TOKKO NAVAL CRAFT: HUMAN TORPEDOES

The concept of a human-guided torpedo that could be launched underwater from a submarine was first proposed in Japan by junior officers of the midget submarine force at the Kure Navy Yard in 1942, but was rejected. One of these officers, Lt (jg) Hiroshi Kuroki, made a formal presentation to the Naval Technical Staff in Tokyo in February 1944, and under the worsening circumstances, the concept received more serious consideration.

Design work on a human torpedo began in March 1944 at Kure under the direction of Capt K. Mimizuka as the Kaiten Type 1. Kaiten was a contraction of the phrase “Kaiten igyo” or “Great Undertaking.” The Kaiten was essentially a conversion of the Type 93 Model 3 “Long Lance” torpedo with an enlarged forward section to accommodate a single pilot and a massive 3,410lb (1,550kg) warhead. In contrast to some earlier manned torpedoes, such as the Italian Chariot, the pilot sat within the pressure hull, not as a frogman riding outside the torpedo. The first Kaiten Type 1 was completed at the Kure Naval Yard in June 1944 and began trials.

The intention was to develop a weapon that could be carried into the combat zone by a parent submarine, then launched from underwater. Later in the war, plans were also underway to deploy Kaiten from amphibious landing ships and also from shore installations. The main attraction of the Kaiten over a conventional torpedo was that the range was extended, the warhead was more powerful, and the guidance was expected to be more precise. Manufacture of the Kaiten Type 1 began in August 1944 at the Kure Navy Yard, and later at the Yokosuka and Hikari yards. Total production was about 330 Kaiten consisting of about 100 of the original Type 1 in 1944 and 230 of the improved Type 1 Mod 1 in 1945.

Various improvements were incorporated into the weapon during the course of production to remedy flaws discovered during training and operations. The most serious problem was the damage caused by leaks. Typically, the parent submarines traveled the first few days on the surface, but would then submerge when entering areas where US ships were active. It was during the submerged portion of the trip that the leakage issues became most serious. Although a number of changes were made in hopes of stopping this problem, leakage remained an issue through the war, often rendering the Kaiten inoperable.

center KI-115 KO TSURAGI
The Nakajima Ki-115 Tsuragi was the army’s low-cost, low-tech alternative to the navy’s Ohka special attack aircraft. Unlike the Ohka, it used a simple aircraft configuration and so did not require a parent aircraft for launch. The prototype aircraft had a somewhat different undercarriage, and this illustration shows the production Ki-115 Ko configuration with the reinforced undercarriage. Due to lingering technical problems, no Tsuragi were issued to combat units, so no unit markings were carried.
image
image

The Kaiten were manufactured at the Kure, Yokosuka, and Hikari navy yards; here is was one of the assembly areas at Yokosuka at the end of the war, with several camouflaged Kaiten Type 1 seen on rail-trolleys. (NARA)

The Kaiten were operated by 18–20-year-old volunteers, mainly reservists and petty officers from the navy aviation training groups. The 1st Special Submarine Training Base Group was established at Otsujima on Tokuyama Bay on July 10, 1944, under the command of Rear Adm Mitsuri Nagai. The initial training was conducted using modified Shinyo (Ocean Quake) crash boats configured so that the pilot had to use a periscope and magnetic compass for steering. After this training was complete, the trainees transitioned to modified Kaiten, which had an exercise warhead substituted for the real one.

The parent submarines were modified for the deployment role with a special watertight tunnel through the deck into the belly hatch on the Kaiten, so that the pilot could be transferred into the Kaiten during the underwater approach to the target. In practice, some of the initial conversions lacked this feature, so the submarine had to surface to transfer the pilot aboard. The Kaiten could operate to depths of about 200ft (60m), and during the terminal approach the pilot usually leveled off around 16ft (5m) in order to use the periscope. It had a maximum underwater speed of about 30 knots (34.5mph/55.5km/h).

