The entry says that Hurricane Esther was the pilot, and that Project Stormfury was established on the strength of it. This piece is about how the pilot was made, from the Naval Ordnance Test Station’s own report, written at China Lake in May 1962 by Pierre Saint-Amand, who ran it, and Graeme Henderson of the station’s technical information department.

A smoke

The compound came out of another job. In 1957 Lohr Burkardt, William Finnegan and Rex Smith, all of the station, “were investigating new possibilities for producing a dense, colored smoke that would be visible at high altitudes. During the experiments, it occurred to them that pyrotechnic compositions containing silver iodate would burn to give off finely divided silver iodide. The three inventors patented this concept. No immediate application was found for this process at NOTS.”

In May 1961 Burkardt and Finnegan took it to Saint-Amand, head of the Earth and Planetary Sciences Division, who had been thinking about seeding the storms of the Pacific north-west “with the idea of diverting some of the rain belt farther south to bring more rain to Southern California.” He discussed it with Louis Battan of the University of Arizona, “who felt that the method had great promise, and suggested that the experiment be tried on hurricanes,” and sent him to Robert Simpson at the Weather Bureau in Washington. Saint-Amand “explained the proposal, and burned a small quantity of the compound in Mr. Simpson’s office.” Simpson “was extremely enthusiastic and said that he would like NOTS to perform the experiment in cooperation with the Weather Bureau during the forthcoming hurricane season in the West Indies. He stated that the Weather Bureau had tried seeding hurricanes with silver iodide but they had neither the ability to generate sufficient quantities of silver iodide nor such a method of dissemination as was proposed by NOTS.”

The station’s commander, Captain C. Blenman, decided that the project, “as the project was then named,” Cyclops, “would be made a maximum-effort undertaking in order to meet the deadline of the start of the hurricane season.” A rear admiral at the Bureau of Naval Weapons intervened with the Chief of Naval Operations to expedite the paperwork. Saint-Amand was made project manager. The preparatory work began on 16 June.

Eight weeks

“Only about 8 weeks were available, before the estimated start of the hurricane season, in which to produce and test a satisfactory unit. Therefore, there was no possibility of a sequential program of manufacture.” Six things had to happen at once: a propellant grain, a container, a parachute, a fuze and ignition system, a telemetry set, and an aircraft with a modified bomb rack.

The grain gave the most trouble. The binder the inventors had used was in short supply, so the chemists switched to a castable system, and found that the silver iodate they could buy was too fine a powder to use in it. Burkardt worked out a way to grow the crystals larger by evaporating them from an ammoniacal solution. Then there was the problem of quantity. “The average stock of this chemical held by a commercial house is around 500 grams. NOTS required about 500 pounds.” Three hundred pounds were made to order in Baton Rouge and arrived unrecrystallised. “There was no possibility of having this consignment recrystallized commercially,” so “the Chemistry Division at NOTS jury-rigged a large number of wooden boxes like nursery-plant flats, lined them with polyethylene sheet, and proceeded to grow crystals in every conceivable corner.” The crystals were dried in an oven, ground to 50 to 100 microns, and sent to be made into grains.

The grains burned too fast, so the design was changed from an internal-burning to an end-burning grain, and the flame had to be led through a central tube to the front of the unit. The first ground tests showed the compound hard to ignite, too cool to vaporise all the silver iodide, and unwilling to burn below atmospheric pressure. Three formulations with 5, 10 and 15 per cent aluminium were tried, and 5 per cent worked. The fuze train passed every ground test and then failed on the first live drops at 40,000 feet, where the outside temperature was 65 degrees below zero Fahrenheit. The post-mortem found a gap of an eighth of an inch at the transfer point, the cold, and too little gas in the fuze composition to carry the energy across. Half a gramme of aluminium and potassium perchlorate was added at the transfer point and the gap closed to a thirty-second of an inch; units chilled to 65 below and fired in a pressure chamber at a simulated 40,000 feet then lit, and the next live drops gave seven ignitions in eight.

