Share This Article

In the final year of the Pacific War, the Japanese worked undetected on a massive underground fortification project that might have convinced Emperor Hirohito to fight on— even after the atomic bombs were dropped.

A mushroom cloud billowing up toward the heavens is fixed in our imaginations as a symbol of ultimate destruction. With two blasts, the United States brought its bitterest opponent to its knees after years of brutal combat and added a measure of human reality to the Atomic Age, which had begun in a lab in 1942. In the first, on August 6, 1945, 60 percent of Hiroshima and an estimated 130,000 of its inhabitants were incinerated in a flash. A second detonation at Nagasaki three days later leveled a third of the city and caused another 75,000 casualties. The devastation wrought on Hiroshima and Nagasaki was significant, but it was not total.

Deep beneath the hills of Nagasaki, a tunnel complex shielded 2,500 Mitsubishi factory workers from the bomb’s catastrophic effects. Buildings directly outside the tunnel entrances and a nearby aboveground plant were destroyed with high loss of life, but those toiling away deep inside the earth remained unscathed, the only damage being tunnel entrances partially pushed in and some exterior doors shattered by the blast. Other underground factories in Nagasaki also withstood the bombing.

While the Japanese government had not designed those facilities to withstand an atomic bomb, their survival was a testament to a hastily planned yet ambitiously pursued underground construction program that—given more time—could have profoundly affected Japanese confidence and influenced Emperor Hirohito’s decision to agree to the unthinkable and sue for peace. Considering Allied ignorance of the extent of the Japanese underground effort, had the planned invasion of the Home Islands—Operation Olympic— actually occurred, estimates of expected casualties into the millions might have seemed paltry. In the end, the bunkers had little impact on the ultimate outcome of the war, but their story says a great deal about Japan’s determination to resist and the potential costs of incomplete intelligence.

The scope and strength of German fortifications—the West Wall, Siegfried Line and Adolf Hitler’s Berlin Bunker—are well ensconced in popular imagination, as are the less substantial cave, coral and coconut log fortifications used by the emperor’s soldiers on Tarawa, Peleliu, Iwo Jima, Okinawa and other remote but fiercely contested Pacific islands. But that is only half the story. By mid-1944 the Japanese government was implementing a strategic underground facility program nearly rivaling Nazi Germany’s and likely surpassing a similar program in the Soviet Union.

The paper trail on the Japanese effort is long and thorough. U.S. intelligence reports declassified later and high-level Japanese documents that survived the war reveal that the Japanese government, with Hirohito’s approval, designed an extensive underground program that was intended to protect the emperor, his family and staff, top-level civilian and military leaders, military industry, fleet and army command posts, and communications centers. Other sources also point to the program’s use to protect underground biological weapons laboratories in Manchuria. In fact, by August 1945 more than 200 strategic underground facilities of substantial size and scope were either planned, under construction or operational. As thorough as their intelligence was, the Allies had no idea of the extent of these facilities until after Imperial Japan’s final defeat.

Strategic deep-underground facilities were built within the Home Islands between the late 1930s and 1943, but it was not until mid-1944 as the war situation worsened that the government decided to build underground facilities for virtually every government agency and military entity. On July 10, 1944, the Japanese cabinet approved a proposal to create an emergency construction corps within the underground construction unit of the Railways Secretariat. Prime Minister Hideki Tojo, just a week before his resignation, agreed and signed the order for the minister of transportation and communications to recruit workers from railroad companies to build these underground structures. Railroad workers were likely chosen because they had experience tunneling through mountains. Even with the Japanese war machine working at maximum effort, special provisions ensured that labor and materials were available for the organization and operation of this new corps.

Construction of the bunkers was prioritized to maintain the “functionality of domestically important facilities.” At the top were the army, navy, Ministry of Home Affairs and Ministry of Munitions. The order may have also supported underground tactical military sites, like those built to repel an expected ground attack on Kyushu.

Six months after the creation of the corps, the minister of transportation and communications requested an imperial order to support the corps and increase its funding and staff. On January 12, 1945, the cabinet agreed to add more than 200 staff members, including an imperial appointee. By July 31, 1945, 1,744 workers were assigned to the corps.

