7. Battle of the Komandorski Islands

7-1 The situation before the battle

Due to the unexpected appearance of the US fleet near Attu Island on February 19 and 20, 1943, the rest of the transportation of the "A-Gou" Operation (the Japanese transport operation for Attu), which was planned during February, was stopped. For the Japanese navy, it was shocking that a transport ship "Akagane Maru" was sunk by the US fleet bombardment at night rather than an air attack in midday. The Fifth Fleet, which charged the sea defense around the northern Japan, was forced to completely rethink the transportation operations. As a result, the Fifth Fleet planned aggregate transports to Attu Island with strong escorts. The first aggregate transport was carried out by escorting the seaplane tender "Kimikawa Maru" and the transport vessels "Awata Maru" and "Sakito Maru" with two heavy cruisers, three light cruisers, and two destroyers of the Fifth Fleet. They arrived at Attu Island safely on March 10, successfully transporting six A6M2-N float fighters (Rufe), three reconnaissance floatplanes, the airfield construction materials, and 342 personnel [4, p442]. However, this transport was discovered by an US submarine off Holtz Bay [3, p250].

Subsequently, the second aggregate transport was planned. The main transport fleet consists of the special cruiser "Asaka Maru" which was reformed from the cargo vessel, and the transport ship "Sakito Maru". They were escorted by Main Body of the Fifth Fleet led by the Commander-in-Chief Lieut. Gen. Boshiro Hosogaya; the heavy cruisers "Nachi" and "Maya", the light cruiser "Tama", and the destroyers "Wakaba" and "Hatsushimo", and by the First Torpedo Attack Squadron belonging to the Fifth Fleet from May 1942 (hereafter Torpedo Squadron) led by Maj. Gen. Tomokazu Mori; the light cruiser "Abukuma", the destroyers "Inadzuma" and "Ikadzuchi". In addition, the second transport fleet consisted of the transport ships "Sanko Maru" and the destroyer "Usugumo" [4, p472]. Among these, the transport ships for Attu Island loaded the commander of Second District Unit (Attu Island Garrison), Colonel Yasuyo Yamazaki, the District Unit headquarters, artillery battalion, anti-aircraft battalion headquarters, anti-aircraft company, and part of the field hospital, and 550 personnel as well as artillery, food, and airfield construction materials [19, p197]. The second transport fleet, "Sanko Maru" and "Usugumo", departed Paramushir first at 1600 on March 22 for Attu Island, and the rest of the fleets departed Paramushir separately on March 23 [4, p475].

However, due to the influence of a severe low pressure developed like a typhoon from the early morning on March 24, an easterly wind of 20 m/s (45 mph) blew over the sea near the transport fleet at around 0600. Due to the violent waves and swells, the Commander-in-Chief of the Fifth Fleet postponed the arrival into March 27 at 1515 on March 25 [4, p475]. However, the decision seems to be a little too early. The weather began to recover from the evening of March 25, and it was informed by telegram at 1800 from Attu Island that it would be possible to unload on March 26 [4, p475]. The transport fleet, Main Body, and Torpedo Squadron joined on March 26 after the waves subsided, but the second transport fleet "Sanko Maru" and "Usugumo" that departed earlier could not join them.

Meanwhile, the US Intelligence supposed that the Fifth Fleet would carry out the second transport operation. The US North Pacific Fleet ordered to the Task Force 16.6 led by Rear Admiral McMorris to block it, which consisted of the flagship light cruiser "Richmond", the heavy cruiser "Salt Lake City", and the destroyers "Bailey", "Coghlan", "Dale", and "Monaghan". However, the US intelligence was unaware that the Fifth Fleet had been augmented with the heavy cruiser "Maya" [2, p54]. Following a transport blockage order, the Task Force 16.6 was sailing in the waters south of the Komandorski Islands, far west of Attu Island, before dawn on March 27. On this day, the Japanese aggregate transport encountered the US Task Force, and the Battle of the Komandorski Islands took place. The weather at this time was cloudy, with a slight swell in the southeastern wind of about 4 m/s (9 mph). [10, p64] Visibility was good, but clouds covered at an altitude of about 2000 feet [2, p55].


7-2 Progress of the battle

7.2.1 Situation at the beginning of the battle

With the exception of the transport ship "Sanko Maru" and the destroyer "Usugumo", which could not be joined due to the storm, the Japanese fleet was led by the flagship heavy cruiser "Nachi", followed by Main Body and Torpedo Squadron in column heading north about 300 km west of Attu Island. The transport vessels "Asaka Maru" and "Sakito Maru" were sailing between two destroyers in Torpedo Squadron [4, p477]. Around 0200 before dawn (the sunrise is 0320) on March 27, the light cruiser "Abukuma", the flagship of Torpedo Squadron, received a report of "finding enemy ships" from "Inadzuma" steaming at the southernmost tip. "Abukuma" thought of this as the second transport fleet of "Sanko Maru" and "Usugumo" [4, p477]. "Abukuma" himself found two ships in the southwest direction at 0237, but the commander continued to consider this as the second transport fleet [4, p477]. The Command of Fifth Fleet in "Nachi" did not take any action, believing that the two ships were "Sanko Maru" and "Usugumo", which had been joining to Main Body of the fleet. [4, p480] But, in reality, the second transport fleet, departed 18 hours earlier, was on far northwest of the fleet. The Japanese fleet never thought that the US fleet had advanced so far to the west. [4, p481] At this time, the Japanese fleet had just finished the early morning training, and the crew had just been relieved after the long stormy weather [4, p482].

