that all priorities established for LSTs unloading over IIIAC beaches be suspended until every member of the airfield headquarters, service, construction, and maintenance units had arrived ashore.
Planned unloading priorities were upset, however, by Tenth Army insistence on getting Yontan and Kadena airfields operational at the earliest possible time, and by General Buckner’s authorization on L plus 2 for corps commanders to bring garrison troops ashore at their discretion. Those on board control vessels and shore party personnel soon viewed many situations wherein low priority units and equipment intermingled with the shoreward flow of essential assault matériel. This interruption of supposedly firm unloading schedules was due, in part, to the natural desires of ships’ captains to unload their vessels and to clear the vulnerable Hagushi anchorage as quickly as possible. The inadequacy of the motor transport available to the shore parties and the radical change in the unloading priorities, however, forced many ships to stand off shore with half-empty holds while awaiting the return of boats which were, meanwhile, stacked up at the control vessels.
Further complicating the critical control problem were the efforts of individual landing boat coxswains who, disregarding their instructions, attempted to “get to the beaches at all costs.” Commenting on this matter, one transport group commander said:
There seemed to exist on the part of most coxswains an almost fierce determination to be first ashore with their individual boats, regardless of the orderly assignment to unloading points, which it is the function of the control vessel to carry out. Coxswains simply would not follow orders to form and remain in cargo circles, but jockeyed for positions of advantage from which to come alongside the control vessel. Many even attempted to ignore the control vessel and bypass it, proceeding directly to whatever beach they had a preference for.6
Despite this, the control of ship-to-shore traffic was probably handled better at Okinawa than in previous Pacific operations, except those at Peleliu and Iwo Jima. After observing the assault landings in the Marianas, Admiral Turner was convinced that only “the most experienced personnel obtainable should be used in the Control Parties for assault landings.”7 Consequently, the key members of the control groups which operated in the Palaus and Bonins served on board the control vessels at Okinawa, where their collective experience helped make ICEBERG a more efficient operation.
Although the ship-to-shore cargo transfer procedures were soon ironed out, problems at the beaches still existed. Organization of the northern landing beaches, for example, progressed slowly. In a critical but friendly evaluation of Marine shore party operations, experienced British observers stated that:–
There seemed to be little or no traffic control, no sign posting of roads or dumps, and no orderly lay-out of the beach areas. It has been said already that the speed of advance inland outran the landing of
vehicles. The rapid landing of [motor transport] therefore became an imperative need and there is no doubt ... that the rate of landing could have been greatly accelerated by proper organization. For instance, although vehicles were able to wade ashore at low tide on Yellow 2, they were only using one exit. This had no beach roadway on it, although its gradient and surface were such that a tractor was frequently required to pull vehicles through it. This considerably retarded progress. (It was noted both here and at other beaches that where beach matting had been laid down, it had usually been cut up by tractors. Separate exits for wheels and tracks is not one of the Marine Shore Party rules!)
It is easy to be critical, but the general impression remains that unloading organization in this sector was insufficiently flexible to cope with the unexpected military situation. However, the Shore Party work in this Corps [IIIAC] must be judged by results, and the fact is that after L-plus 1 day, no serious criticism of the unloading progress was made by the Corps Commander.8
The planned and orderly transition of shore party control to progressively higher troop echelons continued as the beachhead expanded. On 3 April, the XXIV Corps commander took charge of the southern beaches, and, three days later, the commander of the III Amphibious Corps Service Group assumed control for the unloading of the Marine divisions.9 After a conference of responsible fleet and troop logistics officers on board Admiral Turner’s flagship on 8 April, arrangements were made for Tenth Army to take over all shore party activities on the Hagushi beaches the following morning. Major General Fred C. Wallace, the Island Commander, was placed in charge and his 1st Engineer Special Brigade was directed to assume control of all beaches, with the exception of the one which had recently been opened at Nago. In order to operate a much-needed forward supply dump for the far-ranging infantry units of the 6th Marine Division, the IIIAC Service Group retained control of this northern landing point.
Many of the shore party troops in the IIIAC zone of action were from replacement drafts. They had trained with the divisions as infantrymen and accompanied the assault echelon to the target. Until needed to replace casualties in the combat units, these Marines fulfilled a vital function while assigned to shore party and ships’ working parties. Although the weather remained perfect until the afternoon of L plus 3, heavy rain and winds during that night and most of the following day hampered unloading activities. With the abatement of high winds on 6 April, a stepped-up
Yellow Beach 3 on L plus 2. As soon as LCTs and LCVPs are unloaded, others arrive to take their place. (USMC 118214)
Causeways relieve logistical problems as tons of supplies are transported inland. (USMC 118304)
unloading pace resulted in the emptying of 13 APAs and AKAs, and 60 LSTs. The day before, in the midst of the storm, 32 empty cargo and transport vessels left the target area. Between L-Day and 11 April, when the first substantial increment of garrison shipping arrived, unloading over the Hagushi beaches was confined primarily to assault shipping. By noon of 11 April, 532,291 measurement tons of cargo had been unloaded, an amount greater than had been put ashore during the entire course of the Marianas campaign.10
Securing the Eastern Islands and Ie Shima11
Since the rapid sweep of the Tenth Army had cleared the shoreline of Chimu Wan and a large section of the upper portion of Nagagusuku Wan by 5 April, Admiral Turner was anxious to utilize the beaches and berths on the east coast as soon as possible. Although minesweepers were clearing the extensive reaches of both anchorages, before unloading operations could be safely started the Japanese strength on the six small islands guarding the mouths of the two bays had to be determined. To acquire this information, the FMFPac Amphibious Reconnaissance Battalion was attached to the Eastern Islands Attack and Fire Support Group and assigned the mission of scouting the islands. (See Map 11.)
Tsugen Shima, the only island suspected of being heavily defended, was the first target of the battalion. Although Tsugen is relatively small, its position southeast of the Katchin Peninsula effectively controls the entrances to Nakagusuku Wan. Aerial observers reported that the village of Tsugen and the high ridge overlooking it contained extensively developed strongpoints. After midnight, early on 6 April, high-speed APDs carrying the battalion arrived off the objective, and Companies A and B embarked in rubber boats to land on the western coast of the island at 0200. Just a short way inland from the landing point, four civilians were encountered; two were made prisoner, but the other two escaped to alert the garrison.12
Enemy reaction came almost immediately. Company A began receiving machine gun fire from the vicinity of Tsugen, while Company B was similarly taken under fire from a trench system in the northwest part of the island. Japanese mortars soon found the range of the landing party, whereupon the Marines withdrew to the beach under an unceasing shower of shells. Since the battalion assignment was to uncover enemy opposition and not engage it Major Jones re-embarked his unit at
Map 11: Reconnaissance and Capture of the Eastern Islands, 6–11 April 1945
0300. Although the Japanese claimed an easy victory over an “inferior” force,13 the scouts had accomplished their mission. Company A lost two Marines killed and eight wounded.