Japan’s military leadership was astute in constructing such facilities deep underground; it made them impervious to conventional aerial bombing and, although they could not know this, nuclear weapons—unless attacked directly. The nuclear survivability of the tunnels was highlighted in a postwar study by the U.S. Strategic Bombing Survey (USSBS) on the “Effects of Atomic Bombs on Hiroshima and Nagasaki,” whose inspectors discovered the existence of many of the bunkers while preparing their study.

When the cabinet order on underground construction was first signed in July 1944, a plan was also approved to relocate key government and military facilities from vulnerable spots, such as Tokyo, to the town of Matsushiro, about 110 miles northwest of Tokyo and five miles south of Nagano. “Temporary” command facilities would be built there. This enormous project entailed three large tunnel systems about 100 feet beneath Mount Minakamiyama, Mount Zozan, and Mount Maizuruyama. The emperor, his family, government agencies and the Imperial General Headquarters would all move to this huge subterranean complex. Military communications units, the central telephone office and the NHK (Japanese Broadcasting Office) would also have space. To ensure secrecy, laborers recruited for the project were told they were building an underground warehouse for the army.

In November 1944, excavation commenced on the imperial headquarters at Mount Maizuruyama. The emperor’s temporary space was opulent and lined with beautiful cypress wood. An order was also issued to build an underground chamber at Matsushiro to protect the imperial regalia—the curved jewel, the sacred mirror and the sacred sword—which symbolized the imperial house’s legitimacy. Construction was almost complete by August 1945, but it was never occupied by the intended tenant.

The tunnel complexes in Matsushiro were designed by engineers at the Department of the Army. The Ministry of Transportation and Communications’ Underground Construction Unit managed the enormous project. When sufficient domestic labor could not be found, several thousand Korean laborers were conscripted to dig more than seven miles of tunnels at Matsushiro. Other laborers eventually joined the project, including soldiers of the Eastern Division and Construction battalions, the Industrial Reserve Army, the Labor Service Corps, and finally students from the railroad school and local public schools. Conditions during construction were horrendous, particularly for the Korean workers: Estimates of the total number of deaths range from a few hundred to 1,000.

After the war, Hirohito’s adviser, Lord Privy Seal Koichi Kido, stated that the emperor never dreamed of fleeing “to commit suicide in a cave” while the people fought and died to repel the invaders. The construction of the facilities around Matsushiro and the resources put at the builders’ disposal, however, strongly suggest otherwise. Evidence also suggests that Hirohito decided on July 31, 1945, to move to Matsushiro—and to bring along his three essential items of regalia.

As workers labored in the hills around Matsushiro, the future occupants of the bunkers waited in existing facilities in and around Tokyo. Bunkers had been built under the prime minister’s office and below the War Ministry in the Ochigaya area of the capital. A secret underground broadcasting station was set up in case the NHK building was destroyed. At least one of those sites, the War Ministry’s underground headquarters, was tunneled at a depth of more than 50 feet beneath the basement of the War Department’s main building. There was also a bunker on the grounds of the Imperial Palace.

Elsewhere, dozens of additional facilities were built or expanded. These complexes often integrated underground command and communications centers, ammunition storage and even weapons factories. The navy constructed several sophisticated underground complexes at Yokosuka, Sasebo and in tunnels under Keio University, south of Tokyo. Yokosuka, located in Honshu, is now a U.S. Navy base, but during the war it served as a Japanese naval air base. Between 1938 and 1945, more than 20 separate tunnel networks were built there. Underground facilities at Yokosuka included a hospital, electrical power-generating facilities and miles of tunnels for the assembly and manufacture of parts for Japanese naval aircraft. Facilities at the largest air base in Japan, at Atsugi, were to include an underground repair plant, generator and extensive quarters for personnel.

Some components of the underground naval facilities were remarkably well designed and constructed. After the war, inspectors gathering information for the USSBS found the deep-underground ammunition magazines at the Sasebo base in southern Kyushu to be exceptionally sturdy. The most outstanding feature was an inner tunnel structure, which helped to control moisture and maintain temperature. Sasebo also housed a deep-underground command, control and communications center. Allied engineers who inspected the Japanese headquarters beneath Keio found a sophisticated complex with offset adits (tunnel openings) to prevent shell or bomb fragments from entering, 10 very well-protected ventilator shafts, an elaborate water supply and sanitation system, redundant electrical power, reserves of water and fuel, and chemical warfare decontamination equipment. These facilities encountered few of the problems that plagued underground industrial plants hastily constructed and run by corporations. The navy’s success suggests that had there been time to do it, Japan may have overcome deficiencies at those plants as well.