On the other hand, the US fleet initially thought that the target about 15 km north of the SC radar of the destroyer "Coghlan" at 0230 was Japanese destroyers and four transport ships, and they never thought that the opponents were much stronger than themselves [8, p28-34]. The number of destroyers was the same as four, but the number of heavy and light cruisers in the Japanese fleet was twice of the US fleet. The crew of US fleet took up their positions and accelerated ships' speed. Around 0320, the light cruiser "Richmond" found that there were nine vessels in the Japanese fleet including one heavy cruiser and two other cruisers, but the whole picture of the Japanese fleet, which was separated into several groups, was hard to grasp [8, p34]

In the Japanese fleet, at 0313 when the visibility became high at dawn, "Abukuma" reported that "one light cruiser and two destroyers were found", and at this point the Japanese fleet noticed the ships enemy [4, p479]. It seems that the crew's response to the command of "set to the battle arrangement" was slow wondering it was training again or not. The command of Japanese fleet immediately decided to occupy the northeastern side of the US fleet to cut off their retreat way, and to evacuate the transport ships to the northwest [4, p479]. After that, the Japanese fleet changed its course clockwise to the south-southeast and challenged the US fleet in the south. At that time, two destroyers were put into Torpedo Squadron, so Main Body consisted of two heavy cruisers and one light cruiser, and Torpedo Squadron consisted of one light cruiser and four destroyers. Torpedo Squadron was a little late to join the two destroyers, and was on the east side of Main Body, the opposite side to the enemy. Moreover, each ship of Torpedo Squadron was slowing down the boiler combustion to save fuel oil, and it was not possible to increase the speed until around 0400.

Light cruiser "Abukuma", the flag ship of the First Torpedo Attack Squadron 
https://upload.wikimedia.org/wikipedia/commons/9/90/Abukuma_cl1941.jpg

On the other hand, from the view of US fleet, the transport ships were on the north-northwest side, and the Japanese fleet was on the northeast side. 16.6 Task Force Commander McMorris saw the two destroyers joining Torpedo Squadron and thought the Japanese had removed the escort from the transport ships. He viewed the possibility of attacking the transport ships before entering the attack range of the Japanese fleet. The attack on transport ships might make the Japanese fleet to strengthen the transport ships escorts, which could balance the both forces [8, p35]. At this point, the US fleet would still have been possible to evacuate to the east, but McMorris chose to fight. The US fleet shifted its course to north-northwest (330°T) at 0333, advancing northward to attack the transport ships on the west side of the Japanese fleet moving southward [8, p34]. The two forces quickly approached each other, and the heavy cruisers "Nachi" and "Maya" began firing at 0340, targeting the flagship "Richmond," which was at the forefront of the US fleet. Almost at the same time, the US fleet also began firing. The distance to "Nachi" at that time was about 18 km on the SG radar of "Richmond" [8, p36].

At 0350 at the beginning of the war, a shell from the US fleet hit the rear starboard side of the "Nachi" bridge. As a result, 32 people were killed or injured, and the communication device and the electric circuit board were destroyed. In addition, at 0352, another shell hit the torpedo tube room, killing two people and injuring five [4, p482]. Regarding these hits, "Salt Lake City" recorded a hit of a Japanese heavy cruiser by the third or fourth salvo at a distance of 19 km at around 0342, which caused a momentary flame near the bridge [8, p36]. However, Lieutenant Colonel Takeo Inoue, Chief Gunner of "Nachi", said, "Immediately after the first shot (of Nachi), I admitted the gun fires of a light cruiser with four chimneys. His second shot hit the bridge and the communication device was damaged (it might be by a shell from a destroyer)." [4, p484]. However, the Japanese War History Series "Naval Operations in the Northeastern" states that "the shell must be from a destroyer in view of the damage" [4, p486]. The document also states that the range of US destroyers was about 18 km, which was longer than that of Japanese destroyers [4, p509]. Certainly, there is a recollection that the shooting range of the Japanese destroyer "Ikadzuchi" at that time was about 13 km [20, p302].

A crew of heavy cruiser "Nachi", who was surprised at an abrupt meeting up with enemy, turned off the generator by mistake, and the turret power was temporarily off [4, p484]. For this reason, the fire control system of the main gun did not work at the beginning of the battle, thus "Nachi" fired with inferior individual gun aiming for about 30 minutes [4, p482]. Lieutenant Morio Sahashi, Chief of Engine Department, clearly stated that the cause of this individual gun aiming was not caused by a hit, but a crew accidentally turned off the generator, which made the pointer of artillery command center deviate from that of the individual turret gun [4, p484-485]. This confusion was amplified by the damage of the communication device. In addition, for about 30 minutes at the beginning, the heavy cruiser "Maya" fired with a wrong setting ― an approaching enemy heavy cruiser as the shooting target and a distant light cruiser as the range-finding target [4, p485].


7.2.3 Course change of US fleet to the west

"Richmond" was bombarded by a Japanese heavy cruiser at a distance of about 20 km from around 0340, and the shells crossed the ship. The US fleet changed their course to west-northwest (290°T) at 0343 to increase the distance from the Japanese fleet which was rapidly approaching from the northeastern side. At 0346, "Nachi" fired eight torpedoes and "Maya" fired four torpedoes [4, p482] (There is another view that "Maya" fired them at 0407 [4, p485]). The US fleet steered further west-southwest (250°T) at 0348 [8, p36]. The reason for this is not stated in the US materials, but it might be due to avoid the Japanese torpedo attack as described later. There is another recollection that "Nachi" launched torpedoes to the east of the enemy fleet in expecting change of their course to the east [4, p480], but, if so, it doesn't match the U.S. Navy record. The Japanese War History Series concluded that it was unclear whether the torpedo launcher was on the same voyage or counter-voyage for the US fleet [4, p481]. These torpedoes are referred later.

Upon seeing the course change of US fleet, the Japanese fleet turned to the right, headed southwest, then west, and began pursuing the US fleet about 20 km away. As a result, Main Body of the Japanese fleet, which happened to be ahead, was located behind the port side of the US fleet (south rear), and the delayed Torpedo Squadron was located behind the starboard side of the fleet (north rear) [2, p57]. The US fleet fled in a zigzag avoiding artillery and torpedo attacks. The Japanese fleet, however, could not close the distance because it chased diagonally to the enemy's track to use the all turret for shooting [2, p57]. The shells from the Japanese fleet often straddled the US fleet. Each time, US ships plunged into the water column where the shell fell to avoid next shells. Lt. Commander Horn, who was watching it from the destroyer "Dale", described the Japanese fleet's shooting as "beautiful" and said it was a miracle that "Salt Lake City" was able to avoid the shells [8, p39].