On the evening of 6 April, Major Jones’ men resumed their investigation of the rest of the islands in the offshore group. At 0015 on 7 April, the entire battalion landed on Ike Shima, the northernmost island. When no sign of enemy troops or installations and only one civilian was discovered there, Company B went on to Takabanare Shima. Landing at 0530, it discovered that 200 thoroughly frightened Okinawan civilians were the island’s only inhabitants. At about the same time, two platoons of Company A went to Heanza Shima and, using their rubber boats, crossed over to Hamahika Shima. Daylight patrols confirmed the absence of enemy soldiers, but 1,500 more civilians were added to those already counted. These islands were occupied later in April by 3/5.14
After nightfall on 7 April, Company B reboarded its APD, which then circled Tsugen Shima to land the Marines on Kutaka Shima, opposite enemy-held Chinen Peninsula. As the company paddled in to shore, the heavy surf capsized three of the boats and one man drowned. The island had neither enemy troops, installations, nor civilians, and the scouts withdrew shortly after midnight.
While the reconnaissance battalion was searching the rest of the Eastern Islands on 7 April, UDT swimmers checked the proposed landing beach on the east coast of Tsugen Shima preparatory to the assault there. The capture of the Eastern Islands had been assigned to the 27th Infantry Division as its part in Phase I of the Tenth Army preferred invasion plan. The information gained from the 6–7 April reconnaissance indicated that commitment of an entire division was not warranted, and only one regiment was assigned for the operation.15
As the main body of the Army division was landing over the Orange beaches near Kadena on 9 April, the ships of the 105th RCT were rendezvousing at Kerama Retto with the command ship of the Eastern Islands Attack and Fire Support Group. The assault unit selected for the landing on Tsugen was 3/105, while the other two battalions of the RCT were designated floating reserve to be called up from Kerama if needed. Although Tsugen had been pounded intermittently by air and naval gunfire since L-Day, the ships’ guns again blasted the island on 10 April, the day of the landing. Initial resistance was light when the soldiers landed at 0839, but the enemy, strongly entrenched in the stone and rubble of Tsugen, soon engaged the invaders in a day-long fire fight. The battle continued throughout the night, during which time the Army battalion sustained many casualties from the incessant enemy mortar fire coming from the heights above the village.
At daylight on the 11th, the rifle companies of 3/105 made a concerted
Marine reconnaissance personnel prepare rubber boats for landings on Eastern Islands. (USMC 120002)
Dominating Ie Shima is Iegusugu Yama, beyond which are the airfields—the primary objectives of the operation. USN 80-G-315059
attack against stubborn opposition which gradually died out. Organized resistance was eliminated by 1530, and the battalion was ordered to embark shortly thereafter to join the rest of the regiment at Kerama Retto. In a day and a half of fighting, the battalion lost 11 men, had 80 wounded, and 3 missing.16 An estimated 234 Japanese were killed and no prisoners were taken. The seizure of Tsugen Shima opened the approaches to Nakagusuku Wan, and ensured that XXIV Corps would receive supply shipments over the eastern as well as the western beaches. This operation also uncovered beaches in Chimu Wan which were developed by the Seabees and used for unloading the LSTs which brought construction supplies and equipment from the Marianas. This action relieved the load which had been placed on the Hagushi beaches, expedited base development, and hastened the building of additional unloading facilities.17
In its rapid advance leading to the capture of the Motobu Peninsula,, the 6th Marine Division demonstrated that Okinawa north of the Ishikawa Isthmus could be taken by an attack overland. ICEBERG commanders were forced in turn to reappraise the original plans for Phases I and II. They found that naval requirements were now reduced to resupply and fire support operations, and that the ships which might have been needed for an amphibious assault of Motobu Peninsula—a possibility considered in all advance planning—were now available for the capture of Ie Shima. Losing no time, Admiral Turner issued the attack order directing the seizure of the island and its vital airfield, and designated the Northern Attack Force commander, Admiral Reifsnider, as Commander, Ie Shima Attack Group.
Ie Shima was important because its size and physical features permitted extensive airfield development. Three and a half miles northwest of Motobu Peninsula, the island plateau was mostly flat land, broken only by low hills and scattered clumps of trees. Located in the middle of the eastern part of the island was a rugged and extremely steep 600-foot-high limestone mountain, Iegusugu Yama. There were few obstacles to widespread construction of airdromes besides this prominent terrain feature. This factor escaped the attention of neither Japanese nor American planners. The enemy had already laid out three runways, each a mile in length, on the central plateau, and the ICEBERG plan called for the expansion of these existing strips as well as the addition of others which would eventually accommodate an entire wing of very-long-range fighter aircraft.
The landing force selected for the invasion was General Bruce’s 77th Infantry Division. After the Keramas landing, this unit spent two weeks on board ship in a convoy which steamed in circles approximately 300 miles southeast of Okinawa. Without warning, on 2 April enemy aircraft dove out of clouds which had hidden their approach and crash-dived four ships (three of which were command ships), before antiaircraft fire could open up on the intruders. The entire regimental
staff of the 305th Infantry was killed and wounded, and the total number of casualties listed in this one attack was 17 soldiers killed, 38 wounded, and 10 missing.18 Ten days after this disaster, the division was committed to land on 16 April, its second assault landing in less than a month. (See Map 12.)
Major Jones’ Amphibious Reconnaissance Battalion was assigned to execute the first mission of the operation. His unit was directed to seize and occupy Minna Shima, a small crescent-shaped island lying 6,500 yards southeast of the main target. Two 105-mm and one 155-mm howitzer battalions from 77th Division artillery were to be emplaced there to provide supporting fires during the Ie Shima battle.
The Marine scouts landed at 0445 on 13 April and within two hours had swept the island. They discovered 30 civilians but found no enemy soldiers. The battalion remained on the island the rest of the 13th and, on the morning of the 14th, occupied positions from which it covered UDT preparations of the reef and beach for the landing of artillery. By noon of 14 April, Major Jones had re-embarked his men on board the APDs. Three days later, the battalion was released from attachment to the 77th Division and attached to IIIAC.