The Kaiten Type 1 was not entirely satisfactory due to its elementary design, and the IJN decided to start design of a much more sophisticated version, the Type 2, through a collaborative effort between the IJN’s Kure Torpedo Department and Mitsubishi’s Nagasaki-Heiki facility. Its revolutionary Type 6 powerplant was influenced by German hydrogen-peroxide torpedo engines, and a variety of configurations were developed and tested. A pair of Type 2 were completed in December 1944, but the advanced propulsion system never proved practical and the project was abandoned. Instead, it was reconfigured as the Kaiten Type 4 in January–March 1945, but using a more conventional kerosene/oxygen torpedo engine. Although the expected top speed of the Kaiten 4 was 40 knots (46mph/74km/h), trials indicated it could barely make 20 knots (23mph/37km/h), even though it was lighter than the Kaiten Type 2. A total of five Kaiten Type 4 were built before the program was abandoned. The Kaiten Type 10 was a reversion to the original Kaiten Type 1 idea, but based on the smaller Type 92 electric torpedo. It was a far simpler conversion than the Type 1, but also considerably more limited in performance, with only a 660lb (300kg) warhead. It had not entered production before the end of the war and only about six were completed.

Kaiten Operations

The Kaiten operations were codenamed Gen and were scheduled to begin in September 1944 against US Navy anchorages because these contained a large volume of relatively static targets. Due to delays in manufacturing and training, the first Gen operation did not leave Otsujima until November 8, 1944, consisting of three submarines each with four Kaiten. Submarine I-37 was sunk by US warships on November 19, 1944, before launching its Kaiten, but I-36 and I-47 managed to release all eight of their Kaiten on November 20, 1944, outside Ulithi Atoll. Many of the Kaiten were spotted and sunk, but one struck the fleet oiler Mississinewa, which led to a massive fire due to its cargo of aviation fuel. The parent submarines spotted the enormous column of smoke over the harbor and assumed that several ships had been sunk.

The exaggerated assessment of the first raid encouraged the dispatch of a second Kaiten operation, which departed the Inland Sea on New Year’s Day and included I-36 and I-48 heading to Ulithi, I-47 to Hollandia, I-53 to Palau, I-56 to the Admiralties, and I-58 to Guam, with all attacks scheduled for “X Day” – January 12, 1945. I-48 was sunk before it reached its destination, but the other submarines launched their attacks and reported large numbers of explosions and smoke columns. In fact, no Kaiten had succeeded, but Tokyo assessed the results as 18 ships destroyed, including carriers, battleships, and transports. The exaggerated assessments of these two missions led to plans to shift most of IJN submarine operations to Kaiten missions over the next few months.

image

An enormous plume of smoke marks the sinking of the oiler USS Mississinewa by a Kaiten on November 20, 1944, at Ulithi Atoll, as seen from the bridge of battleship USS North Carolina. The apparent success of this attack encouraged a costly but largely futile Kaiten campaign over the next few months. (NARA)

image

The crew wave from on top of their Kaiten submarines as they depart on the Kikusui 3 mission on February 21, 1945 from Hikari harbor. The I-370 was a Type D-1 Tei-Gata class transport submarine converted to the Kaiten role in January 1945. It was lost on this mission when depth-charged by the destroyer USS Finnegan near Iwo Jima on February 26, 1945. As is typical of a submarine on a Kaiten mission, it has the Kikusui marking painted on the conning tower.

The next Gen operation – off Iwo Jima in February 1945 – was hastily improvised and led to the loss of several of the participating submarines and a cancellation of the operation on March 6. Another Gen mission was dispatched in March during the Okinawa campaign using four submarines, but once again it was ineffective due to vigorous US Navy anti-submarine tactics. This failure led to a strenuous debate within the IJN over whether the primary targets of the Kaiten should be ships at anchorage or ships on the open sea; subsequent missions began to shift towards the latter. An operation into the Philippine Sea in July 1945 scored the last Kaiten victory, when the destroyer escort USS Underhill was blown in half by a Kaiten launched by I-53. In total, the IJN’s 6th Fleet conducted eight major and two minor Kaiten operations, losing four fleet submarines and four transport submarines. A total of 147 Kaiten were dispatched on these missions, of which 79 were launched in combat. They sank two US ships and possibly hit or damaged five or six more.

