Review Article
Historical Review of Research Activities toward Typhoons/Hurricanes Modification in Japan and the United States
2025 Volume 103 Issue 3 Pages 305-320
Details
2025 Volume 103 Issue 3 Pages 305-320
This study summarizes the discussions on typhoons or hurricanes modification in Japan and the United States (US) from the 1940s to the present, based on a survey of past literature and interviews with relevant personnel. Research on hurricane modification began approximately 80 years ago with Project Cirrus (1947–1952) and Project Stormfury (1962–1983) run by the US government. This project was initiated following a US proposal to Japan in 1965, which aimed to conduct field experiments using cloud seeding techniques for typhoons over the western North Pacific. The proposal sparked in-depth discussions in both academia and the National Diet of Japan. In 1971, the typhoon committee conditionally approved the field experiment in the western North Pacific, but ultimately, the typhoon field experiment was not conducted. This paper identifies the factors that led to the decision not to proceed with the typhoon field experiment despite significant progress, as well as the reasons underlying the decline of typhoon modification and general weather modification research in Japan from that period onward.
In 2022, “Research on Typhoon Control for a Safe and Prosperous Society” was launched as a core research project under Moonshot Goal 8 of the Japan Science and Technology Agency. In this project, methods and technologies for so-called typhoon control or modification (hereinafter typhoon modification), which involves the reduction of typhoon intensity through human intervention, will be developed and examined for their social implementation.
Figure 1 summarizes the history of typhoon modification research in Japan and the United States (US). Research focused on typhoon modification has been conducted since the latter half of the 20th century, beginning in the 1940s and continuing to the present. In the US, this research was part of the “Project Stormfury (1962–1983),” which aimed to modify hurricanes as stated in the objectives of the project; in Japan, the aim was “artificial control” or “artificial modulation” of typhoons. In 1965, when the US proposed conducting field experiments for cloud seeding for typhoons over the western North Pacific, Japan (led by the National Research Center for Earth Science and Disaster Prevention) extensively discussed the idea in academic circles and in the National Diet. At the session of the Economic Commission for Asia and the Far East (ECAFE) and World Meteorological Organization (WMO) Typhoon Committee in 1971, Japan formally expressed its support for the project with the condition of obtaining consent from the affected countries. However, bilateral negotiations between the US and the affected countries did not reach a satisfactory conclusion; thus, the US proposal for the typhoon field experiments was withdrawn. Project Stormfury triggered an increase in weather modification research in Japan, including typhoon modification, but interest in this field rapidly declined after 1971.
Through a survey of past literature and interviews with those involved, this paper first summarizes how exchanges related to typhoon modification research occurred in the US and Japan beginning in the 1940s, as well as how the US decided to propose field experiments for seeding typhoons over the western North Pacific. Then, it discusses why the US proposal for typhoon modification field experiments—which had shown significant progress for some time—was ultimately withdrawn; it also discusses factors that led to the decline of typhoon and weather modification research in Japan.
The concept of hurricane modification arose as a byproduct of wartime research that began in the 1940s. Schaefer and Langmuir, at the request of the Army Air Corps in 1944, began research that aimed to prevent icing on aircraft (Corbridge and Moses 1968; Havens 1952). In July 1946, while working in a cloud chamber, Schaefer and Langmuir accidentally discovered that ice crystals formed instantaneously when dry ice was placed in a cloud of supercooled water droplets (Havens 1952; Murakami 2015). This observation formed the basis of the cloud seeding concept.
Trends related to typhoons and hurricanes modification research in (blue) Japan and (green) the US in the second half of the 20th century.
On 13 November 1946, the first outdoor experiment of dry ice seeding into supercooled clouds was conducted (Havens 1952; Langmuir et al. 1948; Takahashi 1969). Schaefer and General Electric test pilot Talbot used a small aircraft to seed 13 kg dry ice in a supercooled cloud extending approximately 30 n mi (48 km) over Greylock Mountain, Massachusetts. A few minutes after seeding, the supercooled cloud became a snow cloud (Isono 1960).
Vonnegut joined the supercooling study in the fall of 1945 (Havens 1952) and tested three substances with crystal structures resembling ice: lead iodide, antimony, and silver iodide. Through experiments in which the substances were dropped into a cloud chamber, silver iodide was identified as a very effective seeding substance.
In 1947, the Project Cirrus—a cloud physics collaboration between the US Army Signal Corps and the Office of Naval Research in consultation with General Electric—was initiated with the goal of obtaining a more complete understanding of physical processes in the atmosphere, particularly cloud formation, growth, and extinction (Bulletin American Meteorological Society 1950; Corbridge and Moses 1968; Langmuir 1948).