As scheduled, the preliminary bombardment of Ie Shima began at dawn on 16 April and was stepped up at 0725 when missions in direct support of the landing were fired. Five minutes before S-Hour (as the landing time was designated for this operation, 16 fighter planes made a strafing and napalm attack on the beaches while other fighters and bombers orbited over the island, ready to protect the attack group and support the ground assault.
Although there was little opposition to the landing, the troops experienced stiffening resistance by afternoon when enemy delaying groups, concealed in caves and fortified tombs, started to contest every yard of advance. For a period of six days, 77th Division ground forces struggled. Initially making only slight grains, in many cases, they fought hand-to-hand with defenders who contested every inch of ground. As the battle unfolded, it was found that Japanese defenses were centered about Iegusugu Yama and the small village of Ie, which lay at the foot of the southern slope of the mountain. A masterful camouflage job had been performed by the Ie Shima garrison, for nearly 7,000 people were concealed on the island. The mountain contained a maze of hidden firing positions; Ie itself had been converted into a veritable fortress. The ground approaching the mountain and the town was honeycombed with caves, tunnels, bunkers, and spider holes on which the Japanese had expended their great industry and defensive skills. The advance route to the core of enemy defenses was open land and uphill all the way, flanked by Japanese positions in the village and dominated by emplacements located in a reinforced concrete building on a steep rise facing the attacking troops. The infantry soon named this structure “Government House” and the terrain on which it stood “Bloody Ridge.”
Map 12: Assault and Capture of Ie Shima
On 20 April, after a grim grenade and bayonet battle, the top of Bloody Ridge was finally gained and Government House taken. The island was declared secure on 21 April after the 77th Division had won a victory for which a heavy price was exacted; 239 Americans were killed, 879 wounded, and 19 missing.19 Japanese losses were 4,706 killed and 149 captured.
For the next four days, scattered Japanese and Okinawan soldiers were hunted down and, on the 25th, LSTs began shuttling units of the division to Okinawa, where their extra strength was needed in helping the XXIV Corps maintain pressure on enemy defenses in front of Shuri. Remaining in garrison on Ie Shima were the regimental headquarters and the Mt Battalion of the 305th, This force was considered adequate to handle the rest of the cleanup operations in the island.
The Marines’ “Guerrilla Wars”20
The capture of Motobu Peninsula constituted the major portion of IIIAC offensive operations in April. A lesser but continuing Marine task during the period was ridding the area of the pesky and omnipresent guerrillas. Irregulars attempted to harass, delay, and wear down American units by partisan tactics classically employed against patrols, convoys, or isolated detachments.
Once Yae Take fell and Marines advanced to the northernmost reaches of Okinawa, guerrilla activities increased in scope and intensity. Under the conditions offered by the rugged and primitive wilderness of the north, the lack of roads there, and a shortage of information, a modern force of superior strength and armament was unable to engage the guerrilla decisively in his own element.
In the southernmost area of the IIIAC zone, aside from picking off occasional stragglers, Marines were kept busy improving the road net, sealing burial vaults, and closing the honeycomb of caves. To the north, however, as advance elements of General Shepherd’s fast-moving division approached Motobu Peninsula, and the lines of communication were extended progressively, guerrillas took advantage of the situation. During the night of 8–9 April, a group of marauders broke into the area of IIIAC Artillery, near Onna, and destroyed a trailer and a small power plant. Following this attack at dawn, other enemy groups attempted to disrupt north-south traffic passing through Onna by rolling crudely devised demolition charges down upon passing vehicles from the cliffs above.21
Marine rocket launchers in support of the drive south. (USMC 121342)
Awacha pocket, showing the gorge which was the scene of hard fighting by the 5th Marines. (USMC 121104)
In the south of the 6th Division zone, on 7 April the 7th Marines (less 3/7), in corps reserve at Ishikawa, was assigned to patrol tasks.22 The northern half of the regimental patrol sector was covered by Lieutenant Colonel John J. Gormley’s 1/7, which had moved to Chimu, while Berger’s 2nd Battalion and certain designated regimental troop% in a perimeter defense around the bombed-out ruins of Ishikawa, had a related mission of patrolling north and inland from the village.23
One 7th Marines task was warding off nightly infiltration attempts by individual or small groups of Japanese and Okinawan irregulars in search of food. Most of them were killed or wounded either entering the village or leaving it. The initial patrols in the region were without incident, but, as pressure was applied to Colonel Udo’s force in the mountain fastnesses of Motobu Peninsula, the quiet that had prevailed in the supposed-rear zone was dispelled. On 12 April, a 2/7 patrol fell victim to a well-planned ambush on Ishikawa Take, the highest point on the isthmus. By the time that the entrapped Marines were able to pull out under cover of the fires of the regimental weapons company, 5 men had been killed and 30 wounded.
The next day, Lieutenant Colonel Berger sent two companies into the ambush zone and occupied it against only token resistance. In customary partisan fashion, the elusive guerrillas had departed the area, seemingly swallowed up by the heavy vegetation, deep gorges, and spiny ridges of the complex terrain.
After spending a quiet night on the twin peaks of the heights, the two companies, E and F, were withdrawn to approach the guerrilla lair from a different direction. While retiring, the Marines were fired upon from above by the reappearing enemy, and a number of men were hit. After circling to the far (west) side of the island and establishing a skirmish line, the two companies moved in on the commanding ground where the guerrillas were well dug-in and concealed. The irregulars were engaged, but “did not appear to be well organized.”24 Those of the enemy who escaped were hunted down by patrols.
