The battleship Arizona was moored with six other battleships of the U.S. Pacific Fleet on the northeastern edge of Ford Island, primary targets of an Imperial Japanese Navy air and submarine attack. As part of its intricate planning effort, Japanese naval staff members were well aware of the mooring profiles of the Pacific Fleet's capital ships, and the attack plan took these into account.
The battleships usually were moored in pairs, with the outboard ship vulnerable to torpedo strikes. Conversely, the inboard ships were screened from torpedo attack. To negate this layered mooring profile tactically, the Japanese used high-level horizontal bombers. Operating at or near 10,000 feet, ten flights, consisting of five bombers each, approached the battleships from the south and flew northward, parallel to Battleship Row. The lead bombardier of each triangle-shaped formation signalled the other four bombers to drop their bombs on his command. All five bombers then dropped their single modified bombs simultaneously, hoping the warships below would receive one to three hits.1 The attackers scored several times against the battleships. For the Arizona, moored inboard of the repair ship Vestal (AR-1), the tactic employed was devastating.
Just prior to the Arizona's receiving her fatal blow, Captain Eric Haakensen, U.S. Army Medical Corps, apparently had gone to his stateroom and picked up his 16-mm motion-picture camera, intent on recording the developing attack. At that time, he was on board the hospital ship Solace, (AH-5)moored off the Arizona's starboard quarter, with her stern pointing toward Ford Island. Together with another man (whose hand can be seen in the photography), Haakensen captured one of the most cataclysmic events in warfare and film history. According to National Park Service's Daniel Martinez, historian at the Arizona Memorial, "the Haakensen film is like that of Abraham Zapruder, who filmed the assassination of President John F. Kennedy; Eric Haakensen, 22 years earlier, captured the death of the battleship Arizona. Both images are vivid and only one has undergone intensive photographic scrutiny."2 This analysis helps us understand how the Arizona was lost.
General Methodology
This study called for the discipline and talents of experienced imagery intelligence analysts, skilled in battle damage assessment and in military and naval operations; photogrammetrists/image scientists, whose experience finds them working with premiere aerial and space-based imaging sensors in the United States; and a respected naval architect/engineer who has specialized in naval damage analysis and the effects of explosives on warships.
The methodology adopted comprised four inter-related phases. In Phase 1, the Timeline Analysis, the team created an imagery timeline using many of the pertinent Japanese aerial photographs in combination with Haakensen's motion picture photography. To the best of the team's knowledge, this is the first fusion of multiple imagery sources of the attack to create a timeline, which complements the film, torpedo hit, and forensic analyses.
The second part of Phase 1 involved a mathematical analysis of the explosive effects of the Arizona's forward black-powder and 14-inch gun smokeless powder magazines. This also involved an analysis of the Arizona's hull and interior compartmentation.
Phase 2 had three inter-related avenues of research. The first involved two traditionally complementary disciplines in imagery intelligence interpretation and analysis. The Haakensen film was cleaned and digitized and the camera motion stabilized. Obvious imperfections in the photography (created by accumulations of dirt, hair, smudges, etc.) were removed. The digitized product was then examined from a photo-intelligence-analysis standpoint. During this phase, the team also reviewed collateral photographic sources. The team looked at each frame of the film to detect and identify activity within the entire field of view, including comparison of color images with black-and-white frames. This continued through Phase 3, in which the team selected images critical to their respective analyses.
In Phase 4, the team conducted photogrammetric analyses in response to the findings of forensic interpreters. Critical measurements ranged from determining Haakensen's location with respect to the Arizona to calculating the proximity of smoke plumes seen within the various frames.
Determining the Imagery Timeline
All Japanese aerial attack photographs received by the imagery intelligence team (Rodgaard, Lucas, and Biache) came from the Arizona Memorial film library. The first selected for the timeline was an image showing the effects of the first aerial torpedo strike against Battleship Row. It also shows the presence of a midget submarine. In its earlier work for the Park Service, the team had calculated this photograph to have been taken at approximately 0803, plus or minus 30 seconds. With this as the anchor, the team then determined the sequence of the photography taken by the Japanese horizontal bombers. This timeline then incorporated the Haakensen film. The result was a multi-camera perspective of the last moments of the Arizona.
