The ‘I-Go’ Operation

Admiral Isoroku Yamamoto, commander of Combined Fleet, transferred headquarters to Rabaul for Operation I-Go.

A2-2-152 was flown by a wingman of carrier Junyo.

2-1-202 was flown by a wingman of carrier Junyo. Solomon islands, during I-GO operation, April 1943. This plane shows two vertical red bands that means when Junyo was second carrier of 2nd division.
Beginning with the I-Go Operation in April 1943, photos indicate that the aircraft of either of the two carrier divisions assigned to temporary duty in the Solomon islands over painted both the A prefix in their tail codes and the fuselage vertical bands in some aircraft.


The Allies now possessed the initiative in New Guinea and among the Solomons. Their supremacy rested upon their air superiority, due not so much to the quality of material as to the tactics forced upon AirSols by the aggressive Japanese, whose losses in actual combat were far exceeded by the losses involved in all-weather flying at great ranges from their bases. While many American aviators were rescued, or were even able to regain their bases in damaged aircraft, damaged Japanese aircraft were less often able to return the 350 to 650 miles to Buin, Buka or Rabaul, and their chances of rescue after a forced landing were slim.


With the departure of the IJAAF fighters from Rabaul, only the IJNAF’s 11th Air Fleet remained to provide aircraft for the South-West Pacific area. The backbone of the force was made up by eighty-six A6Ms and seventy-two G4M ‘Bettys’, supported by lesser numbers of D3As and B5Ns. The Japanese Army had given up its offensive in New Guinea, but was now intending to stabilise a deteriorating situation by strengthening existing garrisons. At the same time, concern was felt about the American intentions out in the Solomons. To carry the necessary New Guinea movements and to check the Solomons build-up it was clear that temporary air superiority would have to be achieved over both areas. As the IJAAF Staff were intransigent over the matter of providing air units, the IJNAF would have to find aircraft from its own resources.


Only one source was immediately available. The Air Flotillas in the East Indies, Marianas and on Formosa had all had their most experienced personnel siphoned off for the 11th Air Fleet, and only the air groups of the First Carrier Division, Zuikaku, Shokaku and Zuiho, could provide a balanced reinforcement. The groups had been trained after the Battle of Santa Cruz and were only just attaining the level of expertise necessary to engage the American carriers, so that there was considerable opposition from the Combined Fleet (Admiral I. Yamamoto) and the Third Fleet (the operational carriers) commanded by Vice Admiral T. Ozawa, who had relieved Vice Admiral Nagumo in November 1942. The disaster in the Bismarck Sea forced the IJN to take the first fatal step at the end of March 1943, and by 1 April 161 A6Ms and D3As and a score of B5Ns had arrived at Rabaul, Buka, Ballale and Kahili. The dive-bombers and torpedo-bombers were on the Bougainville airfields to bring them within range of Guadalcanal, but the ‘Zeros’ had sufficient range to accompany the ‘Bettys’ all the way from Rabaul to the targets in New Guinea and the Solomons.


Operation PA, as the Solomons neutralisation was codenamed, started on 1 April 1943, with a sweep by fifty-eight A6Ms of the 11th Air Fleet. These aircraft were supposed to inflict maximum casualties upon any American fighters found airborne, but over the Russell Islands they were met by a mixed force of forty-one USMC F4F Wildcats, USAAF Lockheed P-38 Lightnings and USMC F4U-1 Corsairs, the last based at Henderson Field since mid-February. Only six of the defending fighters were lost, but Japanese losses from all causes amounted to eighteen, nearly a third of the force, while the returning pilots made extravagant claims.


The disembarked carrier aircraft were not ready for their first attack until 7 April, when eighty-nine fighters and forty-nine dive-bombers, reinforced by twenty-one fighters and eighteen dive-bombers of the 11th Air Fleet, rendezvoused over Kahili and made for Guadalcanal to attack shipping. The Japanese had chosen a good day for their strike, for there were some forty worthwhile targets in Lunga Roads. Coastwatchers gave ample warning of the approach of the raid, for three hours elapsed between the first report and the development of the attack, which was carried out in a series of waves.


Seventy-six American fighters intercepted, but of these only twenty went for the sixty-seven ‘Vals’, which, not surprisingly, broke through and attacked the ships. A Royal New Zealand Navy A/S trawler was sunk in Tulagi harbour, as were an oiler and a destroyer. The only other damage was to the hulk of a previous casualty, in use as a fuelling wharf. A dozen or so ‘Vals’ were shot down by fighters and the ships’ AA fire, as well as nine of the escorting A6Ms; the American fighters had claimed twenty-seven. The Japanese aircrew who returned overclaimed once again, to the extent that Admiral Yamamoto, who was directing the ‘I’ Operation, believed that the strike had achieved a major success, drawing the teeth of AirSols and weakening Allied merchant and warship potential for some while to come. In fact the fighters defending Lunga Roads had lost only seven of their number, and all but one of the pilots had been recovered.