On 13 October 1947, the first field experiment of seeding was conducted on a hurricane, code named “King” by the Naval Hurricane Office (Havens 1952; Langmuir 1948; Tsuchiya 1970). Approximately 86 kg of dry ice were seeded near the center of the hurricane from an altitude of 8,000 m. The minimum central pressure of the hurricane was 973.9 hPa and the maximum instantaneous wind speed was ~ 43 m s−1 (Sumner 1947). Immediately after seeding, Hurricane King seemed to partially weaken, but it developed again the next day. The hurricane made landfall in Georgia and South Carolina, causing 2 million U.S. dollars in damage. The public claimed that the damage in Georgia had been caused by a seeding experiment, and a lawsuit was filed. However, meteorologists pointed out that hurricanes in 1906 and a week ago followed nearly the same path (Suburbs and Disaster Prevention Editorial Committee 1966), indicating that the path of Hurricane King had begun to change before seeding was performed. Because of the lawsuit, Project Cirrus discontinued seeding experiments. The Project did carry out further non-hurricane seeding after Hurricane King.
Despite the concerns about hurricane seeding raised by Project Cirrus, hurricane modification studies were not terminated because six hurricanes—Carol, Edna, and Hazel in 1954 and Connie, Diane, and Ione in 1955—caused extensive damage in the US (Willoughby et al. 1985). After the 1955 hurricane season, the US Weather Bureau launched the National Hurricane Research Project (NHRP). The purpose of the program was to study the structure and dynamics of hurricanes and to research ways to regulate them and to improve forecasting. During its first 3 years, NHRP, conducted hurricane observations using Air Force aircraft; these investigations revealed a large amount of super-cooled liquid water within hurricanes (Simpson et al. 1962). Within 6 years of establishment, the NHRP had its own aircraft (Willoughby et al. 1985).
In 1961, for the second time in the world, the field experiment of seeding against hurricane was conducted on Hurricane Esther, a typical mature and severe hurricane with a minimum central pressure of 927 hPa and maximum wind speeds of over 65 m s−1. The purpose of the experiment was to seed the hurricane eye wall cloud to produce negative absolute vorticity and circulation instability (Gentry 1970a; Simpson et al. 1962).
The project was conducted over a total of 2 days, 16 and 17 September 1961, with silver iodide seeding in the wall cloud on the 16th and outside of the wall cloud on the 17th; the seeding on the 17th did not reduce the maximum wind speed, but the seeding on the 16th was found to reduce the maximum wind speed by ~ 10 % over a 2-hour period (Sheets 1981; Simpson et al. 1962). This decrease in wind speed corresponded to outward diffusion of kinetic energy (Fig. 2). Several additional experiments were proposed to see if results from the 16th were caused by seeding or by natural variations in the structure and intensity of the hurricane. It was also requested that silver iodide be continuously applied during the additional experiments. After 2 days of experimentation, there was an obvious change in intensity due to seeding, and the experiment was considered a success.
Distribution of mean values of kinetic energy before and after seeding and distance from the hurricane center in the experiment on Hurricane Esther on 16 September. Right: altitude 20,000 ft (~ 6,100 m); left: altitude, 7,000 ft (~ 2,130 m) (Simpson et al. 1962).
On 30 July 1962, following the successful seeding experiment on Hurricane Esther, the US Navy and the Department of Commerce initiated the Project Stormfury, a joint effort led by Simpson and Gentry, who had coordinated the Hurricane Esther experiment (U.S. Department of Commerce/Environmental Science Services Administration 1967). There were three main reasons for public support of Project Stormfury at that time (Gentry 1969, 1970a). First, advances in hurricane research had improved the understanding of hurricanes and their developmental mechanisms, making modification methods seem realistic; there was a belief that aircraft observations, combined with seeding, would enhance understanding of hurricane mechanisms. Second, the amount of hurricane-related damage in the US was rapidly increasing. Including the six hurricanes of 1954 and 1955, the average annual damage from 1960 to 1969 jumped to 432 million USD, a 650 % increase in less than 50 years. Finally, the perceived benefits of seeding were thought to greatly outweigh the costs of developing and implementing the technology. A 10 % reduction in damage from a single hurricane comparable to Hurricane Betsy in 1965, which caused 1.4 billion USD in damage, could yield benefits exceeding 1,000 % of the project’s cost after 10 years of ongoing hurricane research.
However, there were few actual opportunities for field experiments during Project Stormfury. Only three experiments were conducted, involving Hurricane Beulah (1963), Hurricane Debbie (1969), and Hurricane Ginger (1971). Figure 3 shows the tracks and seeded locations of these three hurricanes, along with Hurricane Esther (1961).
Tracks of Hurricane Esther (1961), Hurricane Beulah (1963), Hurricane Debbie (1969), and Hurricane Ginger (1971) and seeded locations (Sheets 1981).
In the field experiment for Hurricane Beulah, a total of approximately 455 kg of silver iodide were seeded on 23 and 24 August 1963 (Dunn et al. 1963; Sheets 1981; Simpson and Malkus 1964). The minimum central pressure and maximum wind speed of Hurricane Beulah were 958 hPa and 54 m s−1, before seeding (Dunn 1963). The minimum central pressure and maximum wind speed did not change significantly after the first seeding, but after the second seeding, the minimum central pressure decreased by 15 hPa and the maximum wind speed by over 30 kt (15.4 m s−1). These differences were attributed to the fact that the first silver iodide seeding procedure was performed in a nearly cloudless area (Simpson and Malkus 1964).