center KAITEN TYPE 1 SUBMARINE
The Kaiten shown here is in training markings with the upper portion of the hull painted in white. It also carries the Kikusui marking insignia adopted by the Kaiten Corps. The combat Kaiten seldom carried markings and were painted in black for camouflage. The training Kaiten had an exercise warhead, fitted in place of the normal warhead and of the same shape and weight.
  Key
  1.    Inertial pistol for warhead detonation
  2.    Main warhead
  3.    Electric detonator
  4.    Air chamber tanks for steering
  5.    Type 93 torpedo oxygen vessel
  6.    Forward trim tank
  7.    Lower hatch for connection to parent submarine
  8.    Electric gyroscope for steering
  9.    Rear trim tank
  10.  Type 93 Model 1 torpedo propulsion system
  11.  Periscope
  12.  Upper hatch
  13.  Pilot
  14.  Steering rudder
image
image

One of the last submarine Tokko attacks underway in the final days of the war was Operation Arashi (Storm), an attempt to launch suicide aircraft against the US fleet anchorage at Ulithi Atoll in August 1945. The large submarines I-400 and I-401 were originally intended for attacks on the Panama Canal using their Seiran aircraft, but the assignment was changed on June 25, 1945. The war ended before the operation was carried out, and here the submarines are seen tied up alongside the USS Proteus in Yokosuka harbor on September 7, 1945, with their capacious aircraft hanger doors open in front of the conning tower. I-14, in the foreground, was supposed to deliver scout planes to Truk in advance of the main attack. (NARA)

Kaiten Missions, 1944–45
Mission Date Submarines Kaiten group Kaiten embarked Kaiten launched
Gen 1 Nov 11, 1944 I-36, I-37, I-47 Kikusui 12 8
Gen 2 Dec 19, 1944 I-36, I-47, I-48, I-53, I-57, I-58 Kongo 24 19
Gen 3 Feb 20, 1945 I-44, I-368, I-370 Chihaya 14 10
Gen 4 Mar 1–2, 1945 I-36, I-58 Kamitake 8 0
Gen 5 Mar 29–31, 1945 I-44, I-47, I-56, I-58 Tatara 20 10
Gen 6 Apr 20, 1945 I-36, I-47 Amatake/Tembu 12 8
  May 5, 1945 I-367   5 2
Gen 7 May 24–Jun 15, 1945 I-36, I-165, I-361, I-363 Todoroki 18 10
Gen 8 Jul 14–Aug 8, 1945 I-47, I-53, I-58, I-363, I-366, I-367 Tamon 32 12
  Aug 16, 1945 I-159   2 0

Midget Submarines

The IJN had used midget submarines since the beginning of the war, for example in the Pearly Harbor attack. Two types of midget coastal defense submarines were allotted to the Tokko mission in 1944. The Koryu and Kairyu were both designed to carry two torpedoes: the Koryu carried them stacked one above the other in bow launch tubes, while the Kairyu had the torpedoes externally mounted on the lower hull casing. Although they could fire their torpedoes and escape, they were expected to use ramming tactics to ensure the destruction of the target ship. The Koryu was considered the most effective Tokko submarine, with an anticipated exchange ratio of two ships sunk for every three Koryus, compared to only 1:3 for Kaitens and Kairyus. A total of about 110 Kairyus and 250 Koryus were built by the end of the war, but they were reserved for the final Operation Ketsu-Go and so did not see combat use.

image

One of the more common Tokko submarines was the Koryu Type D; there were plans to complete 570 by the end of September 1945 at the Yokosuka navy yard. It had a crew of five and was armed with two 45cm torpedoes stacked one atop each other in the bow. (NARA)

IJN Tokko Naval Weapon Production in 1945
  April May June April–June (sub-total) April–June (planned) July–September 1945 (planned) October 1945–March 1946 (planned)
Small submarines 0 2 2 4 6 14 20
Koryu 12 12 20 44 110 430 1,000
Kairyu 9 42 74 125 300 550 1,000
Kaiten 66 42 51 159 250    
Shinyo Type 1 68 316 284 668 1,200 1,350 1,000
Shinyo Type 5 181 167 111 459 450 900 200