With Guadalcanal neutralised to his satisfaction, Admiral Yamamoto switched his aircraft to the New Guinea combat zone. On 11, 12 and 14 April no fewer than 456 sorties were flown against Allied transports in Oro Bay, near Buna, and in Milne Bay, and against the Port Moresby airfields. These raids, four altogether, sank one ship and damaged five others: five USAAF and RAAF fighters were shot down, and the IJNAF lost eighteen aircraft. With air reports of a substantial victory, amounting to three warships and twenty-five transports sunk and many others damaged, and 175 Allied aircraft shot down, Admiral Yamamoto decided that the ‘I’ Operation had achieved its aim by 16 April, and it was terminated. A few days later the carrier air groups flew out to Truk, less the thirty aircraft lost during their five strikes. On 18 April, shortly before their departure, they had been standing by at Kahili awaiting a review by Admiral Yamamoto. The USN had learnt of this visit before the Japanese at Kahili, and with the USAAF achieved a masterpiece of intelligence gathering, planning and, literally, execution. Eighteeen P-38s of the 339th Fighter Squadron, USAAF, met Yamamoto’s ‘Betty’ and its escort a few miles short of Kahili and the chosen marksman, Captain Thomas Lanphier, destroyed not only the C-in-C’s aircraft, but also a second ‘Betty’ carrying members of his staff. The loss to the IJN was incalculable, for Admiral Yamamoto had been the only flag officer with experience and expertise enough to direct a holding campaign against the growing strength of the USN. Admiral M. Koga, his successor, was a competent commander, but he had had virtually no sea experience during the war so far, and as yet he was unable to command the respect which had been earned by Yamamoto. Morale was sagging, particularly in the Rabaul area, and his death occurred at an inopportune moment, with the Allies preparing for their next amphibious leap.


The use of CarDiv 1’s aircraft from shore bases had been moderately successful. The losses, some 15 per cent of the aircraft sent to the Solomons, had not been excessive in view of the results claimed. A precedent had been set for the use of carrier groups in land-based operations. The premises upon which subsequent deployments were based were false, and the Japanese carrier fleet’s effectiveness suffered accordingly.

New Georgia

On 17 May 1943 Admiral W. F. Halsey’s Third Fleet in the South-West Pacific was reinforced by the arrival of HMS Victorious and the battleship North Carolina. The carrier had been despatched from Pearl Harbor to join Saratoga in providing distant cover for the forthcoming invasion of New Georgia, a 160-mile leap forward from Guadalcanal, and one which was expected to be resisted bitterly by the enemy.


The Japanese had built airfields on New Georgia Island and Kolombangara Island towards the end of 1942, and IJNAF aircraft had been based there, to strike at Guadalcanal and to protect warships in the ‘Slot’. However, repeated attacks by AirSols had forced their withdrawal as a permanent force, and Munda and Vila airfields were used mainly for staging aircraft based at Rabaul, Buka and the Bougainville airfields. In April 1943, following the withdrawal of CarDiv 1’s aircraft, the IJAAF had been obliged to despatch more aircraft to the Solomons, about sixty in all, most of which were sent forward to Munda. Losses were made good to a very limited extent: replacements arrived at Truk in some numbers, but the extended flight from Truk to Rabaul led to some remarkable non-combat casualty statistics, particularly among the IJAAF aircrew. Up to April 1943 some 6 to 7 per cent of aircraft on ferry sorties failed to arrive, but thereafter the loss rate rose to about 25 per cent for IJNAF aircraft, and in June only two out of one flight of twenty-four IJAAF aircraft reached Rabaul.


At the beginning of June 1943 the IJNAF 11th Air Fleet and the IJAAF 4th Air Army possessed approximately 280 aircraft for operations, nearly 100 of which were G4Ms. On the 10th half-a-dozen of these ‘Bettys’ attacked a convoy bound for Guadalcanal to the southeast of the island. All important shipping movements to and from Guadalcanal had been accompanied by one or more of the escort carriers based on Noumea, and on this occasion Suwanee was in company. Good warning of the raid had been given by the coastwatchers, and the USMC Corsair patrols from Henderson Field shot down all six of the torpedo-bombers before they could close the CVE, their primary target. Aware of Allied interest in New Georgia, and that a major operation was forthcoming, the 11th Air Fleet made a determined attempt to destroy this convoy, first by a night torpedo attack on 10/11 June, and then at midday on 11th, as the ships were entering Lunga Roads. The night attack, by eighteen G4Ms, was evaded, the ships manouevring to force the aircraft to drop their torpedoes from unfavourable positions, and the day attack caused the loss of only one ship. Twenty-four bombers and up to seventy-two A6Ms took part in the daylight raid, and of the former only one returned to Rabaul. Further attacks during the month added nothing to the Allies’ casualty lists but cost the Japanese more aircraft than they could afford; 138 IJNAF aircraft were lost during the month, together with their crews. With a further forty IJAAF aircraft lost, and the high level of non-combat losses, further urgent reinforcements were required, over and above replacements.