In the field experiment for Hurricane Debbie, seeding was conducted five times using approximately 371 kg of silver iodide over 2 days, on 18 and 20 August 1969 (Gentry 1970b; U.S. Department of Commerce and National Oceanic and Atmospheric Administration 1977). The minimum central pressure of Hurricane Debbie was 950 hPa, and the maximum wind speed was 57 m s−1 before seeding (Chafee et al. 1970; Simpson et al. 1970). Figure 4 shows the wind speed distribution before and after seedings. After seeding, the wind strength varied, but on average, the wind speed decreased from just after the second seeding to 5 hours or 6 hours after the fifth seeding (Gentry 1970b). The wind speed was recorded at 43 m s−1 on 20 August 1969 (Chafee et al. 1970).
Change in wind speed when Hurricane Debbie was seeded (Chafee et al. 1970).
On 26 September 1971, a seeding experiment was conducted on Hurricane Ginger (Dorst 2007; Wilford 1971). Fourteen aircraft, manned by 250 people, were used for this experiment. Ginger, being a small hurricane with light winds and no strong convection over the entire area, had seeding conducted in the outer rainbands rather than near the eyewall cloud. Postseeding, radar signals and winds showed no objective changes due to seeding. Hurricane Ginger made landfall in North Carolina on 30 September 1971 (Omoto 1971a).
2.3 Criteria for the experiments and proposal to conduct field experiment for typhoons over the western North PacificDuring Project Stormfury, a seeding experiment in Hurricane Betsy in September 1965 was considered; however, it was abandoned because Hurricane Betsy was particularly close to land (Willoughby 1985). Hurricane Betsy eventually made landfall in south Florida and Louisiana; it caused over 80 fatalities and 1.4 billion USD in damage, earning the nickname “Billion Dollar Betsy” (Kiner 2015; National Oceanic and Atmospheric Administration 2015). On 31 August 1965, Hurricane Betsy was forecast to be within Project Stormfury’s experimental area, and a seeding experiment was scheduled for 1 September 1965. In the morning, aircraft departed from Roosevelt Roads Naval Air Station. However, the National Hurricane Center reported that Hurricane Betsy had changed course overnight, with a southward track that would avoid the experimental area. Consequently, the seeding experiment was cancelled, and the mission was altered to a rehearsal. The aircraft flew as originally planned, but no silver iodide was released. Because neither Project Stormfury nor the Weather Bureau had informed the public or the press of the seeding cancellation, many people believed it had been carried out and a link to its unusual path seemed plausible. Although it was later clarified that the seeding had not occurred, doubts lingered regarding its potential impact (National Oceanic and Atmospheric Administration 2015).
In 1967, LaSeur from Florida State University and 4 other members of the Stormfury Advisory Board recommended setting criteria for field experiments (Science 1967). Previously, hurricanes were considered for seeding only if they were located in a specific geographic area between Bermuda and Puerto Rico. An attempt was then made to select storms for experimentation based on hurricane tracks and location predictions. In accordance with this recommendation, Project Stormfury members established the following criteria: a tropical cyclone in the southwestern North Atlantic Ocean, Gulf of Mexico, or Caribbean Sea is considered eligible for seeding as long as there is only a small probability (10 % or less) of the hurricane center coming within 50 n mi (93 km) of a populated land area within 24 hours after seeding (Chafee et al. 1969). The dotted line in Fig. 5 represents the 50 n mi distance from inhabited areas.
Experimental area in the Atlantic (Chafee et al. 1969).
Seeding was not implemented on or near land for two main reasons. First, seeding at sea would allow the hurricane to revert to its natural state before reaching land. Second, hurricanes undergo substantial structural changes when passing over land, complicating the scientific evaluation of seeding effects (Chafee et al. 1969, 1970).
Beginning in 1970, the scope of Project Stormfury was expanded, and the criteria were relaxed to include hurricanes with less than a 10 % chance of coming within 50 n mi of a populated area within 18 hours of the last seeding. The experimental period was extended from the original 1 August, start date and 15 October end date to a new period from late July and the end of October. With these new criteria, it was estimated that an average of two feasible hurricanes could be tested annually.
The US (Project Stormfury members) estimated that using Okinawa and Guam as experimental sites (Fig. 6) and conducting filed experiments against typhoons over the western North Pacific could allow for an average of about 6.0 experiments per year, even under the initial criteria (Black 1971). This frequency would be about three times higher than that in the Atlantic Ocean. Consequently, it was decided to explore conducting field experiments over the western North Pacific and to approach concerned countries, including Japan.
Experimental area in the Pacific (Mallinger 1971).