Explosive Effects of Magazines
Mr. Hsu, a forensic engineer with Anteon Corporation, conducted an analysis of the explosive effects of the Arizona's forward black powder and smokeless gun powder magazines. Before initiating the forensic analysis, he examined carefully the chronology of events and the attack profile based on extant still photos and the Haakensen images. He first identified the speed and bomb-path trajectory of the modified naval armor-piecing shells that destroyed the ship. He then conducted a thorough review of the ship's compartmentation and access to include the physical proximity of the magazines to one another and to the migration of the conflagration through the ship. He conducted a comparison of the ship's hull prior to the attack and drawings of the ship's hull that were commissioned after the attack.
Mr. Hsu traced the weapon's possible path as it penetrated and traveled through the ship's hull and how the explosion expanded progressively forward through the ship. He concluded that the fatal blow came from a weapon that did not reach terminal velocity and that the weapon struck the forward part of the ship. The modified shell exploded in the vicinity of the starboard 14-inch smokeless powder magazine and fuel tanks. The resulting conflagration went from the starboard fuel tanks and the smokeless powder magazines to the center-lined black-powder magazines.
As the explosion intensified, it expanded to the portside smokeless powder magazines, venting forward through the port and starboard sides of the ship's hull above the side armour belt and aft toward the fireroom's intakes. This was confirmed by subsequent independent photographic interpretation of Haakensen's film frames.
Torpedo Hit Analysis
On the morning of 7 December, the first wave of Imperial Japanese Navy "Kate" torpedo bombers, carrying modified Type 91 torpedoes, attacked Battleship Row at 0755. The battleships were berthed along the southeast shore of Ford Island. The California (BB-44), Oklahoma (BB-37), West Virginia (BB-48), and Nevada (BB-36) all had taken direct hits from torpedoes dropped by the bombers, and the West Virginia and possibly the Oklahoma received a torpedo each from one of the midget submarines at about the same time.3 It also was reported that the Arizona received a torpedo hit. This was logical, because all of the battleships (those that either were moored outboard or singly) sustained torpedo hits. If a torpedo had detonated against the Arizona's hull, however, the damage would not have been limited to the battleship. With the Vestal outboard of the Arizona, the torpedo would have struck her—not the battleship—inflicting substantial and unmistakable hull damage. Reports indicated, however, that a torpedo went under the stern of the Vestal and struck the Arizona on the port bow. In fact, if this had happened the result to the Arizona could have been the same as that suffered by the cruiser Helena (CL-50).
The Helena was berthed alongside 10/10 dock, and a smaller ship, the Oglala (CM-4) was tied outboard of the cruiser. A Japanese torpedo travelled under the Oglala, striking the Helena. The resulting underwater explosion sank the Oglala. Given the same type of torpedo, the same thing would have happened to the Vestal, if a torpedo had in fact travelled under the ship and hit the Arizona.
Underwater explosion shock and warhead holing analysis studied shock-wave and fluid interaction against the ship's hull. Together with the physics associated with torpedo explosions, these analyses do not support the theory. The Vestal's battle damage assessment report and the Arizona's salvage report contain no creditable torpedo detonation damage. In addition, the aerial photographs taken by the Japanese horizontal bombers during their attack runs do not show anything associated with a torpedo hit.
Image Science Analysis
The next step in the process was to see what visible evidence existed within the Haakensen film and the Japanese aerial photographs. Several specific operations were performed on the Haakensen film to assist in the comprehensive analyses. The copy of the film that our team received was not in the best possible condition. Neither was the film a contiguous series of images. Over the years, the original was cut and spliced numerous times. The reproduction process allowed dust and other artifacts to become embedded into the emulsion. Finally, the film was processed backward ("flopped," or "mirror-imaged," reproduced through the emulsion).4
The first step was to digitize the copy into individual frames. With this accomplished, the team's photogrammetric scientist, Larry Dambra, reversed the photography to its original and correct side. He also applied a software package to the digitized photography, which stabilized the individual frames. One of the first things he noticed was a time lag between the first frame showing the first signs of the explosion and the rest of the footage, showing the progress of the conflagration. In addition, he pointed out that Captain Haakensen had moved a noticeable distance, having stopped once and moved forward and to his left when he resumed filming.