The container was a 5½-inch flare body with fins from a high-velocity aircraft rocket, fired by a modified Mark 6 fuze pulled by the line of a ribbon parachute packed in the tail; the Irving Air Chute Company of Glendale was asked for sixty special parachutes on 28 June and delivered the first ten on 5 July. The telemetry was Blenman’s idea. A drop over the station’s ranges could be watched by cameras and the naked eye; a drop “in the wall-cloud of a hurricane” could not, so radiosonde units were cut down to fit the nose of each flare, with a transmitter that sent one tone when the unit left the aircraft and a different one when a thermal switch sensed ignition, so that the crew would know what they had done.

The aircraft was a Douglas A3D Skywarrior, a carrier-borne jet bomber, with an adapter built at the station to hold the units in its bomb bay. Fourteen test drops were made at 40,000 feet between 31 July and 12 September. Three on 31 July tested separation and parachutes, two of which did not blossom. Three on 9 August included the two ignition failures that sent the fuze back to the chamber. On 11 and 12 September eight units were dropped, and seven of them burned, for a minute and ten seconds to three minutes. The final tests were finished at about three in the afternoon on 12 September, the station’s research board considered the results, “and it was decided to proceed immediately with the operation on Hurricane Esther, which was then fully formed and being tracked east of Puerto Rico.”

The plan

The party flew to Miami on 13 September, and the main conference convened at ten the next morning at the Fleet Weather Facility, under Commander John Cork, the Navy’s project coordinator. Simpson “announced that it was estimated that on 16 September Hurricane Esther would be located at about 23° N 60° W, and on 17 September, at about 24° N 64½° W; it was his opinion that the estimated position on the two dates would be ideal.”

The scientific object, as the report recorded it, was “that of testing an entirely new seeding method with a new formula to discover whether the process would, by releasing latent heat in the wall-cloud, exert control of the pressure gradient or disturb the thermodynamics to such an extent that an appreciable change could be observed in the form, wind velocity, or track of the hurricane, the principle involved being that of providing freezing nuclei to turn supercooled water into ice, releasing the latent heat of fusion.” Eight Cyclops units would be dropped into the north-east quadrant of the wall cloud from about 2,000 feet above its top, “to see whether this would cut a pie-shaped wedge from this quadrant.” The second day would be planned after the first had been evaluated.

The aircraft were the Skywarrior; a Navy “Willy Victor” Super Constellation of squadron VW-4 as control ship, flying a forty-mile racetrack at 10,000 feet in and out of the eye and making a penetration every hour, with a lieutenant aboard to vector the jet onto its run; a B-57, two DC-6s and a B-26 of the Weather Bureau to observe; and an Air Force U-2 to photograph the eye from 60,000 feet, if it could be spared. At the briefing at Roosevelt Roads on the evening of the 15th, Commander Eason of the station assigned the code words for what the control ship might find in the storm at drop time: X-ray, “conditions as indicated, only minor changes”; Yankee, “definite change, proceed at your own discretion”; Whisky, “major change, conditions dangerous, evacuate area immediately.”

Two drops

T-hour was 2000 Greenwich time on 16 September 1961. “All support aircraft were in position at the planned times. However, it looked at one stage as if the whole operation would have to be canceled because the A3D engines would not start.” The only suitable starter unit kept cutting out. “With only 10 minutes to go before cancellation would have been necessary,” the crew chief, P. C. DiVillaggio of Heavy Attack Squadron 11, “located the pressure-leak in the system. He stopped the leak with his bare hand, sustaining a bad burn in the process, but he succeeded in getting both engines started. The A3D was airborne with less than 4 minutes leeway.”

“Conditions were far from those expected. On arrival in the eye, two factors made the plan impossible. The eye turned out to be only 15 miles across; therefore the A3D could not orbit. Also the wall-cloud was some 50,000 to 60,000 feet high instead of 40,000, far above the ceiling of the A3D. Lt. Shattuck immediately dumped 6,000 pounds of fuel, to gain maximum altitude, and proceeded out of the eye in the front right quadrant. After turning 180 degrees, he was vectored into the eye again by the command ship, and Lt. Baty released the whole load of eight Cyclops units at 5-second intervals.” The planned orbit of the storm afterwards, to see whether anything could be seen, was no longer possible on the fuel left, and the jet went home.