Crash Boats

As in the case of the human torpedoes, the development of crash boats began in the spring of 1944 with separate army and navy programs. The IJN program was overseen by the 4th Section (Ship Construction) of the Navy Technical Department in Tokyo, which assigned it to Capt S. Makino at the Yokosuka navy yard. The design was a scaled-down version of existing 59ft (18m) motor torpedo boats. The intention was to develop a very simple design that could be produced in large numbers using automobile engines. Two configurations were built in prototype form, six steel prototypes at Yokosuka, and two wooden prototypes at the nearby Tsurumi yard. The first Shinyo boats were completed on May 27, 1944, and put through trials. The steel hull had strength issues and, furthermore, the wooden type was far easier to manufacture so it was selected for production. The original Shinyo Type 1 used a transmission between the engine and the propeller that was costly and in short supply. This led to its redesign as the Shinyo Type 1 Mod 1, which dropped the transmission and reoriented the engine to drive the propeller shaft directly. The Mod 1 boat was smaller than the original 19ft 8in (6m) design – only 16ft 9in (5.1m) – and it became the predominant production type in 1944.

The Shinyo Type 1 Mod 1 was powered by a Toyota KC 6 cylinder automobile engine and had a speed of 23 knots (26.5mph/42.6km/h). Its principal weapon was a 595lb (270kg) charge of Type 98 high explosive in the bow, which could be detonated by an impact fuze or a manual switch. Production of these boats was undertaken at several navy shipyards, private yards, and also at a number of automobile plants. As a result, there were many small design differences between the boats built at different factories. Initial combat deployments led to a number of suggestions for improvements. One of the first changes was to incorporate a simple rocket launcher on the stern, one that could fire the standard naval 12cm shrapnel-incendiary rocket. The launcher had several roles: partly to distract US Navy gunners, partly to give the Shinyo pilot a bit of courage during the approach when under fire, and finally to cause some harm to the target ship before impact. The boats were fitted with either one or two launchers. After initial combat employment in the Philippines in the autumn of 1944, recommendations were made to include a net cutter on the front of the boat to handle protective booms. This system usually included a cable cutter on the bow, another cutter in front of the cockpit, and a deflector in front of the propeller to prevent the blade from becoming entangled in lines. These cutters varied in design from yard to yard and were added from the spring of 1945. The IJN planned to build about 7,000 Shinyo boats by September 1945, and some 6,200 were completed before the end of the war, not only in Japan, but in Singapore, China, and the Dutch East Indies.