The tracks during the battle for the US and Japanese fleet.
https://ww2db.com/image.php?image_id=27110


Several torpedoes passed under the bow of the light cruiser "Richmond" at 0358. Shortly after this, the destroyer "Bailey" also discovered torpedo trails 3.6 km off her starboard quarter [8, p38]. The US fleet and the torpedoes were in almost the same direction. There remains no record of torpedo setting specifications in the Japanese fleet. If "Nachi" launched torpedoes in the southwest direction aiming at the destination of the US fleet, it would not be inconsistent with the U.S. Navy's record that the torpedoes passed under the fleet because the fleet turned its direction to the west-southwest at 0348 after the Japanese fleet had launched the torpedoes. The position of "Richmond" when the torpedo passed would be estimated 17 km from the torpedo launch point if torpedoes ran at about 45 knots. Lieutenant-commander Hajime Kinoshita, Artillery Staff of "Nachi", was recollecting that "Launching torpedoes pressed the enemy's counter-voyage, forcing him to make turn to the southwest" [4, p482]. The recollection could be true because it seems to be consistent with the torpedo track and the US fleet movement. 

The U.S. Navy ''Omaha''-class light cruiser USS ''Richmond'' (CL-9) off the Puget Sound Navy Yard, Bremerton, Washington (USA) on 24 June 1944.
https://en.wikipedia.org/wiki/USS_Richmond_(CL-9)#/media/File:USS_Richmond_(CL-9)_port_side_June_1944.jpg

7.2.4 Damage of heavy cruiser "Salt Lake City" by shells from the Japanese fleet

The Japanese fleet intensively bombarded the largest heavy cruiser "Salt Lake City". Around 0410, a single shell hit the ship. It broke through the oil tank and bulkhead and exploded 1 m behind the engine room, which caused flooding of the engine room with fuel oil [8, p39]. However, at 0427, the cracks in the oil tank were fixed and the outflow of oil almost stopped [8, p40]. The ship was still able to make the speed of over 30 knots. Around 0430, the command of North Pacific Force contacted McMorris that the supporting aircraft would not be able to take off until around 0930. Kincaid, Commander-in-Chief of the North Pacific Force, recommended withdrawal [8, p40], but this suggestion "brought a smile" to McMorris' face [8, p40]. 


Rear Admiral Charles McMorris
https://commons.wikimedia.org/wiki/File:Charles_Horatio_McMorris.jpg

Around 0502, "Salt Lake City" was hit near the bridge, causing injuries in the steering house. At the same time, the rudder no longer responded to the steering wheel due to the blast of his own gun [8, p41]. The rudder control was moved to Steering Aft, but only emergency steering with a maximum rudder width of 10° was possible. "Richmond" also received salvoes, so the course was shifted to 330°T at 0502 to take a distance from the Japanese fleet [8, p41]. At 0510, another shell hit the front of the main deck of "Salt Lake City". At 0512, seeing this situation, McMorris ordered the destroyers "Bailey" and "Coghlan" to use smoke screens to conceal "Salt Lake City" [8, p41].

The US fleet shifted its course to the left (300°T) at 0509 [8, p41]. The reason for this is not described in the US Navy material [8], but "Abukuma" fired four torpedoes at 0507. Regarding this torpedo attack, Lieutenant-commander Goro Iwabuchi, Chief Torpedo Control of "Abukuma", said, "So far, any attack was not possible because of the delayed advance. Then I suggested the commander to launch torpedoes, I didn't expect much success though." [4, p490]. However, he forgot to adjust the torpedo's oblique movement (change of direction while sailing), so "Abukuma" had to change its course to the opposite direction to the enemy when the torpedoes were launched [4, p490]. The torpedoes did not hit, but the Japanese interpreted that the attack caused the US fleet to abandon the escape to the north and changed its course to the west-northwest (300°T) [4, p488]. At 0528, the US fleet shifted the course west-southwest (240°T) to escort "Salt Lake City." [8, p42]. At about the same time, the Japanese fleet also shifted its course, chasing the US fleet from the starboard tail (northeast side).

At 0559, an 8-inch shell hit the catapult on the starboard side of "Salt Lake City," and at 0603, another shell hit the port side of the stern [8, p43]. The shell explosion spread the inundation that occurred on the previous shell hit. Shortly after another shell hit again, oil began to spurt into the gyro room at the stern. Fuel oil and seawater flooded the anti-aircraft switchboard room, the after 5-inch handling room, and the after 5-inch ammunition stowage, causing the ship to tilt 4-5 degrees to port [8, p44]. At 0625, the inundation of "Salt Lake City" reached a depth of 1.5 m, which eased the list but flooded just below the main engine, rendering the rear engine room inoperable. To restore this situation, the ship shut down her rear engine. As the speed of the ship slowed down, shells began to cross the ship. At 0629, the fuel oil and water in the rear engine room were pumped out, and the rear engine gradually returned to normal operation, although the ship leaned to port again. [8, p45].


7.2.5 Change of course of the US fleet to the south at 0602

The US fleet changed its course to the south-southwest (210°T) at 0602 and further to the south (180°T) at 0608 [8, p45]. The use of smoke screens was instructed to prevent the Japanese fleet from noticing the course change. Regarding the reason for the steering to the south, the US Navy material states that it was decided to head south and then east because Japanese bombers might come [8, p43]. However, since it had been within the range of Japanese bombers from the beginning of this battle, the reason would not be reasonable. Actually, the Imperial Navy did not have deployed any aircraft in Paramushir, the nearest Japanese base. The US fleet would eventually confront the Kamchatka Peninsula toward the west, and the north side was blocked by the Komandorski Islands, and the Japanese fleet was chasing from the northeast, so there would be no choice but to run to the south.

This course change brought the Japanese fleet a great opportunity to shorten the distance. The Japanese fleet noticed their course change, then the Commander-in-Chief ordered "All forces, charge at the enemy." around 0600. The Japanese fleet, however, changed its course to the south-southeast after sailing west to the vicinity of the course change point of the US fleet. This provided the same pursuit position as before, and the Japanese fleet missed the opportunity to shorten the distance and could not give a decisive blow to the US fleet [4, p492]. 

In the Japanese War History Series, there are several recollections. Lieutenant Joji Tanaka, Chief of Combat Room of "Nachi", said, "Around 0600, I knew that the enemy had changed the direction to the south, so I repeatedly suggested starboard gun battles." Lieutenant Colonel Takeo Inoue, Chief Gunner of "Nachi", said, "There were suggestions of starboard gun battle from Chief of Combat Room, but I did not deliver them to the bridge." Lieutenant Tadayoshi Sugiyama, Chief Torpedo Control, said, "At that time, I felt that the headquarters' fights had already decreased." Lieutenant-commander Hasjime Kinoshita, Artillery Staff, said, "The delay in turning was due to Torpedo Squadron was getting in the way." However, Torpedo Squadron argued in its Detailed Battle Report, "If Main Body approaches, our squadron would have avoided it, and if we were still in the way, Main Body should have just instructed us to move east" [4, p492].