This task proved to be painstaking and time-consuming, for the vegetation on the western slopes of Ishikawa Isthmus seriously hampered effective patrolling despite the fact that this section was the least precipitous in the neck of the island. Visibility off the trails frequently was limited to five feet, at most, by dense stands of bamboo and scrub conifer. Since flank security was impractical in this terrain, the war dogs accompanying the Marines proved a valuable asset in alerting their masters to enemy hidden in the undergrowth.25 Lack of roads and the difficult terrain here raised resupply problems which were solved by the organization of supply pack trains26 to support 2/7
patrols. Enemy resistance continued here for nearly two weeks, during which time Berger’s Marines killed about 110 of the guerrilla force.27
As the 6th Marine Division closed in on the main Japanese position in the Motobu heights, the tempo of guerrilla activity on the fringe of the battle increased proportionately. A daily occurrence at dusk was the harassing of artillery positions by irregulars, who caused the registration of night defensive fires to be delayed.28 When Major Pace’s 1/15, in direct support of the 22nd Marines, displaced to cover the infantry drive to the northernmost limit of the island, its perimeter was hit almost nightly by sporadic sniping and knee mortar fire. In addition, grenades, demolition charges, and even antipersonnel land mines were thrown into the defensive installations encircling the battalion area. The hills in the rear of the 1/15 position afforded the enemy excellent observation and apparently permitted him to coordinate his attacks on the Marines.29
2 From 14 through 16 April, as the battle for Yae Take was coming to a climax, fires mysteriously broke out in various west coast villages from the southern extremity of Nago Wan to the northern tip of the island. On 17 April at dawn, Nakaoshi was struck by an enemy hit-and-run attack that simultaneously swept over the 6th Engineer Battalion command post (CP), water point, and supply installations nearby. Civilian collaboration with Japanese military forces appeared to be a factor in these incidents, when evidence of native sabotage was uncovered during an investigation of the series of fires on the west coast.
The security threat presented by Okinawan civilians appeared to be pervasive, for it arose within the 1st Marine Division zone also. As early as 9 April, Lieutenant Colonel Miller, the 3/5 commander, reported that many civilians were destroying their passes and appeared to be roaming about freely at night. It was reasonable to assume that they were contacting the Japanese at this time.30
For better zonal security control, the 1st Marine Division began rounding up all civilians on 11 April and herding them into stockades built on Katchin Peninsula. The following day, all able-bodied Okinawan males were taken into custody in order to determine their military status. The prevailing tactical situation in the north at this time, required that organized resistance be broken before Marine control over civilians could be established and combat troops spared for this duty.
From the beginning of the 6th Division drive north, an increasing number of Okinawans was encountered on the roads. Only a few men were of obvious military age and were detained. The others, stopped and questioned, were allowed to continue on with their affairs. At the height of operations in the north, 12–16 April, the division was unable to collect able-bodied males methodically in the manner of the 1st Division in central Okinawa. Civilians
of doubtful character and background, however, were seized. When hostilities on Motobu ceased, the 6th Division organized a civilian control center at Taira where, beginning 16 April, from 500 to 1,500 natives were interned daily until operations in the north were ended.
On 15 April, Hurst’s 3/7 (attached earlier to the 5th Marines) reverted to parent control and began active patrolling from its base at Chuda on the west coast. General del Vane regained the 7th Marines the next day, and, as the 6th Division began meeting increased resistance, the boundary between the Marine divisions was readjusted along the Chuda–Madaira road.31
From 17 to 19 April, it appeared that, parallel to the steady reduction of their positions on Yae Take, the Japanese were shifting from a tactical policy of defense to one based on partisan warfare. After the 6th Division took the mountain redoubt, and following a reorganization of Marine units, General Shepherd’s command moved to assigned garrison areas. Here it began patrolling vigorously to fix and destroy remaining pockets of enemy resistance. To assist the division in securing northern Okinawa, the Amphibious Reconnaissance Battalion, part of IIIAC since 17 April,, was attached with a mission of seizing and occupying the small islands lying off Motobu Peninsula.
In a period of two days, 21–22 April, the battalion reconnoitered the islands of Yagachi and Sesoko with negative results. Though no enemy forces were encountered, the Marines found a leper colony containing some 800 adults and 50 children on Yagachi Shima. Before they landed on Sesoko, the scouts met more than 100 natives moving by canoe from islands to the west in search of food, and “considerable difficulty was involved in corralling and controlling” them.32 On the 23rd, Walker’s 6th Reconnaissance Company scouted Kouri Shima and found no enemy.
While the battle for Yae Take raged, and even after it had ended, 6th Division rear area patrols began making contacts with enemy troops attempting to escape from the fighting on Motobu. On 22 April, near Nakaoshi, 1/22 patrols killed 35 enemy in a fire fight. On the next day, this battalion met a strong force, estimated at three rifle squads, three light machine gun squads, and one mortar squad, firmly entrenched in previously prepared positions, including caves and pillboxes, in the mountainous area east of Nago. Two Marine companies assaulted the Japanese killing 52, before an ammunition shortage forced the battalion to break off the action. It returned to the battle scene on 24 April, this time with 4/15 (Lieutenant Colonel Bruce T. Hemphill) in direct support, and the strongpoint was
reduced.33 The engagement ended towards evening with the deaths of a Japanese officer and two NCOS and the remainder of the group fleeing. The battalion continued patrolling the region on the next day and cleaned out the enemy pocket.34
Intensified patrolling of the Ishikawa Isthmus began on 23 April after a small IIIAC military police group was extricated from an ambush by a 7th Marines detachment. The 2nd Battalion, 1st Marines, reinforced the 7th, and all available 1st Division war dogs were attached to that regiment. At the same time, stricter travel regulations within the IIIAC area were enforced, and the movement of a single vehicles in the corps zone during hours of darkness was forbidden.
In the 6th Division zone, while the 29th Marines remained on Motobu Peninsula, the 4th Marines moved to its assigned area in the northern part of the island. At Kawada, 3/1 was relieved by Hochmuth’s 3/4 and returned to parent control on the 23rd. During the next two days, the rest of Colonel Shapley’s regiment was disposed with Hayden’s 2/4 at Ora, and Beans’ 1/4, regimental troops, and the headquarters complement bivouacked in the vicinity of Genka, a small west coast village located about five miles north of the juncture between Motobu Peninsula and the rest of the island. From this point, Colonel Shapley’s mission was to seek and exterminate stragglers in the southern half of what had been the 22nd Marines area. Upon being relieved, 1/22 prepared to move to the west coast to a point just south of Ichi, which had been the 3/22 patrol base since 16 April.
The mountainous interior of the north was combed continually by Marine patrols for Udo Force survivors and semi-independent guerrilla bands. The 6th Division learned from civilians in the area that small groups of Okinawan home guardsmen were in the hills of the northern part of the island and had been preparing to wage partisan warfare for nearly a year. As part of the preparations, they had reportedly established stockpiles of supplies in the interior, The civilians further stated that some of the guardsmen had returned to their homes and civilian pursuits. They also said that home defense units were being trained in the villages by Okinawan veterans who had served previously in China with Japanese forces.35
Until the afternoon of 27 April, however, patrol results were negative with the exception of an occasional flushing out of individuals or small groups. At this time, a 3/4 reconnaissance patrol sighted a 200-man enemy column moving through the northeastern corner of the Marine regimental zone toward the east coast. It was believed that these Japanese had survived the Motobu Peninsula fighting by infiltrating in groups of 20 to 40 from the combat area by way of Taira and that they were going to try to join up with the main enemy force in the south.