This phase of the study also involved the two color frames that accompanied the black-and-white film received by the analysts.5 The question posed to the team was whether the color frames were part of the original film or taken by another camera. If these were original, where did they match to the black-and-white frames? The color images apparently were the originals, because Captain Haakensen used color film during the attack. This prospect amazed the team because of the logistics that Captain Haakensen had to have to ensure the film did not degrade in the tropical environment. The hospital ship's refrigerators must have stored more than just medically related supplies.
Photographic Interpretation and Analysis
This frame-by-frame phase of the analysis focused on several aspects: the position of the cameraman during the filming of the event; the status of the film reviewed; the field of view for the camera; and ancillary information dealing with the position of the Solace in relation to the Arizona.
The Solace was moored in the East Loch, with her stern pointing toward Ford Island. Using photography from the timeline, we calculated the distance of the ship to the island and to the Arizona. The distance of the Solace from Ford Island was approximately .35 nautical miles. From the Arizona, she was .45 nautical miles away. The horizontal field of view for the camera at the time the motion picture photography was taken was approximately 15°. The left edge of the frame was looking due south, while the right edge of the frame was looking south-southwest.
The team reviewed all frames and of those only the frames that showed apparent change were subjected to detailed analysis. In the initial review process, the team determined a sequence of events. First, the photography we received from the Park Service was a reproduction in which every frame had been duplicated. The even-number frames were the copies, so the team analyzed the odd-number frames. Second, the images appear cropped from the originals. Third, the team confirmed the initial assessment that a gap occurred between frames 129 and 131. Apparently, Captain Haakensen did not run his camera continuously (or a later portion of the film had been edited or cut).
The analysis of each frame was prefaced with an overall assessment of the scene, specifically the image activity and scene changes of the Arizona within each frame. The team followed this description with what it detected as clear evidence, throughout the entire image scene, of the results from the attack on the Arizona. The following details how the team treated each of the critical frames, starting with the original (frame 55 was the database image).
Frame 55
The Arizona is in the center of the frame, with the eastern part of the naval base (the submarine base) in the background and left of the ship. The repair ship Vestal can be seen behind the battleship. Her masts and funnel are visible above the superstructure of the Arizona. To the ship's right and in the background is the main part of the naval base, where three water tanks can be seen. Between the Arizona's masts is a large 200-ton lift crane, and to the ship's right and in the background from Haakensen's camera position are two cage masts of the battleship West Virginia. To the left of the Arizona is the bow of the battleship Nevada. She has the false-painted off-white bow wave camouflage on her starboard bow and the forward mooring quay is visible to the left of the wave. The left format edge of this frame cuts through the Nevada's turret number 1, with only the right gun barrel showing (the other two are hidden behind the right barrel). Running across the frame and situated between Mokunui Island in the foreground and Battleship Row is the elevated pipe. In this view, it is positioned nearer to the island than to the battleships. (Other earlier overhead U.S. Navy photography shows the pipe to be closely adjacent to the island. This photography was obtained from the National Archives subsequent to the Haakensen film analysis.)
Moving right from the island, the pipe seems to merge with the head of a small pier. Mokunui Island can be seen, again, near the bottom of the frame and in the center of the format. A geometric, possibly concrete structure is sitting atop of the island. Poles can be seen sticking above the water and around the island. They appear to be metallic, and in all likelihood they were navigation markers, warning ships of the shallows surrounding the island. The water is rippled but calm in the lee of the island and between the island and the Arizona. A large canted structure is to the immediate right of the cameraman (Captain Haakensen), and it dominates that side of the frame. It is chipped, scratched and possibly unpainted. This may be the lower part of one of the Solace's boat davits.
Dark smoke rises above the ships, drifting left to right (east to west).6 A possible bomb splash or smoke can be seen near the stern of the Arizona but on the starboard side of the Vestal. A bomb splash can be seen forward of the Arizona and broad off her starboard bow. Airborne debris and smoke can be seen on the Arizona's foredeck, forward of turret 1. The team equates this to a probable bomb penetration. The West Virginia is listing severely to port, estimated at approximately 20°. In the background and between the Arizona's masts, a smoke plume shoots skyward from exploding ordnance. This is probably the result of bomb strikes against Hickam Airfield.