At the first ECAFE WMO Experts Meeting in Manila, Philippines, in December 1965, the US made this proposal to concerned countries, including Japan. The Japanese representatives were Michio Yanai (Japan Meteorological Agency: JMA) and Akira Mizuno (Embassy of Japan in Thailand). Japan was the only country to oppose the US proposal (Mizuno 1971). Reasons for this opposition included a lack of full theoretical understanding of the experiment; issues of compensation and liability for unexpected damage; concerns about altering typhoon paths and precipitation patterns, which are important for Japan’s water resources; and a belief that conventional engineering methods were sufficient to address flood damage. Although other participating countries, including Taiwan, Hong Kong, South Korea, Laos, the Philippines, Vietnam, and Thailand favored the proposal, Japan’s opposition resulted in the decision not to conduct field experiments against typhoons in the western North Pacific.
Typhoon modification research in Japan began after Typhoon Vera (1959), also known as the Isewan Typhoon. This typhoon made landfall west of Cape Ushio in Wakayama Prefecture on 26 September 1959, causing storm surge and levee breaches in low-lying coastal areas that resulted in 4,697 fatalities nationwide. In response, an extraordinary typhoon science committee meeting was held, where the concept of typhoon modification in Japan was mentioned for the first time in November 1959 (Nakasone 1960); records indicate that the first mention of typhoon modification occurred at this meeting.
The report of the ad hoc typhoon science committee emphasized that “establishing a scientific basis is the first priority for the time being, and furthermore, research on typhoon modification, such as reducing the destructive power of typhoons at sea or altering their direction, should be considered. Promoting research on artificial rainfall and other basic research for typhoon modification is also important.” Article 8, paragraph (2), item (ix) of the Basic Act on Disaster Management, promulgated in November 1961, stipulates that “The State and local governments must particularly endeavor to carry out the following matters: matters on international cooperation with respect to human control of typhoons….”
Although the Japanese government opposed field experiments against typhoons at the 1965 conference, this opposition led to the initiation of full-scale weather modification research in Japan, spearheaded by the National Research Institute for Earth Science and Disaster Prevention (NIED). At that time, research concerning artificial interference with the weather was termed weather modification research, encompassing areas such as typhoon modification, artificial rainfall, artificial snowfall, hail suppression, and fog dissipation.
In 1965, the NIED initiated a 2-year project titled “Research on Weather Modification” (Ozawa et al. 1978), and the First Conference on Weather Modification Research was held in 1967. This conference served as a platform for Japanese researchers in the field to gather, report research findings, and exchange ideas. Participants included researchers from the NIED, the Atmosphere and Ocean Research Institute of the University of Tokyo, Tohoku University, Nagoya University, the JMA, the Meteorological Research Institute, the National Agricultural Technology Research Center, and the Japan Weather Association. The second Meteorological Modification Research Roundtable Meeting was convened the following year.
Beginning in 1968, a 5-year special research project titled “Research on the Prevention of Hailstorms by Artificial Modification of Cumulonimbus Clouds” was begun (Ozawa et al. 1978). Although it was regarded as “a cloud catching story” by the Ministry of Finance, the project received a substantial budget exceeding 2.7 million Japanese yen (Iwabuchi 1988). The primary artificial modification method for cumulonimbus clouds considered was seeding. The conventional methods—ground fuming and aircraft spraying—were hindered by efficiency issues and a lack of suitable aircraft for experiments. Thus, new methods were explored, such as using rockets to spray silver iodide. Considering Japan’s small size, large population, and residential density, the development of an annihilation rocket—which would disappear mid-air after launch without falling to the ground—was deemed necessary. Nissan Motor Co., Ltd. developed such a rocket, nearly completing it, and field seeding experiments were conducted at the Ground Self-Defense Force Hinodedai Training Area, the Gunma Prefectural Asama Farm, and the Ground Self-Defense Force Somagahara Training Area beginning in 1969 (Ozawa et al. 1978). The rocket was small—weighing ~ 2.9 kg and measuring ~ 78 cm in length—and cost 430,000 JPY per unit at that time. However, if production of about ~ 50,000 rockets per year could be sustained for several years, the cost was realistically estimated to be less than 50,000 JPY per unit (Iwabuchi 1988).
3.2 Consideration of safety criteria for the field experiment of typhoons in the western North PacificIn 1969, Norihiko Fukuta, a Japanese researcher at the University of Denver, visited Japan to investigate the state of weather modification research and discuss future directions with Kazuhiko Terada, Yasuo Omoto, and others at the NIED. Upon return to the US, Fukuta proposed organizing a Japan–US scientific cooperation seminar concerning the artificial modification of hurricanes and typhoons. In response, the Center for Science and Technology for Disaster Prevention formed an expert committee to consider this proposal. However, the committee concluded that a seminar on typhoon modification was not feasible, based on the current status of Japanese research and its previous history. The US researchers expressed substantial interest in this seminar and suggested focusing on themes of “artificial modification of cumulus clouds” or “weather regulation”; the seminar was viewed as a step toward future hurricane and typhoon modification efforts. Ultimately, the chosen seminar theme was “Cumulonimbus Modification of Tropical Nature,” and the first Japan–US scientific cooperation seminar was held in 1970.
