Alan D. Zimm Read online

  Three carriers were assigned to the Pacific Fleet, Enterprise, Lexington, and Saratoga. Saratoga was at San Diego. She had not been operating out of Pearl Harbor while in drydock at Bremerton. The intelligence officer’s reports would not have mentioned her presence since October.

  The attack against the carrier moorings would involve a fairly easy approach along the length of the loch past Pearl City, with plenty of room to set up the attack and establish the proper release conditions. The approach would be at an angle so the targets would be foreshortened, reducing the target cross section by about 30%. The approach would pass several nests of destroyers, so surprise would be needed to avoid AA fire.

  These 16 bombers represented 40% of the torpedo-armed bombers. The maximum number of carriers the Japanese could expect to find in port would be three, two more likely, fewer still considering the carriers spent a lot of time underway. At 40% hits, this group would be expected to score six hits, enough to sink two carriers; at 68%, ten hits, which could account for all three carriers or gut two; 82% would give thirteen hits, over four hits for each carrier assigned to the Pacific Fleet.

  This allocation provided killing firepower against the carriers.

  Weighting the Attack

  The battleship anchorages might expose six battleships along battleship row. The Japanese assigned 24 aircraft to attack this group, four per potential target. The other group of 16 bombers would be attacking, at most, three carriers, more likely only two, for five to eight torpedo bombers per potential target. They maintained this overweight even after they received an intelligence report that none of the carriers were in port, in the hopes that the carriers might return.

  Carriers did not have the built-in anti-torpedo protection of a battleship, and could sustain fewer hits before sinking. The Japanese felt that three to four torpedo hits were enough to sink a carrier compared to four to five for a battleship. The Japanese plan sent sixteen torpedo bombers to achieve the eight hits needed to guarantee the sinking of two carriers (two bombers per needed hit), but only 24 bombers to get the 30 hits needed to guarantee the sinking of six battleships (0.8 of a bomber per needed hit).

  The D3A Val dive-bombers added more firepower against the carriers. Eighty-one Vals were to be sent against fleet targets in the second wave, and their first priority was aircraft carriers. If the carriers had been sunk or capsized by the torpedo attack, the dive-bombers were instructed to bomb the hulks. The planners added suspenders to their belt to ensure that the carriers would never sail again.

  This was likely the aviation commander Genda’s work. Genda had previously made clear his views that battleships were obsolete and that carriers were the most important objectives of the attack. In initial estimates for Yamamoto in February of 1941, Genda made carriers the attack’s first priority, but was overruled.

  Genda gave voice to his opinions the day before the attack. Aboard Akagi, he learned there were no carriers in Pearl Harbor. However, another staff officer pointed out that the carriers might return. Genda cheered up, saying, “If that happened, I don’t care if all eight battleships are away.”48

  Another thing that suggests that the weighting was intentional was the numbers involved. The carrier attack moorings were attacked by two groups of eight carrier attack planes. Eight is an unusual number in Japanese air doctrine—it is not a multiple of the usual 3-plane shotai. This indicates that the number of planes sent against the carrier moorings was not a product of doctrine but a conscious decision.

  Additional evidence comes from Abe Zenji, a dive-bomber pilot in the second wave, who stated, “If we could not find the carriers, our secondary targets would be cruisers.” Speaking of the attack as a whole, he went on to say, “We missed our main objective, the aircraft carriers, since they were at sea.”49

  Genda and his aviators had their own priorities, and killing carriers was at the top of the list. They would go for battleships, but they were doubly interested in pulverizing their own US counterparts.

  Did Genda slip this allocation into the plan, or was it made with Yamamoto’s knowledge and consent? It is hard to say, because Japanese staffs had a much different culture from Western staffs. The Japanese military vested an unusual amount of independence to middle grade officers, to the extent that some started wars or initiated invasions or even assassinated political leaders on their own initiative. The planning environment allowed little interplay between seniors and juniors. Criticism would imply that the work was inadequate, a failure by the junior officer. Failure was anathema, a shame that in some cases might only be expunged by ritual suicide.