Though most of Japan’s strategic underground facilities were built to protect its leaders, command and control, and conventional military industry, it turns out that some were used to conduct horrific biological weapons–related experiments on humans. These tunnel complexes were located in Manchuria but nonetheless were of strategic importance. Japan’s “Epidemic Prevention and Water-Supply Unit,” better known as Unit 731, was established in 1936 to research, develop and manufacture biological weapons. From 1936 until 1945, Unit 731 conducted experiments on live human subjects that resulted in at least 3,000 gruesome deaths.

Underground facilities were built at the main complex near Harbin, known as Ping Fang, and at other sites in Manchuria. At Ping Fang, extensive underground facilities were used for human experimentation, according to Wang Peng, the curator of Unit 731’s Criminal Evidence Museum, which is located at the Ping Fang complex’s former main office. The museum staff is now excavating in search of deep-underground laboratories and associated tunnels. Local villagers contend that when excavating building foundations, they often come across hidden openings to underground shafts. Additional sources confirm that experiments took place in tunnels at Ping Fang. At the Unit 100 biological weapons facility at Changchun, tunnels housed laboratories and allegedly contained bunkers that stored such weapons.

Biological weapons experimentation was the most atrocious component of Japan’s underground program, but the most extensive and impressive was the empire’s effort to bury a significant portion of its military industry. Erroneously believing that the U.S. Army Air Forces would have little impact on production, Japan’s industrial establishment did not begin dispersing its factories on a large scale until the fall of 1944, as the U.S. strategic bombing effort intensified. At that time most aircraft production was concentrated in and around Tokyo, Nagoya and Osaka. But some companies, in anticipation of a government order to move their production to safer locations, had already begun to dismantle their plants. Following the bombing attacks of November and December, an urgent dispersal of the aircraft industry began, without government direction or control. In February 1945, shortly after the imperial ordinance was issued to increase resources for underground construction, the government assumed control of the underground and surface dispersal effort.

By April 1945, the dispersal of industry was taking place on a fairly large scale. The government planned to move 993 manufacturing plants, 172 of them to underground sites. The majority of the plants going below ground were to produce aircraft and aircraft equipment. At least 73 sites were newly excavated tunnel complexes, while the others were either expanded and reconfigured from former mines and caves, or based in abandoned railway, highway and streetcar tunnels.

Declassified U.S. documents, including a study by the former Military Intelligence Service (MIS) published in 1946, show that the Japanese encountered insurmountable obstacles to underground production. The Germans overcame some of the problems that plagued the Japanese, such as drainage, ventilation and lighting. Insufficient time was considered a major factor confronting the Japanese as they began to build underground industrial plants too late in the war.

According to the MIS report, by the time the Japanese realized the necessity of dispersing and moving industry underground, they were engrossed in production commitments and took action only after numerous cities had been bombed. Other problems included inadequate transportation, shortages of materials, poor site selection, insufficient storage space and high humidity—which rusted valuable tools and products. Mitsubishi’s Matsumoto airframe-works plant in Nagano prefecture exemplifies the problems. The decision was made in February 1945 to build an underground plant, and work began on the tunnels in April under the supervision of the army. The original geological site survey concluded that shoring was unnecessary, but in actuality the unstable rock required extensive timber support. Timber was very scarce at that point in the war. Transportation became the most significant problem, first in the movement of machine tools and equipment from Nagoya to Matsumoto and later to transport materials underground. The facility was scheduled to be 50 percent complete by June and fully operational within two months, but by August only 40 percent of the facility was completed. Nakajima’s underground aircraft factory at Shiroyama was similarly unfinished by the end of the war. Manufacturing began there in late April 1945, but only half the planned facility was completed by war’s end. Actual airframe production amounted to only four fuselages and four wing assemblies.

Many underground sites had inadequate rail facilities, and the ones in the hills would have been inaccessible during winter snows. The Mitsubishi Nukatani underground aircraft engine plant was accessible only by foot up a narrow road, which was often washed out in many places by heavy rains. The company initiated road, residence and bridge building projects. Beginning in July 1945, machine tools were laboriously pulled up the mountain road by hand.