Japanese attendees included Terada and Omoto from the Disaster Prevention Center, Kitaoka, Fujiwara, and Ono from the Meteorological Research Institute, Magono from Hokkaido University, and Takeda from Kyushu University. The US attendees were Gentry, Fujita, Fukuta, Ooyama, Hawkins, and Mallinger.
The seminar also discussed field experiments involving typhoons over the western North Pacific, with varying opinions throughout. By the end of the seminar, the Japanese attendees expressed their preference for conducting the experiments under conditions that would not affect Far Eastern countries; they agreed to consult with the affected countries to determine appropriate safety criteria upon their return to Japan. At that time, Japan favored experiments in the western North Pacific for three main reasons: the absence of meteorological observation aircraft: Japanese researchers expected the US Project Stormfury members would provide aircrafts to conduct field experiments, reported minimal lasting effects from the US experiment on Hurricane Debbie in 1969, and significant advancements in tropical cyclone theory (Omoto 1971b).
After the seminar, participants sought to establish Japanese hurricane safety criteria similar to criteria utilized in the US. Discussions with various ministry officials were held, and recorded in the minutes of Diet members’ meetings (House of Councillors 1970; House of Representatives 1967, 1971). Terada’s statements included a desire to conduct joint typhoon and hurricane research involving the US and Japan. Despite extensive discussions, Japan’s safety criteria were not established within 1 year.
Additionally, the minutes of House of Representatives (1971) include a discussion on typhoon modification using nuclear power. A committee member inquired about the problematic nature of this approach with respect to peaceful use of nuclear energy. Terada responded that the idea had been considered shortly after the war due to similarities in energy between typhoons and atomic bombs. However, he acknowledged that then-current methods of reducing typhoon force were entirely different and represented a significant advancement over the use of nuclear power. Omoto (1971b) noted that although promises were made at the Japan–US talks in Miami to establish the desired safety criteria and inform Project Stormfury Director Gentry, no conclusion had been reached by October, despite repeated discussions.
At the 4th session of the ECAFE and WMO Typhoon Committee in Tokyo, 4–11 October 1971, members discussed conducting a typhoon modification experiment in the western North Pacific in 1972. In principle, all members welcomed the proposed transfer of Project Stormfury to the western North Pacific in 1972, provided that agreements were made concerning appropriate criteria and restrictive conditions for seeding experiments (Economic Commission for Asia and the Far East and World Meteorological Organization 1971; Mizuno 1971). In particular, the South Korean and Philippine representatives affirmed their full support for the US proposal (Japan ECAFE Association 1971). Although the Typhoon Committee’s minutes indicated agreement in principle, the JMA’s representatives expressed reservations. They supported experiments with certain restrictions to increase typhoon knowledge but emphasized the need for clarity about potential adverse effects before they could confirm agreement (Japan Meteorological Agency 1975).
During the meeting, Japan’s representatives acknowledged the scientific value of the US plan but expressed concerns about effects on public opinion in Japan. They suggested establishing safety criteria in consultation with other countries and conducting experiments only on agreed-upon typhoons (Omoto 1971a). Japan’s concerns included insufficient observational data from previous Project Stormfury experiments, administrative issues that could arise if problems occurred, legal aspects under international law, and differences between hurricanes and typhoons (Mizuno 1971).
The committee noted that the US would take the initiative in approaching the affected countries (Japan, Taiwan, and the Philippines) to seek bilateral agreement on criteria and restrictive conditions (Economic Commission for Asia and the Far East and World Meteorological Organization 1971). All experimental data would be made publicly available to the Typhoon Committee and other affected countries (Inada 1971; Mizuno 1971).
However, a serious issue arose: the US needed to decide on conducting the experiment by 1 January 1972, to secure the US military support due to budget constraints. This urgency led to hurried bilateral negotiations (Mizuno 1971). No records of these negotiations have been found, but a 14 July 1972 Asahi Shimbun article revealed Japan’s opposition to the experiment in 1972. It reported that the project, scheduled to begin in August, had been postponed for at least 1 year due to Japanese reluctance (Asahi Shimbun 1972).
Preliminary evaluations by Japan’s Ministry of Construction and the JMA suggested more risks than benefits. The Ministry’s concerns included potential deviations in typhoon paths if modified, limited effectiveness in reducing typhoon force, water crisis risks if typhoons disappeared, and possible the US motives for experimenting on typhoons instead of their own hurricanes (Fujiwara 1974).
At the 28th ECAFE General Assembly in Bangkok, March 1972, the US announced no field experiments would be conducted in 1972 because of multiple difficulties (Matsumoto 1972). At the 5th session of ECAFE and WMO Typhoon Committee in November 1972, the US representatives reported that aircraft operations could not continue without military support; considering the planned Global Atmospheric Research Program in 1974, experiments were unlikely in 1973 and 1974 (Economic Commission for Asia and the Far East and World Meteorological Organization 1972). After the 1972 hurricane season, Project Stormfury funding was substantially reduced due to Vietnam Warrelated economic conditions (Dorst 2007; Fujiwara 1974).