According to the order of charge from the Commander-in-Chief, "Maya" launched 4 torpedoes at 0605, "Nachi" launched 8 torpedoes, and "Abukuma" launched 4 torpedoes. After that, the Japanese fleet seemed to change its course to the south. All torpedoes did not hit. The Japanese War History Series concluded that the reason of delayed turn of the Japanese fleet seemed in order to perform the torpedo attack and to concentrate its fire to the enemy ships. It also states that there is a quesiton to have launched torpedoes at this moment [4, p492-493].

The US fleet, which once shifted to 160°T at 0621, turned to the south (180°T) again at 0629 to take the distance from the Japanese fleet [8, p45]. McMorris also instructed the destroyers at 0632 to prepare for a torpedo attack on the Japanese fleet in order to smoothly retreat "Salt Lake City" that her engine had begun to operate again. However, at 0638, "Salt Lake City" was able to make 26 knots, so the order of torpedo attack was revoked. On the other hand, the Japanese fleet, which saw the destroyers' maneuver toward the torpedo attack, temporarily sharply steered to the right (west) [8, p46].


7.2.6 Standstill of "Salt Lake City"

The heavy cruiser "Salt Lake City," which continued to bombard the Japanese fleet on the north side with its rear turret while heading south, began to run out of ammunition in the rear turret around 0630. The rear turret was an only recourse to prevent the Japanese fleet. Shells were transferred from the front ammunition depot to the rear using wheelbarrows. The powder bags were handed over by lines of crew members [8, p46]. Around 0640, the rear turret was depleted of shells for armor (AP ammunition) and began firing with high explosives (HC projectiles). The violent spray of this landing on the sea seems to have given the Japanese fleet the misunderstanding that an air attack was taking place from somewhere above the low clouds. The US Navy material states that Japanese heavy cruisers begun to intensely fire anti-aircraft guns [8, p46]. The battle record of Torpedo Squadron said "The enemy fired shrapnel at "Nachi" at the beginning and end of the battle, and the shells exploded within 200 m of "Nachi" at altitudes of 50 to 100 m" [4, p509]. This seems to refer to the "Salt Lake City" bombardment with high explosives.

In the meantime, around 0647, several shells hit the broadside below the water line of "Salt Lake City" [8, p46]. Sea water had been seeping into the fuel tanks. The fuel oil mixed with seawater supplied to the boilers extinguished them one by one. The ship began to fall on her legs, her speed was 8 knots at 0653, and stopped completely at 0655 [8, p46]. There is another theory that the cause of boiler extinction was that the seawater injected to correct the inclination of the ship was mixed in the refueling pipe to the boiler [2, p59]. "Salt Lake City" raised the signal flag "my speed zero", which was immediately torn by the Japanese's fierce shells. Bertram Rodgers, Captain of "Salt Lake City", once ordered "Abandon ship" following the suspension of her ship, but he reconsidered before it was broadcasted in the ship [2, p59].

USS Salt Lake City (CA-25) in action during the battle, with an enemy salvo landing astern on 26 March 1943.
https://www.history.navy.mil/our-collections/photography/numerical-list-of-images/nara-series/80-g/80-G-70000/80-G-73827.html

The Commander of destroyer fleet immediately suggested McMorris to attack the Japanese fleet [2, p59]. In response, McMorris ordered the destroyers "Bailey," "Coghlan," and "Monahan" at 0655 to execute a covering torpedo attack on the Japanese fleet. "Dale" was instructed to continue to put up her smoke screen near the stern (north) of "Salt Lake City" [8, p47]. The three destroyers that received the order lined up and rushed to the north toward the Japanese fleet, firing a 5-inch gun. Upon seeing this, the Japanese fleet shifted its shooting target from "Salt Lake City" to the destroyers. The destroyer "Bailey", which was closest to the Japanese fleet, was hit several 8-inch shells in quick succession at 0700.

Concerned about hitting a shell to the torpedo tube, the commander of destroyer fleet ordered immediately the torpedo fire at 0703 [2, p62]. Bailey launched five torpedoes at the second heavy cruiser, 8700 m (9500 yards) away [8, p49]. Main Body of Japanese fleet confirmed four torpedo trails behind it [4, p494]. As worried, at 0704, just a few seconds after the torpedoes were fired, "Bailey" was hit by 8-inch shell fragments". They seriously injured the executive officer, and inflicted minor wounds on two enlisted men. The debris also pierced the hull, through which seawater flowed and the starboard engine stopped [8, p49]. The other destroyers did not fire torpedoes because their targets were too far away. 

USS Bailey (DD-492)
https://en.wikipedia.org/wiki/USS_Bailey_(DD-492)#/media/File:USS_Bailey_(DD-492)_underway_on_12_December_1942_(80-G-264956).jpg

7.2.7 Reason of the Japanese fleet turning to the southwest around 0638

The Japanese fleet was delayed in turning south toward the US fleet, but from around 0620, it followed the US fleet toward the south-southeast ― the port rear (northeast side) of the US fleet. Therefore, it is believed that if the Japanese fleet continued to go straight to the south, it would block the retreat way of the US fleet and create an advantageous situation [4, p491, 499]. However, around 0638, Main Body of the Japanese fleet suddenly turned its course to the southwest, followed by Torpedo Squadron [4, p493]. This led to a rapid increase in the distance between the US fleet and the Japanese fleet.

Since the "Salt Lake City" was damaged in the US fleet, the distance between the two forces would have shortened rapidly if the Japanese fleet continued to pursue south. Then, this naval battle would have entered the next new phase. Regarding the reason why the Japanese fleet changed its course to the southwest, Chief of Torpedo Attack Management Sugiyama and Chief of Air Staff Miura of "Nachi" recalled that there was a report that the shells of the main gun ran out after 0630 [page 4: 495]. However, Nachi's artillery officials said that armor-piercing shells were gone, but there were still other shells. The Japanese War History Series has stated, "Judging from the recollections of these artillery personnel, the Command of Fifth Fleet may have misjudged the report 'no armor-piercing shells' as 'no shells'." as one of the reasons for the course turning. However, is such a judgment concerning whole fleet made only in the situation of "Nachi"? Other ships in the Japanese fleet continued to fire until 0714. 

