The paper is eight pages long and appeared in the Monthly Weather Review for July 1957, ten years after the storm. Its authors were Conrad P. Mook, Eugene W. Hoover and Robert A. Hoover of the Weather Bureau’s forecast centre at Washington National Airport; by the time it was printed Mook had moved to an Army ordnance laboratory. Its title promises an analysis of the movement of a hurricane off the east coast between 12 and 14 October 1947, and its first paragraph says why one was needed. The storm had crossed southern Florida and gone out to sea near Miami “on a course which seemed to offer no further threat to the North Atlantic coast of the United States. However, on October 13 the hurricane turned rather sharply to the left, and on October 15 it moved inland near Savannah, Ga.” The course it took “as it made its historic turn has been in doubt due to the scarcity of reports received at the time from ships sailing near the storm center.” Several reconnaissance aircraft had been near it on the 13th, “but even their reports did not give a complete enough picture,” so that Tannehill’s standard work on hurricanes showed the path as unknown during the turn.

The paper does not say the word seeding until its second section, and then it says it plainly. “Of particular interest is the fact that three of the aircraft which flew in the vicinity of the storm on October 13, 1947, were a part of a cloud seeding expedition.” Irving Langmuir’s published account placed the storm at the time of seeding, 11:38 to 12:08 Eastern time, at a position the authors reproduced as their figure 1, with the Society’s permission, complete with Langmuir’s dashed circle for the hurricane’s clouds and his dotted track of the hurricane of October 1906. That position, they noticed, “is identical to that shown on the map for 1230 GMT of that date as published in the Northern Hemisphere Sea Level Synoptic series,” a chart drawn for a time five hours earlier. “The implication, which one could draw from this, is that at the time of seeding the storm had stopped moving, which might be interpreted further as an indication that the storm was about to make a change in its course.”

That is the inference on which Langmuir’s whole case rested: a storm that had stopped, and then, within six hours of the dry ice, turned through 120 degrees. The forecasters tested it by asking where the 12:30 position had come from. The answer was that it “hinges almost entirely on a report received by radio from a vessel SS Lisieux. No other reports were listed as having been received from this vessel.” A single radioed observation, from a ship nobody had heard from before or after, had fixed the storm’s position on the official chart, and from that chart Langmuir had drawn his circle.

Finding the ship

“Therefore, before any conclusion could be reached regarding the accuracy of the 1230 GMT hurricane position, it was decided to attempt to verify the report as received by radio.” The National Weather Records Center at Asheville, North Carolina, had no record of it. The Lisieux had belonged to the French Line, the Compagnie Générale Transatlantique, and the forecasters wrote to Paris. “Finally through the cooperation of the French Line, which owned the vessel at the time, the report was tracked down,” from the ship’s present owners, and the company sent the relevant pages of the log, which the paper prints as its Table 1, in French, a cargo-passenger voyage from Galveston to Le Havre.

The log answered the question in one line. At 12:30 Greenwich time on 13 October the Lisieux “had been experiencing a wind from the east, rather than from the west,” so that “the 1230 GMT position of the hurricane, as shown on the Northern Hemisphere map, is in error.” A hurricane’s winds blow anticlockwise round its centre; a ship with the wind from the east is north of the centre, not south of it. The report had been garbled in transmission or in plotting, and the storm had been placed on the wrong side of the one ship that saw it.

Having found one log, they went for the rest. The Bureau’s New York office “made inquiry to many other shipowners which had vessels in the vicinity of the storm,” and visited the shipping lines in the harbour; the acknowledgments name C. J. McGregor and L. E. Truitt, who “devoted many extra hours” to it. Ships’ logs carry an entry every four hours, not only at the synoptic hours when a report is radioed, so with the logs in hand the forecasters could draw charts for midnight and noon and for four in the morning and four in the afternoon of both days, twice as many as anyone had in 1947. Their figure 3 is twelve such charts, from half past midnight Greenwich on the 13th to half past six in the evening on the 14th.

Correcting the barometers

Ships’ barometers are not all right, and the analysis depended on pressures of a millibar or two. So the forecasters did what a careful experimenter does with a dozen uncalibrated instruments: they set aside the noon data for the 13th, drew the other charts, and “by a process of trial and error, a set of average corrections was developed,” a table of offsets for thirty-odd ships, from zero for the Coastal Ranger to minus five millibars for the Santa Margarita, “from which came the consistent and symmetrical picture of the storm circulation shown in the maps.” Then they applied the same corrections to the noon reports, the ones nearest the seeding, which they had kept back. “As a result of the corrections, the hurricane position was found to be somewhat to the west of the position where it was originally believed to have been located (fig. 1), which means that some change in course of the storm was then in progress.”