Steps were taken immediately to destroy the group. Two battalions of the 22nd Marines were ordered to the south to block the column, while 3/4 moved
inland from Kawada. Further ringing the escape-minded enemy was Donohoo’s 3/22, which proceeded toward the interior on a cross-island trail 1,000 yards north of, and parallel to, the 1/22 advance from Hentona. Since it was anticipated that the fugitives would be apprehended in the 22nd Marines zone, 3/4 was attached to that regiment, Additionally, two artillery battalions were to support the pursuers.
The first contact was made just prior to noon on 28 April, when one of 3/4’s companies engaged the escaping Japanese in a fire fight. At the end of the three-hour contest, 109 enemy soldiers were dead; 1 Marine was killed and 8 wounded.36 The other pursuing units were unable to reach the scene of the action because of the difficult terrain; 1/22 encountered small scattered groups as it advanced, while 3/22 was still underway when 3/4 radioed that it had destroyed the enemy. Thereupon, Colonel Schneider ordered his 3rd Battalion to continue on to the east coast, and Colonel Shapley’s 3/4 returned to Kawada and parent control.37
Even though the guerrillas in the IIIAC area had forced the Marines to remain constantly on the alert General Geiger was able to declare the end of organized resistance in the north on 20 April. Continuous patrolling remained the general order, however. As usual in counter-guerrilla operations, the number of combat troops employed was out of proportion to the size and number of guerrillas hunted. In most cases, it was a one-sided fight, for a substantial percentage of the partisan ranks were filled with the poorly trained and equipped Boeitai. The primary contribution of native Okinawans to the guerrilla effort was a knowledge of the land over which they fought; their offensive efforts were limited mainly to night forays against supply installations, disrupting communications systems and centrals, and attacking water points and hospitals. Although these destructive attempts usually ended in failure, they forced friendly units to maintain extensive security detachments, sometimes in platoon or company strength.
As of the 20th, when Motobu Peninsula was reportedly cleared of enemy troops, the Tenth Army began to pay greater attention to the native population in occupied sections of the island. All civilians, irrespective of age or sex, found in the areas of combat units were to be interned. Furthermore, Okinawans were prohibited from moving about freely unless accompanied by an armed guard.38 General Geiger established eight internment camps in the IIIAC zone, but the number of collection points in the Marine area was later reduced
to three; Katchin Peninsula, Chimu, and Taira.
Although tighter security controls prevailed in the corps zone, isolated incidents behind the battle line still occurred. In the last week of April, a 7th Marines patrol killed a Japanese corporal who was wearing a kimono over his uniform. Intelligence agencies found evidence of a Japanese-planned and -sponsored program of espionage and sabotage for the rear areas. In the XXIV Corps zone, the following document was recovered:
Permit
Army line probational officer Inoye Kuchi and two others: The above mentioned are permitted to wear plain clothes for the purpose of penetrating and raiding enemy territory from April 25, 1945, until the accomplishment of their mission.39
TAF Operations in April and the Kamikaze Threat40
Owing to the early and unopposed capture of Yontan and Kadena airfields, Tactical Air Force, Tenth Army, began land-based operations sooner than expected. On 2 April, General Mulcahy and his staff went ashore and selected a CP site midway between the two fields. General Wallace’s ADC headquarters was dug in nearby.
While TAF personnel were kept busy constructing camp and repair facilities, Marine engineers and Seabees began repairing the runways on Yontan and Kadena. The airfields were found to be lightly surfaced and badly damaged by naval gunfire and bombings. Hurried grading permitted the use of Yontan by 7 April, but the problems at Kadena were more extensive. Damage here was greater, and the source of coral for surfacing was at some distance from the field. Nevertheless, the strips on Kadena were ready for dry-weather use two days after those on Yontan and, by 1 May, they were all-weather operational.
Three weeks earlier, the ADC Air Defense Control Center (ADCC) had come ashore and, on 7 April, begun operating from three LVTs specially rigged to serve as the defense command CP and to function as both an ADCC and the Air Defense/Fighter Command operations center. On 19 April the center moved to more spacious quarters in an abandoned farmhouse nearby.
When General Wallace opened his CP on Okinawa, the air defense commander became the land-based agent of CASCU, which continued operating on board Admiral Turner’s flagship. Under ADC operational control were land-based aircraft, radar air warning and control installations, and antiaircraft artillery units. It was the air defense commander’s primary mission to coordinate the combined efforts of these three disparate support activities so that they meshed with the operations of the overall air defense system of the expeditionary force. ICEBERG plans had stipulated that TAF would assume full responsibility for the air defense of Okinawa when the amphibious landings were completed, but, because of “the all-out efforts of Japanese aircraft and the success of their kamikaze suicide attacks directed against naval units, operational control of aircraft in the Ryukyus remained with the Navy until the area was secured.”41
General Wallace believed that the major tactical task of ADC was to meet the Kamikaze threat. From 7 April, when VMF-311 pilots scored the first TAF kill42 of a suicider as they flew in to Yontan from their CVE lift, ADC efforts were directed toward confronting and stopping the destructive enemy air attacks. The fighter squadrons of MAG-31 and -33 mounted combat air patrols from Yontan and Kadena fields on the first days that they arrived at these bases.
As the battle was joined on Okinawa by the Tenth Army and General Ushijima’s forces, the American fleet in surrounding waters was engaged in a desperate battle of its own. The Japanese air attacks on the Kerama Retto invasion group merely heralded even greater enemy attempts to destroy the radar pickets and support vessels safeguarding the troops on Okinawa. Many of these enemy aircraft were on either conventional bombing or reconnaissance missions; others in the aerial attacks were part of the Special Attack Force, the Kamikazes.
As the success of American operations in the Philippines became apparent and MacArthur’s air strength reigned supreme, enemy naval air commanders saw that there was no prospect of any advantage to be gained in the sky while Japanese squadrons continued employing orthodox tactics. The Kamikaze effort evolved as a result of these considerations. Appearing first in the Philippines,43 this was an organized and
desperate attempt by suicide-bent Japanese naval aviators to deprive American shipping at Leyte of aerial protection by crashing the flattops of the covering carrier force. The enemy anticipated that the success of their tactics would then guarantee a Japanese surface victory in the event of an all-out engagement with United States naval forces. Although Japanese commanders felt that suicide missions were a “temporary expedient” only, used “because we were incapable of combatting you by other means ...,”44 initial success gave added impetus to their fuller employment.