The view of the US Navy is different. As mentioned earlier, the US destroyers were in the torpedo attack position at 0632. Eventually the order was revoked, but the US Navy believes that the Japanese fleet turned southwest after seeing the destroyers' torpedo attack posture. This reason seems more convincing. Eventually, the Japanese fleet turned further west at 0707 to avoid the torpedoes launched by the rushing US destroyer "Bailey". The Japanese fleet disappeared to the west without changing its course, and the battle ended.


7.2.8 Aftermath of the US fleet

Commander McMorris proposed to Kincaid that the damaged "Salt Lake City" and "Bailey" be sent to Adak Island and the rest of the fleet follow the transport ships to Attu Island [Page 8: 53]. However, Kincaid was concerned about whether the Japanese fleet had actually left and about attacks by Japanese aircrafts and submarines. In response, McMorris instructed his fleet to head to Dutch Harbor and Adak [8, p54]. It is amazing to see McMorris' fighting spirit proposing to advance the fleet to Attu Island to attack the transport ships after engaging the sea battle for several hours and finally escaping the predicament.


7.2.9 Result of Battle of Komandorski Islands

This sea battle was the last one in World War II in which ships alone tried to decide the game. The battle left 7 dead and 20 injured in the US fleet and 14 dead and 26 injured in the Japanese fleet [21, p64]. Even if the transport ships reached Attu Island, it was expected that they would receive a large-scale air raid in the bay. Also, the destroyers consumed a large amount of fuel in this battle. So the all vessels returned to Paramushir and the second aggregate transport operation ended unsuccessful.

The destroyer "Usugumo", which was escorting the transport ship "Sanko Maru", found the smoke rising to the south at 0355, so it evacuated "Sanko Maru" to the west and headed south. However, "Usugumo" was not in time for the battle, and was ordered to escort the transport ships and return to Paramushir at 0725.

Due to the failure of this naval battle, Boshiro Hosogaya, Commander-in-Chief of the Fifth Fleet, was removed and went into the reserve, and Lieut. Gen. Shiro Kawase was newly appointed as Commander-in-Chief of the Fifth Fleet.


7-3 Consideration for naval warfare

7.3.1 Reasons for the stop of battle by Commander-in-Chief of the Fifth Fleet

At the end of battle, Hosogaya, Commander-in-Chief of the Fifth Fleet, had thought that it would be difficult to expect decisive results even if the battle continued. He raised the following five reasons for stopping the battle [4, p496].

  1.  The enemy fleet suffered considerable damage, but we could not confirm the result of the battle due to the enemy's smoke screen.
  2. Radio intelligence revealed that enemy aircraft had left Amchitka Island.
  3. Despite consuming most of the main gun shells, we could give little damages to the enemy.
  4. Little fuel left on the destroyers, so there was a worry about the continuation of the battle.
  5. Torpedo Squadron's advance was slow and inactive.

Regarding 1 and 3, the commander of Torpedo Squadron reported to Main Body of the fleet by telephone at 0659 that "one of the enemy ships is shedding a large amount of fuel oil". The Detailed Report of Torpedo Squadron also reported a fire on an enemy cruiser around 0700, so the Japanese War History Series guesses that the command of Main Body had known the considerable damage of the US heavy cruiser. [4, p497]. 

Regarding 2, there is no such a telegram record in the Detailed Battle Report of Torpedo Squadron. Rather, in the same report, it was informed from Kiska Island at 0656 that "the enemy army planes did not appear to have taken off" and at 0705 that "there is no evidence that the enemy army planes are flying from their communication so far". [4, p498]. Regarding the enemy air attack, there was a recollection of Shigefusa Hashimoto, Chief Staff of Communication of Fifth Fleet, "I guessed that enemy airplanes would come from 7 to 9 o'clock, and wanted to go out of their range before that." Therefore, it is believed that the possibility of an air attack was taken into consideration for retreat rather than actual enemy aircraft coming [4, p498]. However, as mentioned before, the HE bombardment of the "Salt Lake City", which began around 0638, may have contributed to the fear of air strikes.

Regarding 4, the fuel of each destroyer was not in a state of concern after a few more hours of pursuit [4, p499]. Regarding 5, Torpedo Squadron reported as follows. "In the southern pursuit after 0620, the fleet only sailed to the left rear of Main Body of the fleet. In 0645, the fleet shifted the course to the west according to Main Body, resulting in the opposite side of the enemy." At the post-combat study meeting, Torpedo Squadron Senior Staff Lieutenant Colonel Rokuji Arichika was criticized by the participants [4, p498]. The study meeting concluded that the behavior of Torpedo Squadron was "inactive and poor coordination with Main Body of the fleet" [4, p499].


7.3.2 Lessons and findings of the battle in the Japanese War History Series "Imperial Navy operations in Northeastern" 

Here, the war lessons and findings of this sea battle from the Japanese War History Series "Imperial Navy operations in Northeastern" are described. In the document, the section of war lessons and findings is divided into three categories: "general matters," "gun battle-related," and "torpedo battle-related," but seven of the total 11 pages are characterized by detailed descriptions of gun battle. The "general matters" is consisted of two sources: one is the postwar memoir of Lieutenant-commander Hasjime Kinoshita, Artillery Staff of the Fifth Fleet, and the other is Detailed Battle Report of Torpedo Squadron. The latter seems to be more objective and more suggestive as the official record than the memoir, so I cite a part of it below [4, p504-505].

After the base sortie, the operational fleet should not disperse its vessels as much as possible and should advance in a concentrated or close cooperation. This should be done especially when there is a high probability to meet enemies or where the weather is prone to sudden changes, such as in the northern sea.