  So, whether the firepower distribution was as Yamamoto wanted, or as Genda slipped in, is hard to discern. It likely was Genda’s. After all, with an opportunity to assign additional firepower to targets when he granted the midget submarines a place in the attack, Yamamoto assigned them to strike battleships. Yamamoto wanted more hits on battleships; Genda wanted more on carriers.

  It is never good when the objectives of the boss and his chief planner are not aligned.

  There is another consideration. The torpedo bomber allocations were made early in the planning process. Some Japanese documents refer to the possibility that there might be four American carriers in port. If that was their expectation, then the Japanese were sending four torpedo bombers per carrier, the same as what they allocated against the battleships. This is still an overweight against the carriers, since carriers were more vulnerable than battleships. But if this was the case, it demonstrated again the incredible inflexibility in Japanese planning. They knew from latest intelligence reports that there were only two carriers operating out of Pearl Harbor, and should have adjusted their allocations accordingly. Their failure to do so would have tactical implications—as will be seen, it contributed to much of the confusion of the torpedo attack.

  Attack Routes and Sequencing

  The torpedo units were to attack in single file, a decision that Fuchida claimed to have made. This was a remarkable choice. Line abreast was the usual torpedo attack formation against warships, to make it more difficult for ships to maneuver and avoid several torpedoes dropped simultaneously, and to spread the defensive AA fire. Fuchida claimed that the formation was chosen because it was thought that “the long thin line [was] especially well suited to Pearl Harbor with its narrow channel and many obstacles.”50

  (6) Torpedo Bomber Planned Attack Routes

  Gordon Prange’s At Dawn We Slept contains a chart giving Fuchida’s representation of the planned lines of approach for the torpedo bombers. The dotted lines show the attack paths assigned to the Battleship Row attack, while the solid lines were the routes assigned those attacking the carrier anchorage.

  There were 18 different attack paths, six against Battleship Row, two against potential battleship moorings along 1010 Dock, and five against potential carrier moorings northwest of Ford Island. Five were directed against potential cruiser locations. This scheme was likely formulated by Murata, a torpedo expert, who had been delegated responsibility for much of the planning for the torpedo bombers.

  The Japanese would not know with surety which of the berths would be occupied with priority targets. Their last intelligence would be 24 to 48 hours late. Ships could change their moorings or enter or depart the harbor after the last update. Consequently, the aircraft commanders were to decide their own target selection and routes upon arrival. Aircraft were not pre-assigned to particular routes.

  The problem that leaps from the chart is that 11 of the 18 attack routes cross, setting up the potential of mutual interference. Each group had routes that crossed, a problem that might be controlled within the aircraft of a single group through good communications and tight leadership. However, considering that each group was to attack in two lines with aircraft at 500 meter intervals, that each aircraft commander had the authority to choose whichever route he felt best, and that aircraft would be out of communications with their leaders, the possibility that aircraft would select crossing r
outes remained.

  The greater potential for mutual interference came from the fact that routes belonging to different groups also crossed. This potentially sets up mid-air collisions or air-to-air near misses that could disrupt the attack—in modern terminology, the torpedo bombers’ attack routes were not deconflicted.

  The most significant crossing route was one assigned to Hiryu’s and Soryu’s bombers going against the carrier moorings. They were given the option to deliver an attack against the fleet flagship’s berth along 1010 dock, which might be occupied by either a carrier or a battleship. Aircraft choosing this route would recover from their run across the approach paths of bombers going against Battleship Row.