center KAITEN PATROL OFF IWO JIMA, JULY 1945
I-58 was a B-3 class submarine launched in June 1943 and attached to the Kure Naval District. In September 1944, it was modified to carry four Kaiten. Commanded by LtCdr Matsushiro Hashimoto, it first saw combat as part of the Kongo (Diamond) Group against Apra harbor on Guam. All four Kaiten were launched on January 12, 1945, but one exploded prematurely; two pillars of smoke were spotted and I-58 was erroneously credited with an escort carrier and an oiler, though no US ships were in fact sunk by Kaiten. In March 1945, I-58 was committed to the Iwo Jima campaign, but diverted to take part in Operation Tan No. 2, the air attack on the Ulithi Atoll, with the submarine serving as a radio beacon. After returning to Japan for more Kaiten training, I-58 departed Kure in late March on Operation Gen 5 as part of the Tatara Group to attack US shipping off Okinawa. I-58 was unsuccessful in its missions there, as it was repeatedly forced to submerge due to the presence of US aircraft. After returning to Kure, the submarine was heavily reconfigured by removing its aircraft hanger to permit it to carry six Kaiten as well as fitting a snorkel to permit undersea cruising to avoid US aircraft; the illustration here shows this final configuration.
  I-58 set out on its most eventful Kaiten mission on July 16, 1945, as part of the Tamon Group with five other submarines, with plans to attack US shipping off the Philippines. On July 28, two Kaiten were launched, one against the armed merchantman SS Wild Hunter and the other against its escort, the USS Lowry. The Lowry engaged one Kaiten with its deck gun and rammed the other. On July 29, 1945, while north of Palau, I-58 picked up a large warship that it misidentified as a battleship. It was in fact the cruiser USS Indianapolis on its way back from the Marianas after delivering the nuclear materiel for the Hiroshima and Nagasaki atomic bombs. Hashimoto decided that the Kaitens would not be able to acquire the ship in the dark and so attacked it with conventional torpedoes. The Indianapolis was struck and sank shortly after midnight. Due to the secrecy of its mission, the loss of the Indianapolis was not realized for days and many of its surviving crewman died in the water from shark attack. Hashimoto continued his mission and on August 9, 1945, attacked a hunter-killer team of Task Group 75.19 led by CVE-96 USS Salamaua. The submarine attempted to launch three Kaiten, but two proved faulty and only two were dispatched. US sailors initially mistook the Kaiten for whales, but when identified one was sunk by gunfire and the other by depth charges. On August 12, one Kaiten was launched from a range of about 9,850yds (9,000m) against USS Oak Hill (LSD-7), but the Kaiten was spotted by the dock landing ship’s escort, the USS T.F. Nickels (DE-587), which rammed it. The Kaiten escaped, but surfaced shortly afterwards and the pilot, Hayashi Toshiaki, set off the warhead. I-58 returned to Kure and was surrendered at the end of the war.
  The precise markings and color of I-58 on its July mission remain obscure. Many Kaiten submarines carried the Kikusui emblem on their conning towers. Other markings also included a Japanese flag and the ship number, but these were often painted on canvas banners that could be attached in home waters and then removed during the patrol. Japanese submarines were initially finished in the usual IJN dark gray, but by the later stage of the war, were usually painted overall black with red-lead antifouling paint on the lower hull. A band of black anti-sonar paint was painted 3–5ft (0.9–1.5m) wide near the waterline.
image
image

The Kairyu was a small, two-man Tokko submarine that could carry either two torpedoes externally or be fitted with an explosive contact charge in the bow. It was inexpensively made, being powered by an Isuzu automobile engine. These examples are seen in Yokosuka harbor, where 207 were built. (NARA)

A number of more elaborate Shinyo boat designs were undertaken in Yokosuka, the focus being on developing faster and more survivable boats. These included hydrofoil designs and even rocket-powered boats. None of these designs proved practical, however. The only other major type built during the war was the Shinyo Type 5, which was intended to serve as a detachment leader’s boat and so it was fitted with two engines and carried a 13mm heavy machine gun. The idea was that this boat would lead the attack, and provide covering fire for the rest of the detachment. The first of these was completed in August 1944 and went into production in the autumn.

image

This Shinyo Type 1 Mod 4 was part of the Kawatana Assault Group in Ogushi harbor near Sasebo on Kyushu. It is fitted with a single 12cm rocket launcher on the starboard stern and has a cable cutter on the bow. (NARA)

image

The Type 5 was a detachment leader boat and was larger than the Type 1. It could be fitted with a 13mm machine gun in front of the cabin, but that is absent on this boat, which had been disarmed prior to the US occupation. This boat was part of the Kawatana Assault Group in Ogushi harbor. (NARA)