1. There is a need for study or research regarding a prompt contact way in the fleet when an enemy is found. Reports of enemy discovery at unexpected times or sea areas, such as this battle, often lag behind by normal communication. Especially, in the case of enemy discovery in the rear in a single column, it is necessary to study alarm-type communication (firing, telephone, special signal, etc.) in order to promptly contact the Commander-in-Chief.
2. We were keenly aware of the need for research on the tactical movements of Torpedo Squadron in the daytime pursuit battles with cruisers of Main Body.
(1) When both the enemy and the ally are at almost constant speed, there are few options for the movement of Torpedo Squadron. Thus, the optimum movement is necessary for the squadron considering both rapid approach and retreat blocking. (Omitted after)
(2) Of course, Main Body and the Torpedo Squadron should always maintain close contact, and the Torpedo Squadron should always consider the intention of Main Body when maneuvering. It is also important for Main Body to take measures such as instructing the squadron in the direction of movement accordingly. (Omitted after)
(3) ・ ・ ・ In a pursuit battle in which both the enemy and the ally consist of cruisers, an opportunity to demonstrate the combat power of Torpedo Squadron is extremely few, so it must hit even a single enemy ship by the first gun or torpedo battle to suppress their retreat actions.
3. The main point of an unexpected battle is that allies must attack earlier than enemy and decide a game at once while enemy is not ready. (Omitted after). 

Regarding the normal communication described in 1, it may refer to using a flag signal or a light signal. It turns out that there was some problem with emergency communication systems in the Japanese fleet. Regarding 2 (2), it is stated that Main Body and Torpedo Squadron needed to cooperate effectively. This is a distinct feature not found in the other parts of this document.


7.3.3 Tactics of the Japanese fleet on the battle

In this battle, the Japanese fleet succeeded in pressing the head of the US fleet and turning it west at the beginning of the battle, eliminating the immediate threat to the transport ships. After that, the pursuit of the American fleet escaping to the west began with firing. But what was the basic tactics of the Japanese fleet after that? It seems that the policy of concentrating firepower on the chasing enemy using all turrets and that of shortening the distance to the enemy are generally in conflict. Which one were they going to prioritize? Or was there another way of thinking to destroy the enemy fleet? It's not clear.

After that, the US fleet had three major course changes. When the US fleet turned north around 0443, the Japanese fleet's pursuit changed from the starboard rear (southeast direction). Next, when the US fleet changed its course to the south-southwest around 0528, the Japanese fleet tracked from the starboard rear (northeast direction). Therefore, when the US fleet turned south at around 0600, the Japanese fleet was about 20 km northeast of it. This seems to have put the US fleet in the greatest crisis, as mentioned in 7.2.5. The US fleet would continue to approach Japan mainland and move out of their air support range, so there would have been no choice but to steer east somewhere and challenge the decisive battle. On the other hand, it seems that the Japanese fleet had the option of going straight south to the east the US fleet to meet it, or letting Torpedo Squadron go south to the east side of the US fleet to pinch and shoot.

This naval battle lasted for four hours. Continuing to think for the right decisions for long time in response to ever-changing situations can be exhausting. This is inevitable for everyone. The naval battle seems to have been chasing behind an enemy fleet escaping at high speed for a long time, trying to hit shells. As a result, in the end, fatigue may have weakened human thinking ability, leading to misunderstandings and negative thinking. But that's why the Japanese commander would need a consistent basic strategy for this battle that wouldn't be affected by various changes in the situation.


As mentioned in 7.2.5, at the greatest opportunity that the US fleet steered south around 0600, the Japanese fleet missed it by keeping the course for the immediate artillery and torpedo attacks by the order of charge at the enemy. Looking at the situation of the Japanese fleet at this point, the number of ships and firepower were superior, the weather was good, the daytime was still long left, and the theater was gradually approaching Japan. The Japanese fleet didn't have to rush the fight. The simple policy of "waiting until the enemy would change its course to the east, and then approaching and destroying them with artillery and torpedo attacks" would have been sufficient. Of course, as long as it was a battle, the result would be unknown. The concerning about air strikes is addressed in 7.3.5.


7.3.4 Actions of Torpedo Squadron

At the time of the counter-voyage of the beginning of the war, Torpedo Squadron (light cruiser "Abukuma" and 4 destroyers) was sailing on the east side of Main Body (two heavy cruisers and a light cruiser) opposite the enemy. After that, it followed Main Body, but it could not make any effective attack until the end because it shifted its course to avoid enemy's shells and it moved away from the enemy to make a torpedo attack. Commander-in-Chief of the Fifth Fleet raised the slow advance and inactivity of Torpedo Squadron as one of the reasons for stopping the battle as mentioned in 7.3.1.

On the other hand, according to Detailed Battle Report of Torpedo Squadron, the reasons why the actions were restricted are as follows. Due to an unexpected encounter with the enemy and a situation that changed rapidly, communication could not catch up and a delay occurred, making it difficult for Torpedo Squadron to cooperate with Main Body; Main Body did not clarify the role of Torpedo Squadron; the torpedo attack separated Torpedo Squadron from the enemy [4, p504-505]. It also states that in the pursuit, Torpedo Squadron was slower than Main Body, so there was no room for any attempt to take action to attack; the flagship, the light cruiser Abukuma, had a shorter gun range than the enemy destroyers' [4, 508]. There is another recollection that Main Body could run faster than 35 knot, but the destroyers "Wakaba" and "Hatsuharu" could make speed less than 34 knot [20, p303].

The US fleet was worried that the Japanese destroyers would rush in and make a skillful and accurate torpedo attack, as it had suffered several times in the Solomon Islands of the South Pacific. The U.S. Navy speculates that the Japanese destroyers did not do so probably because the destroyers were loaded with replenishments and soldiers [8, p54]. However, the Japanese relevant documents do not refer any loading of supplies and soldiers on the destroyers, stating that the destroyers were saving fuel until a charge but had no opportunity of it. [4, p498].


After all, Torpedo Squadron sailed only to wait for the opportunities during the most of this battle, and only the last part it participated in the battle. As mentioned before, it is considered that this is largely because the roles of Torpedo Squadron in the battle was unclear. Torpedo Squadron must decide the own actions while chasing the enemy fleet and grasping the movement of Main Body. Main Body had not clearly indicated its intention or tactics (at least on the record), so the decisions of action during the battle would not be easy for Torpedo Squadron. If Torpedo Squadron behaved incorrectly, it might have even been accused of interfering with the actions and operations of Main Body later. Therefore, it seems too harsh to blame Torpedo Squadron for failing to take positive action. In order to make effective use of Torpedo Squadron for the battle, Commander-in-Chief of the Fifth Fleet should have instructed Torpedo Squadron on the necessary tactics and actions. This is in contrast to the US fleet described in 7.3.7.