The two ships nearest the centre gave the same answer in a different way. On the E. R. Scripps, between four and eight in the morning Eastern time, the pressure fell 2.7 millibars at an hour of the day when it should have been rising, as the hurricane came towards the ship on a north-east course; in the next four hours it fell only 1.7, and the wind backed to the south, “suggesting strongly” that the storm “had shifted its course slightly to the north.” On the Lisieux, the wind went from east-south-east to south-south-east between seven and eleven in the morning, “a further indication of a change in the direction of the movement of the hurricane.” Both changes were under way before 11:38.

Where the storm went that night was harder. A Navy flight on the afternoon of the 13th, after the seeding, under Lieutenant Commander N. H. Rudd with Lieutenant E. H. Ryner as aerologist, reached the eye at 2000 Greenwich at a position fixed by Loran, 32.1 degrees north, 75.1 west, and found it elliptical, sixteen miles by twelve, with a glassy sea in the centre and a lowest pressure of 982 millibars; the forecasters gave that report “considerable weight.” For the small loop they then drew in the track overnight, on the strength of an east wind on the Coastal Ranger at 6:30 the next morning, they claimed little. “The data available in the present instance to demonstrate the phenomenon cannot be considered adequate evidence when compared with radar coverage. The loop is presented as one solution to the data available for analysis in this case.” On their map it is labelled “Loop doubtful, see text.” The squall line that Langmuir had made so much of, and that a reconnaissance pilot south of the centre that morning had reported passing through as a “cold front”, they left alone: “No attempt has been made during the present analysis of this storm to show these and other squall lines and lines of convergence which may have been present.”

The storm nobody seeded

The last section of the paper is the one that matters most to the experiment, and it begins with a storm nobody had thought about. “An examination of tropical cyclone tracks indicated that another tropical storm (fig. 4), though not of full hurricane intensity, made a similar turn on a track shown as having occurred on October 6, 1947. This storm moved northward over the Bahama Islands, then curved westward and moved inland in the vicinity of Jacksonville, Fla. This raises the interesting question as to whether a persistence in the broadscale weather regime could have resulted in such a similar recurrence within a week in approximately the same area.” The Bureau’s own season summary had logged it as storm VIII, a tropical disturbance of 6 and 7 October that came ashore near Brunswick, Georgia, with winds of about fifty miles an hour and no damage. It was a week older than the hurricane, a hundred miles to the south, and about a third as strong, and it had done the same thing.

Then they gave the reason. Jerome Namias had shown that abnormal circulations over the southern North Atlantic persist from month to month, and a colleague had shown that the pattern of October 1947 “favored the movement of existing hurricanes into the southeastern United States.” The 500-millibar chart for the morning of the 13th, and the chart of how the heights had changed in the previous twenty-four hours, showed a high building to the north-east, in the storm’s projected path. “These effects contributed to the blocking of any further movement toward the northeast as the hurricane reached the periphery of the High. The existence of the latter was probably the principal reason why the 500-mb. trough in the Mid-West failed to sweep the hurricane into its circulation as it moved eastward.” The chart for 6 October showed “a similar but less intense blocking High toward the northeast in the projected path of the storm.” The conclusion is one sentence. “It is reasonable to suppose, therefore, that the final tracks of both storms were the normal outcome of accompanying anomalies in the general circulation and their reflection in the circulation patterns making up the environmental steering pattern over eastern North America and the adjacent portion of the North Atlantic Ocean.”

The paper does not mention dry ice again after its second section, does not mention Langmuir’s one in a hundred and ten, and does not say that the seeding did nothing. It did not need to. It had done the thing the experiment could not do for itself: it had found where the storm was when the dry ice fell, from an instrument that had been at sea on its way to France, and it had found the unseeded storm of the week before and the high that steered them both. Langmuir died in August 1957, the month after it appeared.

Sources

• Conrad P. Mook, Eugene W. Hoover and Robert A. Hoover, “An Analysis of the Movement of a Hurricane off the East Coast of the United States, October 12–14, 1947,” Monthly Weather Review 85, no. 7, July 1957, pp. 243–250, with Table 1 (the Lisieux’s log), Table 2 (the barometer corrections), figures 3 to 8 and the acknowledgments. archive.org
• H. C. Sumner, “North Atlantic Hurricanes and Tropical Disturbances of 1947,” Monthly Weather Review 75, no. 12, December 1947, for storm VIII of 6–7 October. archive.org
• Irving Langmuir, “The Growth of Particles in Smokes and Clouds and the Production of Snow from Supercooled Clouds,” Proceedings of the American Philosophical Society 92, no. 3, July 1948, pp. 183–185, for the account and figure 14 that the 1957 paper examined. archive.org