In early August 1974, the US again proposed field experiments in the western North Pacific. Japanese representatives received the proposal through the US Embassy in Japan. In the 1974 proposal, Okinawa was excluded from the list of candidate experimental sites. This was because Japan had insisted that no experiments be conducted on typhoons with the potential for landfall in Japan, noting that such experiments might alter the paths of typhoons or increase their rainfall, the “side effects” (Yomiuri Shimbun 1974).
At the 7th session of the ECAFE and WMO Typhoon Committee in Manila in October 1974, the US again proposed field experiments in the western North Pacific (Nakamura 1975). The US representatives stated that they would conduct bilateral negotiations with each country individually and report back to the General Assembly for a decision regarding whether or not to conduct experiments (Economic Commission for Asia and the Far East and World Meteorological Organization 1974). The representative from the Philippines responded in a positive manner, mentioning that his country had begun its own 5-year typhoon modification research program as a national project in January 1974. The Japanese representative was cautious, noting that the experiment might change the tracks of typhoons and increase rainfall (Nakamura 1975). Suda, Director of the Meteorological Research Institute of the JMA at the time, said, “The experiment itself is really fascinating, and I would like to try it from the standpoint of progress in typhoon research. However, we cannot assure that there will be no side effects, and if public opinion says to do it, we will do it (Nakamura 1975, 1978). An informal discussion of typhoon modification was held on the evening of the first day of the fall meeting of the Meteorological Society of Japan (MSJ) (29 October 1974); the transfer of experiments to the western North Pacific was discussed among members of the Society. The following opinions of the MSJ members at the meeting below (Abe et al. 1975), “Gentry and some researchers say that wind speeds have weakened, but I doubt it because the way if they wanted to do the experiment is unknown. Why don’t they conduct direct verification such as taking pictures of clouds or collecting ice crystals? This does not seem like a scientific attitude. The JMA officials say that the experiment will not have much effect, but if there is a big effect (e.g., a typhoon making landfall), it would be a big problem. Some the US scholars are concerned about the possible changes in the air supply caused by the modification. We need the principles of independence, democracy, and openness. Japan’s independence is necessary, and since interdisciplinary studies are shared by all humankind, we should not have the attitude that it does not affect our country, so it does not matter which way we go.” The majority of the participants agreed that “the results of experiments conducted in the Atlantic Ocean, which are used as a basis for forecasting phenomena resulting from experiments currently planned in the western North Pacific, such as changes in wind speeds, rainfall intensity, and paths of typhoons in the region, are not necessarily clear, and that large-scale experiments should not be conducted in the western North Pacific” (Isono 1975). Abe and his colleagues also commented, “The authorities of the JMA should reflect greatly on their previous attitude. A meeting like this one should have been held a few years ago. Even now, I think the authorities should write an article in the “Weather” section to solicit the opinions of a wider range of academics.” The western North Pacific experiment was to be conducted at sea within a radius of 1,000 km (540 n mi) centered on Guam, and the conditions were set to “allow only typhoons that do not approach within 500 km (270 n mi) of land within 24 hours of their last seeding”. This condition was stricter than the Atlantic experimental condition of “only hurricanes with a 10 % or less chance of coming within 50 n mi of human habitation within 18 hours of last seeding (1970)” (Nakamura 1975, 1978). The Philippines, South Korea, Hong Kong, and Thailand had positive views of the US proposal, but Japan and China were cautious; at the Economic and Social Commission for Asia and the Pacific (ESCAP) meeting in New Delhi in March 1975, China stated that it could not “agree to the US proposal” (Nakamura 1978). The report of the Typhoon Committee meeting in Bangkok in November 1975 concluded that the field experiments on typhoons over the western North Pacific would not be conducted because bilateral negotiations and other discussions had not reached a satisfactory conclusion (Economic and Social Commission for Asia and the Pacific and World Meteorological Organization 1975). There is no record of any discussion of field experiments in the western North Pacific by the Typhoon Committee after 1976.
Taiwan, one of the three countries engaged in bilateral negotiations with the US, experienced an important change. On 25 October 1971, at the United Nations (UN) General Assembly, the People’s Republic of China (PRC) joined the UN and was recognized as the sole legitimate government of China, replacing the Republic of China (Taiwan) in the UN system and other related international organizations. Consequently, Taiwan lost its seat in the UN General Assembly, its membership in the ECAFE, and its membership in the Typhoon Committee. Details of the bilateral negotiations between Taiwan and the US are unclear, but Taiwan’s loss of membership in the Typhoon Committee likely impacted the outcome.
During the implementation of Project Stormfury, discussions primarily between the US and the Union of Soviet Socialist Republics (USSR) focused on preventing the military use of weather modification technology, especially in the context of the Vietnam War (1954–1975). In Operation Popeye, the US attempted to disrupt North Vietnamese troop movements and suppress missile fire using artificial rainfall in Vietnam, Laos, Thailand, and Cambodia. However, some individuals strongly opposed using artificial rainfall in warfare (Wilford 1972). After these events during the Vietnam War, discussions began in October 1971 to establish a treaty on banning the hostile use of environment-modifying technologies.