Frame 105
The scene has not changed significantly since frame 55. Smoke continues to move from east to west, and the debris and smoke from the possible bomb penetration of the forward deck of the Arizona continues to grow. A significant bright spot appears, however, near the forward gun turrets of the ship, having the appearance of a second bomb penetration. This bright spot shows also on frame 107, but this one is much more dim. Its appearance on two different frames, however, reduces significantly the possibility of its being a film artifact (imperfection).
Frame 129
The camera position is approximately the same as frame 55. The frame shows tremendous vibration. The team speculates this is the result of Haakensen being buffeted by shockwaves from the second bomb detonation.
Frame 131
The camera position is changed in relationship to frame 129, and the view is slightly to the right and higher up. Haakensen appears to have moved forward and to his left. There is no indication of vibration within the frame. The possible davit is no longer in the field of view on the right side. The Arizona still is in the center of the frame. The island appears closer and the nearest pole appears at the edge of the island and astern of the Arizona. A portion of a man's left hand is visible at the left edge of the frame. The man is pointing skyward with his left index finger.
Because the camera may have been shut off between frames 129 and 131, or portions of the film may have been cut, the Arizona's magazines already have exploded, and the fireball and smoke are expanding. Note that the light from the fireball is reflecting on the water and is illuminating details on the deck, hull, mast, and superstructure. Along the hull, is the officers' brow and just aft of it is the white canvas top of one of the ship's utility boats. The left side of turret 4 can be seen because the light from the fireball is reflecting off its surface. Part of the awning suspended over the ship's stern and the starboard quarterdeck awning also are visible. The aft mooring quay can be seen to the right of the quarterdeck awning. Forward along the hull and just to the right of the fireball are at least two more utility boats, their white canvas tops prominently in view. To the right are the mastheads of the battleship Tennessee (BB-43).
In the background, the stack of the Vestal is billowing black smoke, and smoke from the Arizona blast is obscuring the West Virginia. A rudimentary calculation of the length of the smoke from the Vestal's stack indicated a length of approximately 100 feet. Smoke from the Vestal's stack is absent from frame 129. At the time of the attack, the trade winds were blowing at 17 miles per hour or approximately 25 feet per second.7 Employing these numbers, the estimated time difference between frames 129 and 131 is a minimum of four seconds.
Disposition of Findings
A report on the interim findings of this analysis will be presented to the National Park Service and the Arizona Memorial during the 60th anniversary commemorative symposium on Pearl Harbor, during the first week of December 2001. In addition, these results will be part of The Discovery Channel's documentary on Pearl Harbor to be aired that same week. Prior to these two events, the team is sharing here some of its findings and conclusions.
*This study stems from a relationship the team has had with the National Park Service since 1993. Discovery Television provided financial backing, Boeing-Autometric provided logistical support, and the Marine Forensic Panel of the Society of Naval Architects and Marine Engineers contributed to the analysis. Erik Nelson of Termite Films incorporated the findings into The Discovery Channel special, "Death of the Arizona," which premieres on 2 December 2001.
1. The bombs used by the horizontal bombers were modified Navy 15-inch armor-piercing gun shells.
2. This quote was taken from the proposal Mr. Martinez drafted to Mr. Rodgaard and Boeing-Autometric, Inc.
3. See December 1999 issue of Naval History, pp. 16-23.
4. This is well known to historians, and we have noted that recent documentaries showing the explosion also have returned the photography to its original perspective. To the best of our knowledge, however, this change has not been well publicized.
5. Mr. Martinez informed the team that the National Archives knows the history behind the film and how it was reproduced and altered from its original color version to black-and-white.
6. Mr. Martinez informed the team the smoke was probably from a downed Japanese aircraft that splashed near the Nevada. The Nevada had yet to suffer damage.
7. The speed of the trade winds at the time of the attack was taken from Japanese sources. Their knowledge of speed and direction of the wind was critical in the attack plan for the horizontal bombers.