7.3.5 Concern about air strikes by US air forces

One of the reasons why the Japanese fleet abandoned the pursuit of the US fleet after 0700 was that they were concerned about air strikes by the US bombers. One of the grounds for the Commander-in-Chief's decision on the withdrawal was that "the radio intelligence revealed that enemy aircraft had left Amchitka Island at 0640" [4, p496]. In fact, at the Amchitka Island airbase, any bomber had not yet been ready to take off at that time. However, as mentioned above, it is believed that consideration of the possibility of enemy-aircraft coming, that is, the guess that the enemy bombers were about to arrive contributed to the abandonment of the pursuit.

At the Amchitka and Adak islands on the day, the air corps had prepared for regular bombing for Kiska Island as usual. Armor-piercing bombs rarely used were frozen in the ground. It took four hours to mount armor-piercing bombs on bombers that were carefully dug out from the ground. Further, they had to wait for another two hours due to the bad weather [10, p67]. As a result, three B-25 bombers and eight P-38 fighters took off at Amchitka Island around 0905, and they actually arrived over the US fleet at 1104 [8, p54]. The same is true for Adak Island, where thirteen B-24 bombers and eight B-25 bombers, which take off the island at 0836, arrived at 1202 [8, p54]. Neither of them was not able to participate the battle at all.

The radar on Kiska Island having about 300 km range had monitored the movement of US aircraft. In reality, the radar unit in Kiska Island reported on radio at around 0700 that the sky over Amchitka Island was as usual and, by the interception of communication, there was no evidence of US aircraft flying. They also reported at 0930 that they had captured the US aircrafts taking off from Amchitka Island [4, p498] as mentioned in 7.2.7. The naval battle area was about 800 km away from Amchitka Island, and it would take about another 2 hours by aircraft to reach the naval battle area. It was not necessary to evacuate at 0700 if the fleet had accurately appreciated the information on the movement of US aircraft sent from Kiska Island. There may be the problem with the communication equipment damage of the flagship "Nachi", but there were no records of Fleet Command attempting to review any information from the radar in Kiska Island. The intention to cooperate with Kiska Island of the Japanese fleet was likely to be weak.


7.3.6 A problem of communication with the spotter seaplane of "Nachi"

A spotter seaplane was launched from "Nachi" at 0354 (the other planes were dumped because of the damages by firing of the main gun [4, p483]). There are no records left about other spotter seaplanes in the Japanese fleet. The US fleet relentlessly chased away every time the spotters approached, as the accuracy of shellfire increased as they approached. The return order from "Nachi" to the spotter was 0720, and in response to that, the spotter headed to Attu Island [4, p501]. Meanwhile, the US fleet shot down a spotter seaplane at around 0430, and later a US aircraft confirmed the wreckage [8, p40]. Therefore, it was possible that another seaplane was launched from other than "Nachi".

The standstill of "Salt Lake City" at 0655 was recognized by the "Nachi" spotter seaplane, but this important information was not reach the Fleet Command. The report on the standstill of "Salt Lake City" from the seaplane was not recorded in Detailed Battle Report of Torpedo Squadron. There is only the record left reported from Attu Island after it arrived there.

According to the recollection of Lieutenant Colonel Takeo Inoue, Chief Gunner of "Nachi", any observation report from the spotter of seaplane of "Nachi" had not reached him [page 4: 506]. Kinoshita Artillery Staff of "Nachi" stated that all reports from the seaplane were not received at "Nachi" due to the communication device damage [4, p491, 507]. However, since the seaplane received the return ordered to Attu Island, it seems that "Nachi" could transmit messages to the seaplane. If seaplane's report could not be received by "Nachi", couldn't "Nachi" be possible to convey it to the seaplane and take alternative communication means? It may be difficult for ships moving around at high speed, but there may have been means such droping reports from seaplanes ― one of the means in normal situations. Kinoshita Artillery Staff stated that the fleet lacked the measures and training to receive seaplane reports on the secondary ship and convey them to the flag ship [4, p507]. The Japanese fleet might have the problems for communication system, such as frequency allocation and emergency communication ways. Anyway, it is a fact that there was some communication problems during the battle.


7.3.7 Problems on communications during the battle

As mentioned in 7.3.4, Torpedo Squadron could not make effective cooperation on the combat due to the lack of close and smooth communication with Main Body. Also, in this battle, the light cruiser "Tama" separated from the two heavy cruisers of Main Body from around 0350, and was fighting alone until around 0630. The reason is not stated in the Japanese War History Series "Imperial Navy operations in the Northeastern". However, the document states that "Tama's battle actions aggressive and always advancing into the enemy were admirable." [4, p490]. The US Navy material [8] also describes the situations that a light cruiser, believed to be "Tama," often tried to poke or approach. However, it is not sure why "Tama" was fighting on a single ship. As a result, the Japanese fleet appears to have attacked the US fleet independently by the two heavy cruisers, the light cruiser "Tama," and Torpedo Squadron without their coordinations and cooperations.

Light cruiser Tama in arctic camouflage, 1942
https://ww2db.com/image.php?image_id=7643

On the other hand, in the US fleet, there was frequent communication between the flagship, the damaged ship, and the escort destroyers, such as the reports of which ships were damaged when, which ships smoked where, and instructions that the destroyer fleet would launch a torpedo attack on the Japanese fleet in crisis. As a result, the Japanese fleet was often obstructed and forced to change course, hindering effective attacks. The US fleet appears to have escaped the crisis with the systematic responses to the combat.

By contrast, it seems that the Japanese fleet was decisively lacking in communication to collaborate for making effective attacks. Even the Japanese War History Series "Naval Operations in the Northeastern" is not discussing following three points: No reports from the spotter seaplane; Main Body did not give specific instructions to Torpedo Squadron; and the three groups acting out of coordination.