A USSBS special report on the underground aircraft industry concluded: “Japanese underground installations were begun too late for them to be able to save the production of aircraft. In any event, their existence could not have overcome other problems such as shortages of vital raw materials and fuel.” The survey did, however, also report that the “profusion of tunnels, caves, and mines is impressive.” One of the most advanced plants visited by inspectors was Nakajima’s Yoshimatsu aircraft engine plant near Matsuyama, southwest of Tokyo. Its tunnels were fairly dry due to the particular quality of the volcanic rock in which they were excavated. In many places concrete floors had been laid. Corrosion became a problem right away and each worker was responsible for keeping his machine from rusting; finished parts were removed from the tunnels immediately. This well-camouflaged facility had two entrances located in sheer cliffs more than 75 feet high.

Despite the haste of building them, many underground sites were cleverly hidden. For example, an entrance to the Mitsubishi underground aircraft parts factory southwest of Nagoya at Hisai was covered by a building and built on fairly flat ground, which proved an unusual and effective means of concealment. This plant, like virtually all underground factories in Japan, went undiscovered by Allied intelligence during the war. The USSBS highlighted this point: “The principal advantages of an underground installation are that it is hard to find, makes a very poor target, and would probably be safe from any weapon used in the Second World War.”

Some information was known by the Allies about the Japanese underground industrial dispersal program, but specific data were scant on the location of such sites. Even less human intelligence was available on the very secret underground leadership facilities and other strategic underground bunkers. Intercepted communications were also silent on the issue of underground facilities.

Aerial photography was scarcely more helpful in bridging the intelligence gap. Most imagery covered cities and did not extend out into the country far enough to cover even one-third of the existing underground industrial plants or other remote facilities. Moreover, photo interpreters were not alerted to look for underground installations. When a Japanese strategic tunneling project was identified on air photos, the significance of the activity was not understood. For instance, a few deep-underground factories under construction were identified on reports as “areas of tunnel activity” or misidentified as storage depots. Only three or four Japanese underground aircraft plants—of a total of 73 newly excavated tunnel sites—were located by photo interpreters during the war, and of those described in published reports, none were correctly identified as aircraft plants.

While there was little human source information on Japan’s strategic underground program, the empire’s use of caves and underground complexes in the Pacific islands should have prompted inquiry into a possible tunneling effort on the Home Islands. The Allies’ relatively successful methods used to detect German underground facilities could have been applied to Japan. In October 1944, nearly nine months after the Germans had begun their underground program, a special underground section was organized at the Allied Central Interpretation Unit (ACIU) at Royal Air Force (RAF) Station Medmenham, England, to detect and characterize underground facilities. The section, known as “B-6,” had a maximum strength of seven photo interpreters, two cartographers and part-time services of a geologist. About 60 underground factories in Germany or German-occupied countries were identified by this unit. Photo interpreters noted a consistent pattern of construction and operation of German underground facilities, which aided in the detection of these sites and their characterization as industrial facilities. Still, 167 German underground factories were not confirmed by photo intelligence during the war.

After the war, U.S. military intelligence was amazed to discover that Japan’s dispersal of aircraft and engine manufacturing plants to underground locations was far more extensive than had been suspected. In an attempt to understand why the United States failed to correctly identify any of these plants, a former B-6 photo interpreter analyzed aerial photos of Japanese underground plants taken during and shortly after the war. He noted that newly excavated Japanese facilities exhibited a pattern strikingly similar to German sites at the same stage of construction. He surmised that if the lessons learned regarding the interpretation of German underground plants had been applied to the Japanese plants, and if a concerted effort had been made to locate and identify such facilities, results could have been much better.

Japanese leaders interviewed after the war revealed that they had some general knowledge of German underground installations but that any similarities between the Japanese and German programs was due to events that forced them both to resort to the only apparent solution: Move the factories where they would be hard to find and difficult to bomb. U.S. analysts noted, “In this, they were largely successful.”

Negligible human intelligence, inadequate aerial coverage and inexperienced photo interpreters were why the Allies failed to detect and characterize Japanese underground facilities. But, according to the USSBS report on the underground construction of Japanese aircraft, it is very difficult to identify such facilities through photo interpretation even when analysts are well trained. Inspectors from the USSBS stated that postwar airphoto interpretation, even when based on information that the USSBS supplied, still could not find most entrances at locations that had been visited by survey inspectors. The report added that it was impossible to determine how extensive an underground plant might be or what activity was being conducted there by airphoto interpretation alone.