The Army Renraku-tei Boats

The IJA had a parallel boat program that was a local initiative by LtGen Yoshiaburo Suzuki from the Shipping Engineers at Ujina, a branch that was responsible for army logistics in the Pacific. In March 1944, he proposed arming small boats with explosives and other weapons to conduct fast raids against invading American amphibious forces. There was still some ambivalence about suicide tactics when the proposal was first made, so the initial idea was to build a boat that could carry a pair of 264lb (120kg) Model 98 depth charges on the stern, which could be rolled off near American landing craft. They would also have an explosive charge that could be detonated on impact or with a time fuze so that the pilot could escape. The boat was designed by the army’s research institute at Himeji near Kobe and tested in Tokyo harbor in June 1944. The crash boats were given the cover name of Renraku-tei (Liaison Boats) or Maru-ni (Large Boats), while the associated sea raiding regiments used the cover name of Akatsuki (Dawn) Regiments. The Renraku-tei boats were similar in overall design to the IJN Shinyo boats and were powered by 85hp (63kW) automobile engines. Boat construction, along with the formation of the first sea-raiding regiments, began in September 1944. The crews were recruited from young16–17-year-old officer cadets who were posthumously advanced to the rank of lieutenant after their mission. Each regiment included three boats companies and a total of 100 boats, including the reserve. By the time these boats were deployed in early 1945, the inhibitions over suicide tactics had evaporated and usually they were used as crash boats like the navy Shinyo vessels.

image

This comparison photo from the stern shows the difference between the Type 1 (left) and Type 5 (right) Shinyo boats. The Type 1 had a single pilot while the Type 5 had a crew of two with a second sailor manning the 13mm machine gun. (NARA)

image

The army Renraku-tei boats can be distinguished from the navy Shinyo by the claw-shaped rack on the stern on either corner, which carried the boat’s depth charge. This particular boat was captured on Aka Shima during the Okinawa campaign. (NARA)

Initial Boat Combat

The first IJN Shinyo units were raised at Sasebo in the summer of 1944 in squadron strength, with each unit consisting of about 200 men and 48 boats. Shinyo Squadrons 1 through 5 were sent to Chichijima and Hahajima in the Bonin Islands, while Shinyo Squadrons 6 to 13 were sent to the Philippines and headquartered at Corregidor. The navy units in the Philippines were reinforced by army Akatsuki regiments with 800 Renraku-tei boats deployed around the Lingayen Gulf, Manila Bay, Batangas, and Lamon Bay on Luzon. Numerous boats were lost during the transfers from Japan and through initial US naval bombardments of key ports and bases. The first attempted attack by Shinyo Squadron 9 ended in disaster when a fire on one of the boats spread among the craft while still in harbor, leading to the detonation of a few warheads, which wiped out the unit. The army was the next to try, with an attack by Akatsuki Regiments 11 and 12 in the Lingayen Gulf in the pre-dawn hours of January 10, 1945. One or more of the army boats executed a depth charge attack on the Landing Ship, Tank LST-925, severely damaging one engine. Four LST and eight transport ships were rammed and damaged by crash boats, but the only vessels sunk were two Landing Craft, Infantry (LCI) vessels modified as gunboats. The Akatsuki units expended 45 boats in the attack, and believed that they had sunk or seriously damaged 20–30 US ships.

The US Navy responded by deploying the “Cactus Navy,” a flotilla of gun-armed PT boats that conducted raids against known Japanese bases, as well as the nightly “Flycatcher” patrol tasked with intercepting the crash boats. The crash boat bases were also struck by aircraft, destroying or damaging many of the craft. US Navy tactics indeed prevented any mass attacks, but sporadic operations continued. Sub-chaser PC-1129 tried attacking a crash boat flotilla around midnight on January 31, 1945, but was rammed and sunk. By mid February 1945, only about 60 of the original1,200 crash boats were still operational, and on the night of February 15, the surviving IJN units on Corregidor staged a final attack after dark against a barrier of gun-armed Landing Craft, Support (LCS) vessels at the mouth of the Mariveles anchorage, sinking three craft and damaging another.

center SHINYO TYPE 1 MOD 4 CRASH BOAT
This version of the Shinyo crash boat, the Type 1 Mod 4, entered widespread production in the spring of 1945. By this stage, it has the full complement of improvements, such as the 12cm rocket launchers and the cable cutters. These crash boats received various paint schemes, a dark green with red-lead anti-fouling paint underneath being one of the more common schemes.
image
image

A US infantryman inspects one of the Renraku-tei crash boats discovered near the shore on Okinawa during the fighting there in the spring of 1945. (NARA)

Two of the IJN Shinyo squadrons in the Bonin Islands were moved to Iwo Jima prior to the US attack on the island, but were overwhelmed by preparatory naval gunfire. Seven IJA Akatsuki regiments and more than 700 boats were also deployed to the Okinawa area in early 1945 on Kerama Retto and the neighboring islands under the command of the 5th Sea Raiding Base Headquarters. The 1st, 2nd, and 3rd Sea Raiding Regiments on the islands were mostly overrun by US forces, which took the Kerama chain in late March 1945 as a preliminary step to the main assault on Okinawa. A few boats were launched with no effect, and most of the 250 vessels there were scuttled or burned by their units.

