The storm perceived the Keys as it approached from the southeast, and what it perceived was geometry.
Not the geometry of a coastline — the long, continuous face that the Galveston and Miami entries had presented to the approaching systems, the broad front where the storm’s wind field engaged the land across miles of shoreline simultaneously. That geometry distributed the storm’s energy. It spread the surge across a wide front and allowed the wind field to work against a large area of built environment at once. The leverage the Galveston and Miami storms had achieved came not from concentration but from the coincidence of force against vulnerability across the full width of the target.
The Keys geometry was different.
The target was a corridor. Narrow, low, linear, oriented roughly northeast to southwest, perpendicular to the storm’s northwest approach track. The islands were coral formations barely above the water — the widest not more than half a mile, the narrowest measured in hundreds of yards, their surfaces flat and featureless except for the vegetation and the structures and the railroad embankment that ran the length of the chain like a raised seam in the geography. The water surrounded them on every side. The Atlantic lay to the southeast, open and deep, offering an unobstructed fetch from the direction of the storm’s approach. Florida Bay lay to the northwest, shallow and warm, a broad basin whose water the storm’s circulation would push across the islands from the bay side as the wind rotated through north and northeast ahead of the center’s arrival.
The storm would strike the corridor from two directions simultaneously.
* * *
The approach from the southeast would drive the ocean surge.
The Atlantic fetch — the distance over open water across which the storm’s wind could build the surge — extended for hundreds of miles to the southeast, uninterrupted by land or shallow water that might attenuate the wave energy before it reached the island chain. The surge that the storm’s eye wall generated in this sector would arrive at the Keys’ eastern shore as a combined storm tide — the astronomical tide plus the wind-driven rise in water level plus the wave action on top of both — at heights the Record was calculating against the storm’s central pressure and wind field.
Storm surge would be driven chiefly by three facts: very low pressure, compact violent wind, and shallow water on the Bay side. Earlier Florida entries offered rough comparisons, but none matched this exact combination.
The Labor Day storm was pushing beyond those earlier comparisons. Its pressure was falling into territory no Florida coast had recently recorded, and its small core meant the highest water would be piled into a narrower and more violent zone.
The Record estimated the combined storm tide at the Keys corridor at ten to eighteen feet above normal.
The islands sat at five to eight feet above normal.
The arithmetic was the same arithmetic the Record had performed at Galveston: subtract the grade elevation from the surge height and the remainder is the depth of water that will stand on the land. At Galveston the remainder had been measured in feet. At the Keys it would be measured in the same units, but the meaning was different. At Galveston, the remainder had determined how high the water would reach on the buildings that stood on the island. At the Keys, the remainder would determine by how much the water exceeded the total height of the islands themselves.
The Keys would not be flooded. They would be submerged.
* * *
The railroad embankment added a variable the Record had not previously been asked to measure in this form.
The embankment fills — the raised sections of track bed that connected the bridge crossings, running at approximately eight to ten feet above mean high water — were the highest continuous features on most of the islands they crossed. In ordinary conditions they served their designed function: supporting the railroad track above the tidal variation. In storm conditions, with water rising from both sides of the narrow islands, the embankments would serve a different function.
On the windward side — the Atlantic side, in the approach phase of the storm — the embankment would collect the rising surge. The water, driven across the island surface by the southeast wind, would encounter the embankment as an obstruction and pile against it, the level on the windward side rising higher than it would have risen if the embankment were not there. The embankment was creating a dam effect — artificially elevating the windward water level by blocking its passage across the island to the bay side.
When the water level exceeded the embankment’s height, the dam would overtop.
The overtopping would not produce a gentle flow. The water standing against the windward side of the embankment represented stored potential energy — the height of the impounded water above the embankment’s crest, multiplied by the volume of water behind it, converted to kinetic energy as the water poured over the crest and down the leeward face. The flow over the embankment would be concentrated, fast, and destructive — a weir flow, in the language of hydraulic engineering, that would scour the ground on the leeward side and carry with it everything that was not anchored to the coral foundation.
The camp structures on Lower Matecumbe stood on the leeward side of the railroad embankment.
The Analyst measured embankment height, probable surge, and camp position together. The line built to connect the islands would, under these conditions, concentrate water against the very men it was supposed to protect.
The railroad was not merely the target. It was the amplifier.
* * *
The Florida Bay side added the second component.
As the storm’s center approached from the southeast, its counterclockwise circulation would push Florida Bay water against the western shores of the Keys through the northeast and north winds that preceded the eye wall’s arrival. The bay was shallow — averaging five to six feet deep across most of its basin — and the shallow depth meant that even moderate sustained winds could push the water level up significantly on the leeward shore. The bay surge would arrive at the Keys’ western shores before the ocean surge arrived at the eastern shores, because the bay’s shallowness made it responsive to wind forcing at lower wind speeds than the deep Atlantic required.
The sequence would be: bay surge rising on the west side of the islands as the north winds built ahead of the eye wall, then ocean surge arriving on the east side as the eye wall’s southeast winds reached the Keys, then — after the eye passed and the winds reversed to southwest — ocean surge from the south crossing the islands from the new windward direction.
The islands would receive water from both sides.
The interval between the two surges would be measured in hours rather than days. On islands a quarter mile wide, the surges would not remain on their respective shores. They would cross the islands and meet.
The Observer and the Analyst agreed on the two-sided surge problem. Earlier Florida strikes had driven water mainly from one dominant face. The Keys would take it from both.
The geometry was converging.
The storm was converging.
The corridor waited.