In 1972, the US renounced the use of weather modification technology for hostile purposes. In 1973, the US Senate passed a bill prohibiting any environmental or geophysical modification activity as a weapon of war. In August 1975, the US and the Soviet Union tabled a draft treaty of a convention to the Geneva Conference on Disarmament. This treaty was adopted by the UN General Assembly at its 31st session on 10 December 1976, as the Convention on the Prohibition of the Hostile Use of Environmental Modification Technology and was later approved by the Japanese Diet on 4 June 1982.
4.2 Termination of Project StormfuryIn the 1980s, two scientific discoveries (1) the absence of supercooled water droplets and (2) the observed reformation of the eyewall in non-seeding hurricanes led to the rejection of the Stormfury hypothesis and the Stormfury members discontinued additional field experiments.
The proposed Stormfury hypothesis involved artificial stimulation of convection outside the eyewall through seeding with silver iodide to release the heat the supercooled water. The invigorated convection, it was argued, would compete with the original eyewall, lead to reformation of the eyewall at larger radius, and thus, through partial conservation of angular momentum, produce a decrease in the strongest winds (Gentry 1970a, b; Simpson 1962). For seeding to be successful, the clouds must contain supercooled water, but observations made in the 1980s suggested that most hurricanes didn’t have enough supercooled water for the Stormfury hypothesis to work (Black and Hallett 1986; Hallett and Mossop 1974).
In seeding experiments on hurricane Eshter, Beulah, and Debbie, the eyewall was observed to have moved outward, as hypothesized. However, since the 1980s, non-seeding hurricanes have also been observed to naturally form outward eyewalls (Willoughby 1990; Willoughby et al. 1982). This phenomenon suggested that it was impossible to isolate the effects of seeding from natural changes in field experiments and prevented additional field experiments.
A special committee of the US National Academy of Sciences concluded that a more complete understanding of the physical processes of hurricanes was needed before additional corrective experiments could be conducted, and Project Stormfury was terminated in 1983.
According to Ozawa et al. (1978), the typhoon and weather modification research committee had not convened for various reasons since its third meeting on 12 June 1970. Additionally, Ozawa et al. (1978) noted that the committee’s composition underwent extensive changes in the 2 years to 3 years after 1971; therefore, the resulting infrastructure for weather modification research was considerably less advanced compared with the period of discussion concerning field experiments in the western North Pacific. Figure 7 shows that 120 references regarding typhoon modification research were published in Japan from 1947 to 2023. The literature was collected from various sources, including the National Diet Library Online, National Diet Library Search, National Diet Library Digital Collection, National Diet Conference Proceedings Search System, MAISAKU, Asahi Shimbun Cross Research, Yomidas History Museum, and National Archives of Japan Digital Archive. The number of references peaked at 21 in 1971, then decreased to 11 in 1972 and 3 in 1973. Considering that academic articles often closely reflect changes in research priorities, it is evident that typhoon modification research declined concurrently with weather modification research. However, there is no literature or reference material that explicitly describes the reasons for these changes. Consequently, a literature review and interview survey were conducted based on the hypothesis that an adverse event impacting weather modification research in Japan occurred around 1971. Several reasons for the decline in meteorological modification research were identified.
Published literature concerning typhoon modification in Japan (1947–2023), including articles in (blue) academic journals, (red) newspapers, (yellow) magazines, (orange) books, and (green) official documents.
First, according to Omoto (1971a), the media coverage of typhoon modification experiments around 1970 in Japan was problematic. After the 4th session of the Typhoon Committee in October 1971, reports inaccurately indicated that artificial typhoon modification experiments were definitively scheduled for the following year (Yomiuri Shimbun 1971). Newspapers and television often oversimplified the experiment’s nature, describing it as “typhoon destroy,” “silver iodide sprinkling,” or “attack” (Asahi Shimbun 1971a, b; Yomiuri Shimbun 1971). Additionally, Science Fiction Magazine (SF Magazine) suggested that compared to the US’s proactive approach, Japan’s lack of action in typhoon modification was frustrating. The magazine speculated that Japan’s negative attitude toward typhoon modification research would not be conducting field experiments for another 10 years, compared to 4 field experiments in the Atlantic and observations in south Pacific conducted in the US (Kato 1971). Although this perspective could be interpreted as support for seeding experiments in the western North Pacific, the negative representation may have contributed to public opposition.
Second, an accident was caused by an artificial landslide experiment in Kawasaki on 11 November 1971. In that experiment, an attempt to artificially recreate a landslide by dumping large amounts of water on a slope led to an unexpected massive slope failure (Kuronuma 1972; Oishi 1971). This resulted in the deaths of 15 researchers and media personnel present at the site. The incident received widespread domestic and international media coverage (New York Times 1971). Since the accident, research and experiments that involve artificial intervention in nature were feared (Asahi Shimbun 1971c). The accident also had a serious impact on public perception and support for weather modification research.