Correctly anticipating that the next invasion attempts would be at Iwo Jima and, after that, Okinawa, Imperial General Headquarters withdrew the remnants of some Army and Navy air units from the Philippines in early January 1945 to strengthen the defense of the Home Islands and the Ryukyus. Upon completion of this transfer, designed to “produce a more unified [defense] strategy,”45 brigades and regiments of the Sixth Air Army and naval squadrons of the Fifth Air Fleet were combined into a single tactical command on 19 March under Admiral Soemu Toyoda, Commander in Chief, Combined Fleet.46 At the outset, it was determined that operations of this combined force of about 1,815 planes47 were to be well planned and organized—a definite contrast to the sporadic, albeit somewhat successful, Kamikaze attacks at Leyte.
One of the first opportunities for the Japanese to mount coordinated suicide and conventional air attacks occurred during the TF 58 raids of 18–19 March 1945 on Japan. Although the carriers were damaged and there were some American casualties, the enemy lost 161 aircraft. Most of this damage, strangely enough, was not caused by Kamikazes. An important result of this raid was the destruction, while still on the ground on Kyushu, of many of the Japanese planes scheduled to be employed in the defense of the Ryukyus. This disaster forced the Fifth Air Fleet to reevaluate its plans. Moreover, a Tenth Army landing relatively unharassed by enemy air raids was guaranteed, for Toyoda’s squadrons were unable to mount a major air offensive until after the beginning of April.48
Scattered conventional and Kamikaze flights from Japan and Formosa carried the attack to the Western Islands Attack Group of the ICEBERG force first; later these planes began swarming all
over the transports and picket line off Okinawa. During the first few days of April, the toll of ships damaged and sunk grew at a steady rate while naval casualties mounted in consequence. By 6 April, Admiral Toyoda was prepared to launch from Kyushu the first of ten carefully planned Kamikaze attacks, which were to be flown over a period ending 22 June. A total of 1,465 sorties emanated from Kyushu to sink 26 American ships and damage 164 others.49 Not included in these loss figures are the victims of small-scale Kamikaze efforts by another 250 planes which rose from Formosa air bases, and the 185 additional sorties flown from Kyushu, independent of the mass attacks.50
The Japanese decision to turn to large-scale air operations was arrived at after Toyoda had studied both his and the Thirty-second Army situations and had found that “it would be futile to turn the tide of battle with present tactics.”51 He therefore dispatched the first and largest coordinated suicide attack—Kikusui Operation No. 152—against ICEBERG forces on 6 April. Spearheading the Kamikazes were 14 planes sent to bomb and strafe Okinawa airfields before dawn in order to destroy Allied aircraft suspected of being there. Apart from their nuisance value, the raids did little damage to the runways and none to TAF planes, for the squadrons had not yet flown ashore. Following the first group of enemy hecklers were more than 100 fighters and bombers sent to engage TF 58 off Amami-O-Shima in order to draw American carrier-based planes away from the suiciders heading for Okinawa.
For a 36-hour period, 6–7 April, the Japanese flew 355 suicide sortie, which were accompanied by nearly an equal amount of conventional cover, reconnaissance, and bombing planes. As these aircraft bore in to crash, torpedo, and bomb the ships at anchor in Hagushi transport area, crewmen in exposed positions and troops on the beaches were subjected to a deadly rain of antiaircraft artillery shell fragments. Friendly fighters were not immune from the effect of the hundreds of guns firing from the beaches and ships; three American pilots were shot down when they followed Japanese planes too closely into the murderous barrage.
The main attack, which began about 1500 on 6 April, spread out all over the combat zone with the outer ring of radar pickets and patrol craft—lacking a protective smoke-screen cover—catching the full fury of the battle. Ships of all types, however, were fair game for
the Kamikazes.53 Before the Okinawa landing, the Japanese confined the direction of their suicide attack efforts to American carrier task forces. After 1 April, the attacks were mounted against convoys, and, just prior to the first Kikusui, the enemy began hitting all surface forces. After the time of the 6–7 April attack, the Japanese reserved the carrier forces for Kamikaze attention while their conventional bombers and fighter craft were directed to hit other American vessels and transports around Okinawa.54 In this first mass suicide attack, Admiral Turner’s forces claimed to have shot down at least 135 Japanese planes, while the pilots from the Fast Carrier Task Force reported splashing approximately 245 more, bringing the total American claims of enemy losses to nearly 400 pilots and planes. Contemporary Japanese sources place the losses in Kikusui No. 1 at 335.55
As a sidelight to the air battle over and the land fighting on Okinawa, the Japanese mounted their only real surface threat to the success of the American invasion. Intending to attack Allied shipping at Okinawa, the 69,100-ton battleship Yamato and a covering group steamed out of the Tokuyama Naval Base, on Honshu, at 1500 on 6 April. Less than two hours later, the enemy vessels were sighted by two U.S. submarines in the screen lying off the east coast of Kyushu. Within 24 hours, TF 58 pilots had administered death blows to the Yamato and a part of her group, and had forced the remainder to scurry home.
Since TAF pilots had not yet begun operations from Okinawa when Kikusui No. 1 struck, the four Marine squadrons on board the carriers Bennington (VMF-112 and -123) and Bunker Hill (VMF-221 and -451) carried the ball for Marine aviation during the time that General Mulcahy’s planes and pilots were still on board their carrier transports. Until late in April, as much as 60 percent of the ground support missions flown for Tenth Army units were carried out by Navy and Marine carrier pilots, while the primary concern of TAF flyers was to blunt the
Kamikaze menace. To at least one TAF air group commander, “it seemed strange for planes off the carriers to come in for close-support missions, passing [Okinawa-based] Marine pilots flying out for CAP duty. ...”56
Almost as soon as Colonel Munn’s MAG-31 squadrons touched down at Yontan, a 12-plane combat air patrol was organized and launched to remain airborne until dark. Prior to the time that TAF joined the fighting, CAPs had been flown by planes from both the Support Carrier Group and TF 58. Originally, a large CAP, varying from 48 planes in relatively quiet periods to 120 or more during critical times, was flown to protect the surface forces from air attacks. Basically, the aircraft were deployed “in a circle in depth” over the invasion and picket craft.57
Generally, TAF planes were airborne from dawn to dusk on CAP flights, and they flew special early morning and twilight CAPs as well. On 14 April, the commander of the ICEBERG operation transferred the responsibility for flying night CAPs from TF 58 to TAF. In addition, TAF was to maintain another four planes constantly on patrol during the hours of darkness. This last mission was assigned alternately to the night fighters of VMF(N)-542 and -543 commanded by Majors William C. Kellum and Clair “C” Chamberlain, respectively. In order to guard the radar picket ships—special objects of the Kamikaze attacks—General Wallace’s fighter command was ordered on 14 April to maintain a continuous two-plane daylight CAP over each of the three picket ships that were stationed offshore northeast of TAF airfields. Each flight leader was to report directly to the captain of the ship he was guarding. In turn, the naval officer would control the flight and ensure that its planes were kept out of range of the ship’s antiaircraft guns. Two days after this mission was first initiated, the number of ships protected by this CAP was increased by two.