For the second time the only source of reserve strength was the carrier fleet, the Third Fleet, and CarDiv 2 was selected to provide their air groups. Less well trained than those of CarDiv 1, they would still be superior to the inexperienced crews arriving as normal replacements. The move was delayed by the torpedoing of Hiyo, hit by one torpedo from the submarine Trigger during the night of 10/11 June. The attack took place fairly close to Yokosuka, a major naval dockyard, so that the problems of immediate salvage and repair were much simplified for the Japanese. Junyo and Ryuho had to embark Hiyo’s aircraft for delivery to Truk, and the short delay which was thus occasioned prevented the 150 A6Ms, D3As and B5Ns from arriving at Rabaul until the invasion of New Georgia was well under way.


For Operation Toenail, the New Georgia landings, Admiral Halsey’s naval forces included three CVEs and two fast carriers. These were to take no direct part in the assault, for with 258 fighters and fighter-bombers and 193 SBDs and TBFs AirSols had sufficient aircraft to provide fighter cover for the transports and the beach-heads, as well as close support and strikes on pre-briefed targets. In point of fact, the distance between the target area and the Russell Islands, the nearest fighter bases, was rather too great (175 miles) and imposed great strain on the pilots. A thirty-two-strong fighter patrol was programmed for daylight hours, for which ninety-six aircraft would be required out of the serviceable total of 213. Losses would reduce the pool required for strike escort and local defence tasks from Guadalcanal to New Georgia. Saratoga and Victorious possessed seventy-two F4F-4s and Martlet IVs (F4F-4Bs) between them, and there were another three dozen aboard the escort carriers. With deductions for self-defence the carriers could have provided another fifty to sixty aircraft for short periods, provided that they themselves were not attacked. The USN had no illusions of complete naval/air superiority, however, in spite of the enemy’s huge losses in June, and the carriers were held back, the fast carriers in company with three fast battleships to cover against an attempt by their Japanese opposite numbers against the invasion shipping. Sangamon, Suwanee and Chenango, with two old battleships, formed a reserve force whose main task was the protection of shipping between Espiritu Santo and Guadalcanal, but which could also fight a delaying action against a Japanese force attempting to break into the Coral Sea by the improbable eastern approach route.


The first landings on New Georgia took place at the eastern end of the island on 21 June 1943. Segi Point was unoccupied by the enemy, and work started on a fighter strip at the end of the month. By 11 July it was ready for emergency operations. No support, other than that of the destroyer-transports which had put the Marines ashore, was provided at Segi Point, and Task Group 36.3, the heavy covering force, did not sail from Noumea until 27 June.


Victorious and Saratoga had exercised together during the earlier part of the month, and, after assessing the particular tactical virtues of the Royal Navy carrier, Rear Admiral D. C. Ramsey, flying his flag in Saratoga, proposed a partial exchange of aircraft between the two ships. Victorious ability to operate the TBF-1 was marginal, in spite of recent modification at Pearl Harbor, but her fighter direction organisation was far superior to that of Saratoga. The advantages of each carrier being responsible for a single task were obvious, so all the TBFs of 832 Squadron were transferred to Saratoga, and twenty-four F4F-4s of the latter’s VF-3 joined the thirty-six Martlets of 882, 896 and 898 Squadrons aboard Victorious. The American carrier retained twelve Wildcats for strike escort and her own selfdefence. Planning, briefing and tactical co-ordination were thus much simplified, and the arrangement whereby one carrier was tasked with defence and the other with search and strike was near to the ideal. How effective the pair might have been in combat was never established, for although TG 36.3 remained at sea for 28 days during Toenail, the circumstances for its direct involvement did not arise.


Landings were made on Rendova Island and two small islands to the east of Munda on 30 June, the first assault on New Georgia Island itself being made on 2 July. Not until mid-afternoon on the first day did 11th Air Fleet aircraft press home an attack, when two dozen G4Ms attacked the transports in the Rendova anchorage, sinking the amphibious headquarters ship McCawley at the cost of all the aircraft which took part. An hour later a similar number of D3As appeared from Kahili, but these concentrated their attentions on the hulk of the HQ ship, scored no hits and lost heavily. Although A6Ms and a few IJAAF aircraft made hit-and-run attacks, these too achieved little at considerable cost.