  Crossing attack routes can be deconflicted by altitude or by time. Altitude was obviously impossible, as the torpedoes had to be released from the same low altitude. The Japanese intended to feed bombers into the attack one at a time from each of the four groups, each sending in an aircraft at about seven-second intervals. But while each aircraft in a group would be separated by time, there was no way to prevent different groups from interfering with each other. Aircraft that were searching for alternate targets or repeating aborted runs would also be a hazard as they tried to re-insert themselves into the attack pattern. The aircraft were too far apart to communicate by hand signals, and the use of voice radios was apparently not considered.

  This formation was a command and control nightmare. Once the attack began, the leaders could not exert any control, and the potential for mutual interference was great. None of this was practiced during the two rehearsals, with the Japanese taking the position that none of this type of interference would occur. Here, as throughout the war, Japanese planning generally assumed that events would be executed as planned, with little consideration for what might go wrong, and no contingency planning.

  While Fuchida’s drawing might suggest many aircraft attacking simultaneously, the natural result of using long strings of aircraft was that only a few routes would be used. Aircraft attacking one at a time in “follow the leader” style would be naturally drawn to select the easiest approaches. For the Akagi and Kaga torpedo bombers going against Battleship Row, the attack would concentrate, as a natural result of a form of self-organizing agent-based behavior, on the southern end of Battleship Row.

  The Japanese noted this propensity during their November “rehearsal.” Nothing positive was done to correct the problem.

  The attack CONOPS (concept of operations) required the torpedo bombers to come into AA range one at a time at seven second intervals, sufficient time for a .50-cal AA machine gun to shift targets. Each bomber would in turn face the concentrated fire of all the automatic weapons along that path, without other aircraft to spread the defenders’ fire.

  The planners’ scheme rejected the usual method of delivering a torpedo attack, with planes approaching in line abreast. Several modern attack aviators examined this aspect of the plan and believed that line-abreast attacks in waves of four against Battleship Row would have been possible, especially if they were echeloned at 50-yard intervals. Each bomber in a wave would attack a different target. The formation would allow the trailing aircraft in each echelon to see the aim point of the aircraft ahead of it, allowing it to select a different target. This would spread rather than concentrate the attack. In this way the attack against Battleship Row could have been completed in under a minute, with better control over the distribution of fire, fewer chances for mutual interference, and greater dispersion of the defensive fire.

  Torpedo Nets

  Torpedo nets had the potential to thwart the torpedo attack. The Americans considered using torpedo nets, particularly after Taranto in November 1940, where British torpedo bombers sank several Italian battleships in port. The idea was rejected by Admiral Kimmel. Kimmel may have felt the harbor was too shallow for torpedoes, although he knew of the successful British torpedo attacks in Taranto harbor and had been warned by the CNO that the depths of water at Pearl Harbor did not make a torpedo attack impossible.51 Kimmel rejected torpedo nets because he believed they would interfere with an emergency sortie.

  While Japanese intelligence indicated that torpedo nets were not being used in Pearl Harbor, they had to consider the possibility that the Americans would install them as the threat of war increased. Tokyo sent a message to the Consul General at Honolulu, received on 2 December, asking that information on torpedo nets be wired on a daily basis.52

  In the days underway while steaming towards Hawaii, the torpedo bomber aircrews considered ways to defeat torpedo nets. They talked of dropping their torpedoes between the nets and the ships, which was impossible for a number of technical reasons. Their final solution was to have aircraft intentionally crash into the nets, a bizarre expedient with little chance of success, but a very Japanese course of action. Pilots were selected for the task.

  One potential solution was to use 250kg bombs from D3A Val dive-bombers to blow the nets apart. Photographs show that the line of floats buoying the nets would be visible, yet as an option it was never mentioned.53 This was another example of Japanese stovepipe thinking. Instead of using the different types of aircraft cooperatively, one type helping the other, the torpedo nets were considered to be a torpedo bomber problem for the torpedo bomber aircrews to solve, even if it required the most extreme expedients.

  Alternately, a B5N Kate could have been loaded with a large number of 60kg GP bombs. It could have dropped a string of them across the nets from low altitude, giving a high probability of blasting holes in the nets. This was another solution evidently not considered.