By the time that TAF had been established ashore, the three Marine Landing Force Air Support Control Units, commanded overall by Colonel Vernon E. Megee, had landed also. Although they were shore-based representatives of CASCU and outside of the TAF chain of command, by the very nature of their functions the LFASCUs worked closely with the Marine aviation units. Once air support operations began, coordinating agencies relayed all orders concerning aircraft missions directly to General Mulcahy’s command in a smoothly functioning system. At Tenth Army headquarters, Megee’s LFASCU-3 screened all requests for air support received from LFASCU-1 (Colonel Kenneth H. Weir) and -2 (Colonel Kenneth D. Kerby) which were working with IIIAC and XXIV Corps respectively. If a review of TAF and carrier aircraft commitments indicated that an air support request was consistent with priority requirements, the mission was approved. At that time, if Marine planes were assigned, LFASCU-3 relayed the
order for the mission directly to the TAF operations section.58
Frontline control of the ground support missions flown by both land- and carrier-based aircraft was provided by Air Liaison Parties (ALPs) from the Joint Assault Signal Companies attached to each division. Ground unit requests for air support were reviewed first with respect to the capabilities and availability of the other supporting arms to fulfill a specific mission, and then passed on to the LFASCU at corps headquarters. If the request was approved here, the LFASCU would requisition the necessary number and types of planes, and stipulate the armament they needed for successful completion of the mission. In addition., the LFASCU provided strike direction and supervised the scheduling of all air support in unit fire support plans.59
Not all close air support missions were ground controlled in this campaign. Employed at Okinawa was an air coordinator, or airborne traffic director, who spotted and marked the ground target for the planes flying the mission. The coordinator would direct the flight to the best target heading, observe attack results, and correct subsequent runs if he decided that they were needed.60 At times when smoke and weather conditions over the target denied the airborne controller suitable visibility, the support mission would be run nevertheless, but directed by the ALP.
It took time to establish land-based radar reporting, control, and homing stations on Okinawa and the outlying islands. In addition to the problems involved in getting the Air Warning Squadrons (AWSs) and their equipment ashore rapidly, initially it proved difficult to net the ground-to-ground communications systems with the overall ship-to-shore warning system. Prior to the establishment of the ADCC, the individual radar stations had reported directly to CASCU aboard the Eldorado. After 8 April, the day on which the control center first began to provide shore-based operational homing facilities, AWS early warning teams began reporting directly to the ADCC which, in turn, passed on to Navy control the reported enemy and friendly plots.
Early warning teams were also assigned temporarily to each assault division and corps headquarters. They then operated in coordination with the AAA units already assigned to the defense of corps and division sectors. Here, the teams monitored ships’ radar telling circuits and local air warning and inter-fighter director nets, from which air raid warning information was obtained and passed on to the ground units. In addition to radar coverage, the air warning squadrons provided radio monitoring services, the results of which figured prominently in and assisted the operation of the Air Defense Command.
The AWSs also worked very closely with and were, in fact, supervised by
the senior Marine AAA officer, Colonel Kenneth W. Benner, commanding the 1st Provisional Antiaircraft Group. He was responsible for coordinating the disposition and operation of his organic radar with that of the AWSs in order to ensure maximum surface and low-angle electronic surveillance for defense against enemy air attacks.61
Because theirs was a vital role in the overall air defense of the ICEBERG forces, land-based AAA units, although attached to the assault corps, were directly under the operational control of General Wallace’s ADCC. On 20 April, the antiaircraft units reverted to the Tenth Army which then assigned the 53rd Antiaircraft Artillery Brigade the mission of coordinating all AAA activities. At the same time, the brigade became the TAF agency for providing the ground forces with early air raid warning services while continuing to fulfill its AAA defense mission.
Initially, the 1st Provisional Antiaircraft Artillery Group was assigned to support IIIAC during Phases I and 11 of the operation by providing AAA defense for corps units, installations, and beaches, and the captured airfields in the corps zone.62 Additionally, the group was to provide antiboat defense of corps beaches, supplement field artillery units in both direct and general support mission, and be prepared to fire seacoast artillery missions.63
The assault elements of the group were the 2nd (Lieutenant Colonel Max C. Chapman) and 16th (Lieutenant Colonel August F. Penzoll, Jr.) Antiaircraft Artillery Battalions, which supported the 6th and 1st Marine Divisions respectively. Scheduled to land on order at later dates, the 5th (Lieutenant Colonel Harry O. Smith, Jr.) and 8th (Lieutenant Colonel James S. O’Halloran) AAA Battalions were to reinforce the group and extend antiaircraft defenses already existing.
Because of the rapid progress of the infantry and the assignment of higher priority to items needed ashore immediately, the landing of the Marine AAA battalions was delayed. Group and battalion reconnaissance parties landed on L plus 2 to select sites, and beginning on 5 April, the units themselves were given an unloading priority. By 12 April, the battalions were in position ashore.