On Okinawa itself, the IJN deployed the 22nd and 42nd Shinyo Squadrons in addition to the four IJA Akatsuki regiments. The IJN boats attempted a sortie on the evening of March 29, 1945, with no effect, but a March 31 attack by about 50 Renraku-tei boats sank the LSM-12 landing ship. An attack by four surviving IJN boats of Shinyo Squadron 22 on the night of April 3 also sank the gunboat LCI(G)-82.

During the first week of fighting around Okinawa, LCI gunboats claimed to have sunk 71 Japanese explosive motor boats (EMBs). Actions around the port of Naha were so frequent that the US Navy kept the harbor illuminated at night with star shells for most of the month to deter EMB attacks. There were sporadic assaults by the army’s 27th, 28th, and 29th Akatsuki Regiments in April, seriously damaging the destroyer USS Hutchins and an LCI gunboat on April 27, and damaging several other ships and craft on other occasions. By early May, the attrition of the boats had been so severe that there were only infrequent scattered attacks, and attempts to reinforce the Okinawa garrison with boat detachments from Taiwan were frustrated when their transport ships were bombed.

Tokko Craft in the Final Defense of Japan

In the wake of the Philippines and Okinawa campaigns, control of both the army Renraku-tei and Shinyo crash boats came under IJN control. The IJN decided to unify the coastal defense Tokko units into assault groups that contained a mixed assortment of midget submarines, manned torpedoes, and crash boats. These were concentrated in the Kyushu and Tokyo areas, where the main US amphibious landings were expected.

IJN Tokkotai Assault Groups in Western Japan, July 1945
Assault Groups Location Shinyo Renraku-tei Kaiten Koryu/Kairyu
  Kyushu        
32 AG Kagoshima Bay 500 200   12
Kawatana AG Nagasaki 200 100    
31 AG Goto Archipelago 50   4  
34 AG Fukuoka 25 100    
24 AG Saiki       12
35 AG Hyuga Sea 125 100 2  
33 AG Shibushi Bay 100 200 34 12
101/102 AG Bungo Strait       18
  Shikoku        
21 AG Sukumo Bay 50   8 12
23 AG Tosa Bay 175   24  
  Honshu        
36 AG Tsushima Strait       2
22 AG Kii Strait 24   4 50
Sub-total (deployed)   1,249 700 76 118
Total (inc. reserve)   2,150 700 119 125

During the Ketsu-Go operation, the IJN Tokko weapons were intended to provide layered defense. The Koryu submarines were expected to carry out the attacks furthest from shore during the initial approach of the US fleet, and these would be followed by actions of the shorter-ranged Kairyu submarines. By this stage in the war, there was little confidence in the Kaiten submarines, based on past performance, and they were intended for use primarily from shore bases against US ships engaging in shore bombardment or against US transports. The Shinyo crash boats were to be committed against transport ships anchored off shore and, again in view of recent experience, they were supposed to be reserved for operations at night or in obscured weather conditions, due to their vulnerability. An IJN report indicated that “The principal operational mission of these forces will be to shatter the enemy invaders at sea prior to landing.” Navy officers expected an attack by about 2,000 transport ships and anticipated an attrition of about a quarter of this force, 470 ships, by the Tokko units, a tally consisting of about 210 transports likely to be sunk by kamikaze aircraft, and 260 by submarines and crash boats. The effectiveness of the Tokko vessels was expected to be about 2:3 for the Koryu, 1:3 for the Kaiten and Kairyu, and 1:10 for the Shinyo.