  It is curious that the aircrews were forced to consider the problem at all. A counter to torpedo nets ought to have been worked out earlier by the planners. As it was, it became a nagging worry hanging over the heads of the aviators for the duration of the 13-day transit to the target. Those selected to crash the nets must have been relieved when they learned, a bare four hours prior to launch, that the Japanese Hawaiian Consulate reported that the harbor had no torpedo nets.54

  Level Bombers

  The payload allocation of B5N Kate carrier attack planes was the key planning decision. If seven Kates with torpedoes were likely to cause the same damage against battleships as 50 to 100 planes with AP bombs, why were so many of the Kates armed with bombs?

  At the time of the allocation decision, torpedoes still had significant uncertainties, and the technical problems had not been solved. The shallow water problem was resolved only a week before the end of training, torpedo nets were only confirmed as absent a day before the attack, and the double-berthing problem remained. The Japanese had to consider the possibility that the battleships might be vulnerable only to AP bombs.

  After intensive training, the Japanese expected exceptional performance from their level bombers, anticipating “4 out of 5 salvoes to hit.”

  In American aviation terminology, a group of bombs dropped simultaneously on a target by a formation of level bombers was called a “salvo.” In the gunnery terminology of the United States and Great Britain, a salvo was considered to have hit if one of the shells hit; it did not imply that all hit. This interpretation is confirmed by Lieutenant Takayoshi Morinaga, who said, “If one plane out of five bombers hit the target their attack would be successful.”55

  A Japanese post-attack evaluation entitled Lessons (air operation) of the Sea Battle off Hawaii (hereafter referred to as Lessons) was written perhaps in late August 1942 by the Battle-Lessons Investigation Committee. Fuchida probably served on the committee.56 Lessons cited two numbers for the level bombers’ expected results, based on a practice attack conducted on 24 October 1941: “Probabilities of hitting targets, 10% (Max. 17),” and “Probability of striking targets, 50% (Max. 75).”57 “Probabilities of hitting targets” was likely the hit percentage per bomber, which would translate into five hits with a maximum of seven. “Probabilities of striking targets” was likely the probability of the salvo hitting. Fifty percent would equate to five hits; 75%
to eight. Thus, the Japanese expected five to eight hits with AP bombs against inboard battleships, with additional hits possible on outboard battleships that might be caught within accurate bomb patterns.

  Lieutenant Commander Sadao Chigusa’s diary, in an entry ten days after the attack, said “The power of the No. 80-5 [AP] bombs… is enormous and sufficient to deal any battleships in existence a mortal blow if hit by 5–6 bombs…”58 If this coincided with the planner’s expectations for the attack, they might have expected enough hits to inflict sufficient cumulative damage to sink one battleship, if the hits were concentrated against one target. Since the comment was made after the attack, Arizona’s fate might have biased Sadao’s opinion.

  There were two different models of ship damage used in this period: a deterministic model and a stochastic model. Which model is chosen is critical to how the potential results of an attack would be calculated.

  In the deterministic model, an estimate is made of the average number of hits that are required to sink a battleship. Damage is cumulative. Ships are sunk by the cumulative effect of many bombs. If 18 hits are thought to be needed to sink a battleship, one hit is worth 5.6% of a battleship. A ship would be considered crippled after 50% cumulative damage (9 hits).

  The Japanese ascribed to the deterministic methodology, as did the US Navy in 1941. The US Naval War College’s “Fighting Strength” methodology assessed the “Life” of a battleship to be about 18 penetrating 14-inch (680-kg) shell hits. The Japanese 800kg armor piercing bombs would be valued at 1.08 equivalent 14-inch hits. They were considered to be able to penetrate 150mm (5.91 inches) of armor when dropped from 8,200 feet, more when dropped from the planned delivery altitude of 3,000 meters (9,843 feet), so the American battleships’ deck and turret top armor was vulnerable to penetration.