The command ship’s radarscope had been filmed at one frame a minute, and the film was processed as soon as the aircraft landed and examined by everyone present. “It was the unanimous opinion that a drastic change in the formation of the eye of the hurricane had taken place after the drop. As seen on radar, the wall-cloud had turned to ice, starting in the drop area, and spreading to almost 180 degrees of the hurricane eye. In vertical formation, as seen by 10-cm radar, the eye no longer resembled a cylinder but had changed shape so that it now more closely resembled a cup. This condition was maintained for a period of about 1½ hours after the drop, when the eye started to re-form, displaying for a time a figure 9 formation until the eye completely re-formed.” Cork and C. L. Jordan of Florida State University, who had watched it from the air, “were of the opinion that the change was dramatic and unprecedented.” Simpson reported that the Weather Bureau’s B-57, flying at 20,000 feet at constant attitude and power, had recorded a steady climb of 2,200 feet a minute through the north wall, and that the pilots flying above the freezing level “all reported heavy icing and turbulence before the seeding, but essentially none thereafter.”

The second day did not repeat the first. The eye had shrunk from about fifteen miles across to seven, and the wall cloud was only three miles thick. At 1215 local time on 17 September the jet made a single pass through the right forward quadrant and released eight more units on a line tangent to the wall cloud. “During the drop the A3D lost contact with the WV 3, and observers contend that the Cyclops units missed the wall-cloud, falling largely in the eye. One plane orbiting in the eye reported a unit falling nearby. No dramatic effects were noted. This experiment should properly be classed as a null.”

What the station concluded

The report’s summary is confident. “The data from the first day’s operation have been analyzed, and, taken with the visual observations of qualified experts such as Cdr. Cork and Dr. Jordan, it is clear that for a period of about 1½ hours a radical change was produced in the thermodynamics of the hurricane. Both visual and instrumented observations show that over a large segment of the hurricane, about 400 cubic miles of supercooled water clouds were converted into ice clouds, comprising about 20,000,000 tons of ice and releasing the energy equivalent to several atom bombs. Spectacular results might have been obtained if more Cyclops units had been available together with more A3Ds as drop ships to maintain the attack once the initial change in the hurricane had been observed.”

The passage that matters more comes two pages earlier, where the report notes that the Weather Bureau’s preliminary assessment, presented by Simpson to the Washington chapter of the American Meteorological Society on 18 October 1961 under the title “Hurricane Esther: Something Old, Something New,” gave no estimate of error, and that “nor were other parts of the hurricane explored during the time of the seeding and the period immediately following, so that comparison with the undisturbed portions is not possible.” The one witness that might have told the seeded wedge from the rest of the storm had not been called. The station recorded the omission and did not dwell on it.

Its recommendations were practical. The units might “grossly overseed their immediate area”; smaller ones, dispensing about a kilogramme each and carried outside the aircraft, would spread the nuclei better. The Skywarrior would carry only nine of the present units, “and it is also clear that the safety factor for the aircraft under these conditions is definitely marginal. An A3D was never designed to open its bomb-bay doors in a hurricane.” For 1962 the station proposed working from an aircraft carrier, starting off the Cape Verde Islands where the Caribbean hurricanes are spawned and keeping station “in perfect safety about 50 miles to the left rear of a hurricane,” which would also let the Navy “play a more effective role in dealing with typhoons as well as hurricanes.” The Hurricane Hunter aircraft, it added, took months to turn their data into numbers, and should be fitted to record continuously so that the analysis could be done in days.

The smaller units came, as the Stormfury pyrotechnics of the later 1960s. The carrier did not. When the project’s scientists looked back at Esther in 1985 they read the first day’s radar and wind profiles the way they read Beulah’s and Debbie’s: an outer wind maximum had been present before the eight canisters went in, and the storm’s change of shape that afternoon looked like what intense hurricanes do on their own. The report of 1962 had already said what was missing. Nobody had looked at the part of the storm that was not seeded.

Sources

• Pierre Saint-Amand and Graeme W. Henderson, Project Cyclops: An Experiment in Hurricane Modification, NOTS TP 2751, US Naval Ordnance Test Station, China Lake, California, 11 May 1962. All quotations above. archive.org
• H. E. Willoughby, D. P. Jorgensen, R. A. Black and S. L. Rosenthal, “Project STORMFURY: A Scientific Chronicle 1962–1983,” Bulletin of the American Meteorological Society 66 (1985), pp. 505–514, for the storm’s deepening before the drop and the outer wind maximum present before seeding on 16 September 1961. web.archive.org