7.3.8 No deployment of air forces to Paramushir

The aggregate transport that caused this naval battle was the operation in response to the appearance of the US fleet near Attu Island in February. The operation was aimed not merely to strengthen the escorts for transportation, but to clean up the US fleet in a naval battle being expected to occur during the transportation [4, p441]. If that is the case, Imperial Navy should have dispatched patrol aircraft and attack aircraft at the airfield of Paramushir on the northern Kuril Islands in preparation for another appearance of US warships. If so, Imperial Navy could have accomplished their purpose. 

Attack aircrafts of Imperial Navy, such as Nell and Betty, had the ranges to make a round-trip to Attu Island. The Fifth Fleet did not have its own attack aircraft. However, was there any means such as asking dispatch attack aircraft to Navy General Staff Office? The reinforcement for this transport only by warships would be inflexible and insufficient. On the other hand, the US fleet thought that it might be attacked by Japanese aircraft during the naval battle. 

Eventually, the attack aircraft and patrol aircraft were not deployed on Paramushir until the Allied troops invaded on Attu Island in May. It seemed that the area around Kuril Islands was thought to have many days with unsuitable weather for flight, but it would not so different from the conditions around Amchitka Island base of the US air forces.


7-4 Aftermath of the battle

7.4.1 Shift of defense strategy

The Japanese fleet was not damaged hardly by this naval battle, but after the naval battle, the heavy cruisers "Nachi" and "Maya" and the 21st destroyer squadron ("Wakaba" and "Hatsushimo") had been decided to head to Yokosuka. On March 30, a strong wind caused the destroyers "Ikadzuchi" and "Wakaba" to contact with each other, and they were severely damaged in the Paramushir harbor. The destroyer "Ikadzuchi" also went to Yokosuka for repair. [20, p306-308]. The light cruisers "Tama" and "Abukuma" also went to Maizuru for repair (or maintenance) in the latter half of April [page 4: 524]. The power of the Fifth Fleet had significantly decreased.

The Fifth Fleet never planned aggregate transports again and passed the command of the sea around Aleutian Islands to the US military. The Japanese War History Series "Imperial Navy operations in the Northeastern" states, as the atmosphere of northern battlefield, that "the northern battlefield had gloomy and negative atmosphere, while the southern battlefield was bright and positive one. The bad weather annoyed us throughout the year. In addition, since the occupation of the western Aleutian Islands, the Japanese troops had continued the passive actions. That is the reasons why the Japanese troops became reluctant."  [4, p500].  This would have been a matter of mood for the Japanese troops, as the weather acts equally on both sides.

On April 12, Imperial General Headquarters decided on the defense policy that Kiska Island, which was difficult to transport, was shifted into the forward base, and the defense group bases would be established on Attu Island and its neighboring Semichi and Agattu Islands [3, p263]. Northern Army Force, the Fifth Fleet, and Imperial General Headquarters held a meeting (Asamushi Conference) at Asamushi in Aomori on April 24 and 25. There, they planned to conduct the second phase of "A-Gou" Operation ― they called it "Fog" transportation, taking advantage of the coming fog season. The operation was to transport soldiers and half a year's worth or one year's worth of supplies to Attu Island and Kiska Island to strengthen the defense using a total of 21 transport vessels from the end of May to the beginning of August [18, p424]. In anticipation of naval battles, they planned to add three heavy cruisers and four destroyers to the Fifth Fleet [18, p428].

As the public announcement had been made in the United States since the early April that the Aleutian operation would be conducted near future, there was a suggestion in Imperial General Headquarters that the Allied troops might invade during the foggy season [18, p417]. However, Imperial General Headquarters didn't think that the situation was so urgent [3, p266]. Neither the Operational Direct Guidelines formulated by Imperial General Headquarters nor the local Army And Navy Agreement described any actions in the case of the enemy invasion [18, p425]. Most of the themata discussing on the meeting were the "Fog" transportation and the affairs after the completion of airfields in May. 

A staff of Northern Sea Defense Corp in Northern Army Force pointed out a possibility that Allied troops would land before the "Fog" transport. He urged in a strong tone, "Because Imperial General Headquarters is ordering to secure, you should show me tangible plans. If there is no such thing, you should change the order to 'desperate defense'." [3, p260]. There is only a document left that "the situation at that time was that there was no other way than promoting transportation based on the default plan for Imperial General Headquarters and Northern Army Force" [3, p259].

After the Allied Forces had actually landed on Attu Island, Imperial General Headquarters decided to abandon Attu Island and to evacuate from Kiska Island in only one week. Considering the above poor situation of the Japanese forces, Imperial General Headquarters should have considered evacuation from both the islands or countermeasures against the Allied troops invaded, because the staff of Northern Sea Defense Corp had asserted at the beginning of April that a invasion to the western Aleutian Islands by the Allied troops might be imminent. The Japanese War History Series "Navy operations in the Northeastern" states that it was mysterious why Imperial General Headquarters did not consider the Allied forces invasion and countermeasures for it [4, p435].


7.4.2 Transportation to the Aleutian Islands after the battle

On April 4, it was reported that Kiska Island had a staple food until July, a side dish until May, and Attu Island food only until the end of April [3, p256]. In early April, the Fifth Fleet attempted twice to transport food and ammunition using two destroyers "Ikadzuchi" and "Usugumo," but these efforts were abandoned due to the bad weather and contacts with US aircraft. [4, p445]. A crew of the destroyer "Usugumo" wrote that "When a US patrol aircraft found Japanese ships near Kiska Island, it immediately contacted the air base on Amchitka Island and drew a large circle over the ships with a smoke to inform airplanes of its location. Then, within 30 minutes, enemy airplanes came to attack. It was difficult for even destroyers to escape from under the circle within 30 minutes" [22, p227].

After the Battle of the Komandorski, transportations by the Fifth Fleet were shifted to be conducted by submarines during heavy fog, and only in emergencies by destroyers [4, p448]. However, the replenishments transported by submarines was very limited. This delay in strengthening the armament became one of the reasons why the Imperial General Headquarters gave up the rescue early after the Allied troops had invaded. Second District Unit (Attu's Garrison) could not hold the island until the Japanese was ready for relief. Colonel Yamazaki, who was appointed to the commander of Second District Unit, headed for Attu Island by the aggregate transportation in March, but could not reach Attu Island due to the Battle of the Komandorski Islands. He managed to reach Attu Island by submarine on April 18.

(Please note that the time and day are Japan Standard Time (GMT+0900))

References are listed in "Reference" in the upper right-hand side.