Initially, one heavy and two light AAA batteries of the 2nd Battalion were assigned a defense sector on 6th Marine Division beaches; the 16th Battalion supported the 1st Division with two heavy and two light batteries. The remaining five 90-mm gun batteries of the group defended Yontan air field.64
When the 53rd Antiaircraft Artillery Brigade assumed control of Tenth Army AAA units, it found that the defenses in the IIIAC and XXIV Corps zones were unbalanced. On 27 April, the brigade adjusted the dispositions and,
in addition, extended AAA defenses across the island to the east coast in order to break up enemy raids coming from that direction. The 1st Provisional Group continued the Yontan area defense, but was made responsible for defense of the entire Yontan–Kadena sector also. To aid in this last mission, Colonel Benner’s group was augmented by two Army AAA-Automatic Weapons battalions.65
Lieutenant Colonel O’Halloran’s 8th Antiaircraft Battalion landed at Nago Wan on 17 April to defend IIIAC units and supply dumps in that area. Its most immediate problem in view of the tactical situation on the Motobu Peninsula, was achieving ground security.66 For that reason, battalion .50-caliber heavy machine guns and some .30-caliber light machine guns obtained locally were assigned a primary mission of ground defense.
Enemy air attacks on shore installations were directed at Yontan and Kadena airfields mainly, and usually took place at night. The only firing opportunities afforded shore-based antiaircraft artillery during daylight occurred when Japanese aircraft, with the obvious intention of attacking the transport area, made their approaches from the landward side of the anchorage.67 Usually, the illumination or visual sighting of an enemy plane, and sometimes even an American one, was the signal for a wave of wild uncontrolled firing both from shipboard and the island. “Carbines, rifles, and even .45 caliber pistols enthusiastically joined the fun on occasion.”68 Under these conditions, casualties and materiel damage resulted from falling shell fragments and wild shots until the Tenth Army insisted upon the enforcement of greater fire discipline by all unit commanders.
From the beginning of ADCC operations, there was no satisfactory communications and control system linking the fighter command and the antiaircraft artillery units. When ADC was heavily engaged with enemy air attacks, liaison with AAA units weakened or broke down completely. At times, permission for the guns to fire on unidentified or enemy planes was withheld, even when the area was definitely under attack. On several occasions, air raid warning flashes were not relayed to the AAA command until after the infantry and shore party units had been informed.69 Of necessity, an efficient control system was soon initiated. By the end of the month, Marine AAA units were credited with the destruction of 15 planes and 8 assists, 5 probably destroyed, and 6 damaged.70
It was noted that, during enemy air attacks, Japanese aviators showed an increasing knowledge of radar evasion measures, and frequently used “window”71 in both conventional and Kamikaze attacks. In commenting on the enemy failure to mount air attacks on the ground forces, one observer stated that “it was difficult to understand why they had not resorted to formation bombing from low altitudes,”72 since low-angle radar detection of approaching aircraft was almost impossible. He concluded that the concentration on suicide attacks was too great; he might have added that the Japanese just did not have enough planes by this time to divert their air strength to missions other than the Kamikaze attacks.
While TAF fighter pilots added to the expanding bag of downed enemy planes, other types of air missions in support of the Tenth Army were performed at the same time by General Mulcahy’s command. Upon its arrival, the Army Air Forces’ 28th Photo Reconnaissance Squadron rephotographed the entire Okinawa Gunto area to obtain more accurate and complete coverage than had been available for the maps used on L-Day. The squadron also provided infantry commanders with enlarged aerial photographs of masked terrain features to their zones. As soon as Major Allan L. Feldmeier’s VMTB-232 arrived on 22 April, it was given tasks other than its original mission of antisubmarine warfare.73 During the remainder of the month, the squadron flew numerous artillery observation missions daily, bombed and strafed enemy lines and installations in southern Okinawa, and conducted heckling raids in the same areas almost nightly.
The second mass Kamikaze attack took place during 12–13 April. Although as frenzied and almost as destructive as the first attack, it was mounted by only 392 planes, on both conventional and suicide missions,74 as opposed to the 699 total in the first attack. As in Kikusui No. 1, TF 58 pilots downed most of the enemy, but carrier-based Marine flyers were active also. Flying Leathernecks from the Bennington shot down 26, and Bunker Hill Marines downed 25. Okinawa-based TAF pilots accounted for 16 more.
During the interval between the first and second mass raids, the Japanese command had recognized the threat presented to their air attacks by American land-based aircraft, so Kadena was bombed early on the 15th by planes that preceded the Kamikazes. TAF personnel and airplanes were endangered further when both of the fields occupied by Tenth Army squadrons were fired upon by an artillery piece, or pieces, nicknamed “Kadena Pete” in not-too-fond memory of
‘Pistol Pete’ at Henderson Field on Guadalcanal.”75
TAF reports evaluating Kikusui No. 2 noted that the evasive tactics employed by the enemy “do not tend to indicate that the flyers were top-flight fighter pilots,” and that “a definite lack of aggressiveness” seemed “to confirm the belief that the pilots were green.”76
A third mass raid of 498 aircraft (196 suiciders) occurred 15–16 April. As the furious air battle carried over into the second day, TAF planes began to score heavily. The largest bag made by land-based aircraft to that date was accomplished by VMF-441 (Major Robert O. White) pilots, who had shot down 17 of the 270 Japanese pilots and planes allegedly splashed on these two days.77
In this attack, a TAF pilot made the first sighting of the so-called “Baka” bomb78 in its maiden appearance over Okinawa, This small, single-engined, wooden craft, powered by rockets, carried a one-man crew and over a ton and a half of explosives. Carried by a twin-engined bomber to a point near the target, the Baka was released when its pilot had verified the weapon’s target and position, oriented his own position, and started the rocket motors. Although the destructive powers of the Baka were real, its employment was erratic and it appeared too late in the war to be influential.
TAF operations for the rest of April tended to fall into a routine of CAPs and support missions. On 22 April, the dusk air patrol was vectored to a point over part of the radar picket line then being attacked by enemy aircraft. When the half-hour battle had ended, Marine pilots claimed 333A Japanese planes. Five days later, during the fourth mass Kamikaze attack (27–28 April), 115 suicide-bent Japanese pilots were launched against friendly shipping and the steadfast radar pickets. On the second day of the attack, at about 1600, the airborne TAF CAP and an additional 36 Corsairs were vectored out 40 miles northwest of Okinawa to intercept an approaching Kamikaze formation. After dark, when the two-hour fight was over, the Marine fighters were credited with downing 35½ enemy planes. Upon
being congratulated by the Tenth Army commander on the accomplishments of his pilots, General Mulcahy Sent a message to the ADC: “Not only brilliant work by fighter pilots but excellent command control and most efficient reservicing by ground personnel were admiration of and inspiration to all.”79 By the end of April, TAF pilots had flown 3,521 CAP sorties and shot down or assisted in the downing of 143¾ enemy aircraft.