The pattern was repeated on the next day (1 July), but after losing over sixty aircraft in the two days the Japanese were obliged to slacken their air effort until sufficient aircraft could be made serviceable for another onslaught. By the end of July the IJNAF had lost over 300 aircraft, in combat and through operational causes, during the ‘Munda campaign’.


The IJN faced another Guadalcanal-type problem. Supplies and reinforcements had to be brought from Rabaul and Bougainville in the face of strong Allied opposition, but without a carrier fleet strong enough to challenge the USN’s the scale of the earlier ‘Tokyo Express’ could not be repeated, even if the IJN had possessed sufficient destroyers for use as transports, and the heavy cruisers to cover them. Nevertheless, Japanese destroyers and landing barges did make a number of runs from Shortland Islands to New Georgia and Kolombangara, succeeding in passing undetected on the first occasion, but encountering opposition on four subsequent occasions. The Battles of Kula Gulf (6 July), Kolombangara (13 July), Vella Gulf (6/7 August) and Vella Lavella (17-18 August) in many ways repeated the pattern of battles around Savo Island during the last four months of 1942, with the Japanese destroyer torpedo’s long range inflicting casualties on opposing radar-equipped cruiser groups. The USN and RNZN, with considerable advantages in detection, communications and gunnery, lost one cruiser outright and another three badly damaged, as well as one destroyer sunk, while the Japanese lost an old light cruiser and four destroyers. Most of the troops carried were landed at or near their intended destinations.


AirSols attacked the Buin-Kahili-Shortlands assembly area for the reinforcement runs, TBFs and SBDs being supported by USAAF B-17s, B-24s and B-25s. Air support in the Munda area was provided by the naval and USMC attack aircraft, as many as 105 TBFs and SBDs taking part in a single raid. Strong fighter escorts were provided for the 130 strikes delivered during the five-week campaign, with up to 114 Corsairs, Wildcats, Lightnings and Curtiss P-40 Warhawks at a time to swamp the intercepting ‘Zeros’. Many Japanese warships and transport craft were damaged, including the heavy cruiser Kumano, and four destroyers and the seaplane carrier Nisshin were sunk by the strikes.


The CVEs covered their convoys without incident. The 11th Air Fleet was sufficiently occupied with events closer to Rabaul, and such submarines as were in the Solomons were being used as transports or to attempt the disruption of shipping around New Georgia. The covering position of TG 36.3 was 360 miles to the south of Guadalcanal (550 miles to the south of Munda at the western limit of the patrol line), but six sweeps were made to the north to within 250 miles of New Georgia. These pushes were preceded by TBF and SBD searches out to 230 miles from the carriers, and it was during one such sortie that the Task Group made its only contact with the enemy, an SBD of VB-3 sighting and attacking a G4M 60 miles to the northwest of the ships. The covering force returned to Noumea on 25 July. The Munda airfield had not yet been captured, but the beach-head was secure and it was clear that the main strength of the Japanese fleet was not going to be committed to the campaign.


The operations off New Georgia were the last extended carrier deployment during the Solomons campaign. Once the airfields at Segi Point and Munda had been developed, Rabaul was within comfortable fighter and dive-bomber range of AirSols. The US Fifth Air Force in New Guinea and Australia was already attacking in strength with its heavy and medium bombers. Japanese air and shipping losses continued to mount, and although the arrival of CarDiv 2’s 150 aircraft brought a temporary stiffening to the 11th Air Fleet in mid-August, this was counterbalanced by the complete withdrawal of the IJAAF component in the area. The occupation of Vella Lavella Island, between the New Georgia group and Bougainville, at the beginning of September provoked strong reaction from the 11th Air Fleet, as a result of which the Buin-based CarDiv 2 air groups suffered heavily and were forced to withdraw to Rabaul to re-form. During the same month Allied troops landed at Lae and Finschafen in New Guinea, and the remaining Japanese air effort had to be divided between the two fronts, with consequent heavy attrition. By early October 1943 it was clear that the last IJNAF reserves would have to be committed if the steady advance on Rabaul was to be checked, let alone stopped. CarDiv 2, Junyo and Ryuho, had been unable to train new air groups as they had been almost continuously engaged in ferrying replacement aircraft from the Home Islands to Truk since sending their own to Rabaul in early August, so CarDiv 1 was obliged to give up its trained crews for the second time in eight months. The IJN could have hardly chosen a less appropriate moment to emasculate its carrier force.