Third, Terada resigned as director of the NIED, based on his assumption of responsibility for the Kawasaki accident. Terada was a key figure in typhoon modification research; he actively promoted relevant studies and organized the Japan–US Science and Technology Seminar. His leadership in the “Research on Weather Modification (1965–1967)” project had considerably enhanced the investigation, extending the project’s initial duration by 1 year (Ozawa et al. 1978). However, after his resignation, Sugawara, who opposed typhoon modification, became the director. Sugawara’s stance was clarified on 14 July 1972, Asahi Shimbun article, where he indicated disagreement with the typhoon modification approach previously considered under Terada’s direction.
Fourth, media portrayal of the accident in Kawasaki had a negative impact. The accident was reported on television news as an investigation of artificial rainfall-induced cliff collapse, which led the public to believe that artificial rainfall was dangerous. However, the actual experiment solely involved spraying water via pumps; no artificial rainfall techniques were used. Nevertheless, the public likely associated the term “artificial rainfall”—and, by extension, weather modification research—with the accident (Nippon Hoso Kyokai 2021; Yamada and Inokuchi 2022).
Fifth, transition of Project Stormfury to the western North Pacific was cancelled. Japan’s participation in the 1972 western North Pacific experiment would have provided detailed typhoon observation data using the US aircraft. The number of references to typhoon modification research in Japan decreased from 1971 to 1972 substantially (Fig. 7), indicating a rapid decline in typhoon modification research in Japan at the same time the experiment cancelled.
Sixth, the termination of the Special Research Project “Research on Hailstorm Prevention by Artificial Modification of Cumulonimbus Clouds” in 1972, along with the discontinuation of all related research and field experiments, primarily arose from changes in meteorological research trends during that period. Meteorological research was transitioning from aircraft-based observations to satellite-based observations, which represented a significant shift. This considerable shift was driven by the launch of the world’s first meteorological satellite, US TIROS-1 (Television Infrared Observation Satellite Program), in April 1960; Japan’s first geostationary meteorological satellite, Himawari, was launched in 1977, and the initiation of the World Weather Watch project was initiated by the WMO in 1963, promoting a global meteorological satellite observation network (Japan Meteorological Agency 1975). This sixth reason highlights a broad transformation in meteorological research trends, which contributed to the overall decline in meteorological modification research.
This review explored the history of typhoons and hurricanes modification research in Japan and the experiments by the US influenced similar research in Japan, but substantial interest in typhoon modification in Japan arose after Typhoon Vera (1959). Despite initial hesitancy, Japan began to explore weather modification, establishing a foundation for future typhoon research.
The turning point for Japanese typhoon modification research was the 1965 ECAFE and WMO Typhoon Expert Meeting, which contained a US proposal for regional field experiments that stimulated debate in Japan. By 1971, Japanese researchers were more amenable to participation in experiments in the western North Pacific, although discussions about safety criteria remained unresolved.
In the early 1970s, there were conflicting views within Japan about participation in field experiments proposed the US. By 1975, due to the lack of agreement, the US abandoned plans for experiments in the western North Pacific. Concurrently, Japan’s weather modification research, active during the Project Stormfury era, sharply declined due to negative media coverage and shifts in meteorological research focus.
Decades of advances in meteorology and numerical simulations now suggest the scientific feasibility of typhoon modification. Moreover, it is quite likely that society would need to mitigate extreme weather events as an option for disaster prevention methods. The various data show that global warming and other environmental challenges will intensify rainfall and winds in the near future (Intergovernmental Panel on Climate Change 2023).
Based on these circumstances, in 2022, a research project on weather modification, the Moonshot Goal 8, was launched by the Cabinet Office, Government of Japan. The ambitious goal of the project is to implement typhoon mitigation technologies by 2050. Prior to implementation, several field experiments should be conducted to verify feasibility. The target year for typhoon field experiments is 2040. By this time, we need to address the scientific challenges left unsolved since Project Stormfury.
Designing the experiments and developing effective modification methods, require a comprehensive understanding of typhoon dynamics, particularly in the areas of generation mechanisms, intensity change, and structural transformation. Furthermore, to accurately evaluate the effects of intervention relative to natural typhoon behavior in field experiments, it is necessary to develop numerical models capable of accurately reproducing turbulence and cloud formation. Establishing an integrated observation system that combines multiple approaches, such as conventional meteorological satellite observations and direct aircraft observations, is also critical to precisely monitor changes in typhoon intensity.
Moreover, building on the international discussions during Project Stormfury, it is necessary to approach the realization of field experiments from a social perspective, including the establishment of safety criteria. Overall, this review is expected to contribute to the advancement of typhoon modification research.
This paper is a review article based on the recent publications. See the paper and/or contact the corresponding author of the paper for the detailed data availability.
The authors would like to thank Drs. T. Furukawa, T. Hashida, M. Murakami, and M. Yamanaka for valuable talks on the history of weather modification research.. The NIED provided information for this research. This work was supported by JST Moonshot R&D Program (JPMJMS2282-11 and JPMJMS2282-03).
