Transportation Infrastructure
Florida Keys Hurricane 1935
Railroad Infrastructure
The 1935 hurricane eye passed near Long Key
in the Central Florida Keys.
Railroad viaducts and dikes
(limestone earthen and rubble fill)
held back ebb-current (storm surge return flow),
causing railroad dikes to burst and wash away.
An early type of mainland-type railroad had been built in the Keys
before the 1935 storm,
using concrete viaducts and earthen fill dikes
that were designed for mainland use
(not the conditions in the Keys).
These dikes
(also called fill, embankments or causeways)
consist of rubble piled up, blocking water flow.
The viaducts are a substitute for bridges,
using only compression (instead of tension like bridges)
resulting in shorter spans than bridges,
with wider piers and wider spandrels
than bridges.
The shorter spans, wider piers and lower spandrels of the viaducts
produce very large water-blocking surfaces,
especially at heights above mean sea level
where storm water would be blocked,
almost blocking as much water as the dikes,
essentially a dike with culverts.
The mainland-style dikes do in fact have culverts,
installed after initial construction,
attempting to alleviate the resulting hydraulic head.
Hydraulic head is the water surface height difference of two sides
of a dike or viaduct,
equal to the amount of force of that weight of water.
In the Central Keys, with tide fluctuation of 1 m,
much of each day and night will have observable tidal hydraulic head,
because of the dikes and viaducts slowing down water flow which causes
the water to bunch up.
Storm ebb current is much faster than normal tide currents
(e.g., may be four times faster),
causing a greater percentage of slowing
by the obstructions, thereby producing
more hydralic head (more bunching of the water).
Figure 1:
Railroad dike with concrete culvert in the Central Keys before 1935.
Water level (hydraulic head) is higher on the left than on the right at this tide,
causing constricted water to flow from left to right through the concrete culvert.
The 1935 hurricane caused water to build up to the top of these
dikes, causing enough hydraulic head to wash away some of the dikes.
Figure 2:
Railroad dike washed out by the Labor Day Hurricane of 1935,
at Snake Creek (between Windley Key and Plantation Key).
Make-shift suspended foot bridge spans where the dike washed out.
Person is standing on side of concrete culvert
(like the culvert of preceding photograph).
“Fluid pressure built up on the gulf side of the fills and eventually they failed quickly
and violently.”
—
Coch, in Coastal Hazards, p. 223
These dikes, also called fills or causeways,
are similar to the military-designed pedraplenes of Cuba:
“Cuban tourism authorities have constructed causeways
(or stone embankments) bridging barrier islands
to the mainland and to one another called
pedraplenes (see Map 10.1 in book).
These pedraplenes
block the movement of water in the intracoastal
waters, exacerbating contamination and destroying coastal
and marine habitats… several colonies of flamingos
that used to nest in the Sabana-Camaguey sub-archipelago
have left this area because of the destruction of their habitat
resulting from tourism facilities and
pedraplenes
and settled in the Bahamas”
—
Sergio Diaz-Briquets and Jorge Perez-Lopez,
Conquering Nature,
University of Pittsburgh Press, 2000 (online book),
p. 264, 274
Figure 3:
Military-designed pedraplen in Cuba,
blocking water flow
like the dikes in the Florida Keys.
The next two openings connecting the Atlantic and Gulf
after Snake Creek – Whale Harbor and Indian / Lignumvitae –
had longer dikes, also washed out by the storm ebb current
(storm water return flow).
Figure 4:
Railroad dike in Whale Harbor
washed out by 1935 hurricane.
Figure 5:
Chesapeake Docks at Whale Harbor in 1962,
showing the dike of the previous image was
rebuilt (right background with automobile traffic),
still a barrier to water flow between the Atlantic and Gulf.
Figure 6:
Photograph taken through airplane window,
the day after the 1935 hurricane,
showing ebb current
(storm water return flow)
breaching the Indian / Lignumvitae dike
between Upper and Lower Matecumbe Keys.
Florida Bay is in background,
Atlantic Ocean in foreground.
Figure 7:
Dike breached by 1935 storm
between Upper and Lower Matecumbe Keys.
The next opening
connecting the Atlantic and Gulf
was previously open water
more than 6 kilometers (km) wide
between Lower Matecumbe Key
and Long Key,
but which had been (and still is)
entirely blocked off with dikes and concrete viaducts
(Coch, figure 8.6).
And after Long Key, there are
many more kilometers of dikes and viaducts,
also blocking every possible opening,
all still in place to block water release from
future hurricanes.
Figure 8:
Train on a viaduct in the Central Keys before 1935.
The dikes and viaducts fill the horizon from the Gulf side to the Ocean.
Figure 9:
Concrete viaduct and earthen fill dike in the Central Keys before 1935.
All of the dikes and viaducts are still in place blocking water flow.
Figure 10:
Long Key Viaduct, SW of Long Key, 1926,
reducing water current velocity even without a storm.
Water level during 1935 hurricane reached the top of these viaducts.
Most of the water was blocked by the viaduct from returning to the ocean,
being forced instead to concentrate backwards (toward Cape Sable).
This and all other viaducts continue to retard water flow.
Figure 11:
Closeup of preceding photograph
showing water turbulence and hydraulic head
caused by the viaduct
(reducing water current velocity).
Figure 12:
Another train on Long Key Viaduct in 1926.
Notice again, water bunching up on one side
of the viaduct without even a storm.
Figure 13:
Another example of viaduct-induced
turbulence causing hydraulic head.
Figure 14:
Closeup of preceding photograph
showing wave reflection causing less
bunching up of water on the wave reflection surface
(bridge pier),
while actually causing additional bunching up of the
overall water flow.
Water bouncing back increases the blocking effect.
The faster the water flow, the additional blocking effect,
making the blocking a greater percentage than the pier width,
due to additional bounce-back waves of faster flow.
This still happens (see newer photograph of this effect
in Coch, “Anthropogenic Amplification of Storm Surge Damage
in the 1935 ‘Labor Day’ Hurricane”,
Coastal Hazards,
Fig. 8.4,
p. 216).
Note: The blocking effect is greater than Coch estimates,
because that estimate does not account for the bounce back waves.
Even with lower estimates, water blockage will be very high.
Figure 15:
More recent photograph of Long Key Viaduct.
Pedestrians are walking on the viaduct, to the right
of a two-lane bridge that was built in 1982.
Gulf of Mexico is on the left, Atlantic Ocean on right.
Poles in upper-right are electric utility poles.
The 1982 Long Key bridge, newer than the viaduct,
has much less environmental impact
than the viaduct.
Notice, in the photograph above,
tidal sedimentation
(lighter color) stripes
on the gulf and ocean floor in line
with the viaduct piers and openings.
Tidal flow retardation
is creating sedimentation stripes.
These stripes are occuring on both sides
of the viaducts,
because the tide changes direction every day.
The following image shows sedimentation stripes on the
Atlantic side of that viaduct:
Figure 16:
Satellite imagery (early 2012) shows
Long Key Viaduct (top)
causing sedimentation stripes
in the Atlantic Ocean (Hawk Channel).
[
USGS]
The stripes are testament to the laminar flow
tendancy of the region, newly restricted by the viaducts
and dikes.