As one example, Mitsubishi’s underground aircraft engine production plant at Sabae would have been almost impossible to detect using aerial photography. Inspectors visiting the plant after the war reported: “From the standpoint of aerial reconnaissance, the facility was ‘exceptionally concealed.’ Only a small amount of spoil was visible and the roads were almost completely concealed by trees.” The inspectors noted that the facility would have been comparatively safe from a direct bombing attack. The same went for another Mitsubishi underground aircraft plant, located at Shakutani. The inspector who visited that site wrote, “So skillfully were the semi-underground plants built into the hillside that one building was almost passed during an inspection trip before it was noticed.”

Japan’s underground program was a serious attempt at strategic survival. Tunnel complexes, such as the alternate naval headquarters at Keio, facilities at Sasebo and Yokosuka, Matsushiro and several industrial plants, were well designed and sophisticated. Had the Japanese military leadership been more willing to see that its industry, government and military were at risk from air attack six months sooner, and ordered a serious tunneling program then, the majority of its underground facilities could have been fully operational by August 1945. Even with the late start, though, the program likely contributed to the military leadership’s overconfidence in the regime’s ability to survive.

A different result would have had profound consequences for the Allies. If a sizable number of unidentified underground factories operated during the war, Japanese weapons production would have increased and the Allies would have miscalculated their enemy’s capabilities. Further, if these facilities had been fully operational and occupied, the emperor and his military leaders may have been less likely to consider surrender as their only option to annihilation, in which case the war could have continued even longer and the defense of the underground network could have caused numbers of casualties beyond even the most gruesome estimates of Allied planners.

Apart from the Mitsubishi workers busy in the underground factory at Nagasaki at the time of that blast, the USSBS noted in its report that the 400 or so people who were properly placed in Nagasaki’s tunnel shelters also survived, including those in tunnels nearly directly under the explosion. The shelters consisted of rough tunnels dug horizontally into the sides of hills with crude, earth-filled blast walls protecting the entrances. The tunnels had a capacity of roughly 100,000 people. The USSBS surmised that the loss of life in Nagasaki would have been substantially lower had these tunnel shelters been filled to capacity: “Without question, shelters can protect those who get to them against anything but a direct hit.”

The survivability of deep-underground facilities and their occupants did not go unnoticed by Japanese military commanders. In fact, after the bombing of Hiroshima, Field Marshal Shunroku Hata, who had visited the devastated city, pointed out to General Korechika Anami, then minister of war, that people who were underground seemed to have escaped the effects of radiation. General Anami asked the field marshal to be sure to report that fact to the emperor, as it “might affect His Majesty’s final decision.”

According to a formerly classified U.S. Air Force photo-interpretation manual on underground installations published in 1954, “It was generally agreed that hostilities in both the European and Pacific theaters of operations would have been at least prolonged, with a resultant greater loss of Allied lives, if the Germans and Japanese had completed their underground programs before hostilities commenced, or at least early in the war.”

Japan’s failure in this regard was not repeated by other countries. During the Cold War, the United States and the Soviet Union—among many others—built deep-underground facilities to protect their most valuable assets. The Soviets had a comprehensive underground program to ensure the survival of the leadership and military command and control that also included the protection of military industry. Most notably, the Soviets tunneled into a mountain in Krasnoyarsk to protect three plutonium production reactors. Meanwhile, the United States built similar underground facilities of its own.

Underground facilities are now built in peacetime throughout the world for defense during times of crisis. Their uses range from protecting weapons of mass destruction and their delivery systems to harboring terrorists and leaders in caves and holes—as was found to be the case with Osama bin Laden and Saddam Hussein. Underground facilities, particularly deep ones, give leaders whose military forces are less technically advanced or potent than their adversary’s a sense of confidence and security. Until technology allows us to identify these facilities with ease, underground networks such as those built by the Japanese will remain a feature of warfare well into the future.


Robin Pekelney Shwetzer is a Defense Threat Reduction Agency contractor working for Science Applications International Corporation (SAIC) on hard-target issues. For further reading, see The Campaigns of the Pacific War, by the USSBS.

Originally published in the August 2006 issue of World War II. To subscribe, click here