Storm Surge Explained: Why the Water, Not the Wind, Is What Actually Kills
When a typhoon is coming, almost every conversation is about wind speed. The category, the gusts, the number in the headline. Wind is what the storm is named for and wind is what the graphics show. But if you look at where people actually die in tropical cyclones, the wind is usually not the answer.
The answer is water. Specifically, it is storm surge, which is the sea being pushed inland ahead of and alongside the storm. It arrives fast, it arrives in places that look nowhere near the ocean on a map, and it is the reason coastal towns get evacuation orders on days when the sky looks unremarkable.
This is a short explanation of what surge actually is, why it hits one town and skips the next, and what it means if you happen to be staying somewhere with a sea view.
What a storm surge actually is
A storm surge is an abnormal rise of water above the normal predicted tide, caused by the storm itself. It is not the same thing as big waves, and it is not rain flooding, although all three can arrive together and make a mess that is hard to tell apart afterwards.
The main mechanism is simply wind pushing water. A tropical cyclone spends hours or days dragging enormous quantities of surface water in the direction it is blowing. When that water reaches a coastline it has nowhere to go but up and inland. There is a second, smaller contribution from the storm's low pressure allowing the sea surface to bulge upward. The rough rule of thumb is about one centimetre of rise for each hectopascal the pressure drops below normal, so a very deep storm at 920 hPa contributes something on the order of a metre from pressure alone. That is not nothing, but set it against the several metres that wind can pile up against a shallow coast and you can see why forecasters do not lead with the barometer. People who track the central pressure number as their main indicator of surge risk are watching the smaller of the two mechanisms.
The distinction that matters practically is between surge and storm tide. Surge is the extra water the storm brings. Storm tide is that surge sitting on top of whatever the tide was already doing. A three metre surge arriving at low tide and a three metre surge arriving at high tide are completely different events, and the difference is not gradual. Timing that a forecaster can predict days ahead determines whether a town is inconvenienced or destroyed.
One more source of confusion is how surge gets reported. Forecasters increasingly give surge as height above ground level rather than height above sea level, because that is the number that tells you whether water will be in your building. Older reporting, and a lot of news writing, still mixes the two. If a forecast says three metres, it is worth knowing which three metres is meant, because the difference between three metres above the sea and three metres above the street you are standing on is the difference between a warning and a catastrophe.
Why one town floods and the next one does not
This is the part that makes surge feel arbitrary from the outside, and it is not arbitrary at all. Four things decide it.
The first is the shape of the sea floor. A coast with a shallow, gently sloping continental shelf lets the water pile up as it comes in, because there is nowhere for it to go downward. A coast where deep water runs close to shore lets much of that energy pass underneath. Two towns the same distance from the same storm can get very different amounts of water for this reason alone.
The second is the shape of the coastline. Funnels are the problem. A bay that narrows as it goes inland concentrates the same volume of water into a smaller and smaller space, and the water level climbs accordingly. Long straight coasts spread the surge out. This is why the same storm can leave one stretch of shoreline wet and put another under several metres.
The third is which side of the storm you are on. In the northern hemisphere the right-hand side of a tropical cyclone's track, relative to the direction it is moving, carries the stronger winds, because the storm's own rotation and its forward motion add together there. On that side the wind is also blowing from the sea toward the land. Stronger wind, pointed the wrong way, over the same shallow shelf. That is the combination that produces the worst surge, and it is why the town to the right of the eye takes the damage while the town to the left reports a windy night.
The fourth is the tide, which is the only one of the four that is precisely predictable years in advance. A storm arriving at high tide on a spring tide is the worst realistic case, and forecasters say so explicitly when it happens.
What it means if you are staying on the coast
For a traveller, surge is the hazard that most often gets misjudged, because the instinct is to look outside and assess the weather.
The most important thing to understand is that surge does not require dramatic conditions where you are standing. The water is being pushed by wind that may be many kilometres offshore, over hours. It is entirely possible to be looking at moderate rain and rising water at the same time. Evacuation orders for coastal zones therefore go out while conditions still look manageable, and the people who ignore them are usually the ones who reason from what they can see out of the window.
Height above sea level matters more than distance from the shore. A hotel two blocks inland but at ground level in a low-lying area can flood while a building directly on the seafront but elevated does not. If you are on a coast during a typhoon and want one number, it is not how far you are from the beach, it is how high your floor is and how high the ground under the building is. Upper floors of a solid modern building are generally where you want to be, which is the opposite of the instinct to get out at ground level and move.
If an evacuation order does come while you are in a hotel, the hotel will usually tell you what to do, and the instruction is often to move upward within the building rather than to leave it. That feels wrong to people who associate evacuation with getting in a vehicle and driving away, but for surge it is frequently the correct call. Roads flood before buildings do, and a car in moving water is far more dangerous than a stairwell. The exception is when authorities order a full area evacuation in advance, hours before conditions turn, and that is an instruction to act on immediately rather than to think about.
It is worth having your passport, phone, charger and any medication together in one bag before the weather arrives, not because you are likely to need to run, but because if you do move floors at short notice you will not be assembling things in the dark. Power cuts are routine in surge events, lifts stop working, and stairwells are not where you want to discover that your charger is still plugged in on the seventh floor.
The other practical point is timing. Surge arrives with the storm and can retreat quickly afterwards, but roads do not become safe simply because the water dropped. Coastal roads get undermined, debris gets deposited, and drainage takes time. The morning after tends to look far more passable than it is, and this is a recurring pattern in how people get hurt after the storm has technically passed.
History keeps making the same point. The deadliest tropical cyclone events on record are overwhelmingly surge events rather than wind events. Typhoon Haiyan in 2013 is the case most people remember, where the water that came into Tacloban was the mechanism for the great majority of the deaths, in a bay whose shape and shallow approach were close to worst case. Korea's Typhoon Maemi in 2003 did the same thing on a smaller scale in Masan, where the surge came up the bay into the city and drowned basement-level spaces. In both cases the wind was severe, and in both cases the wind is not what the death toll came from.
So when a storm is approaching and you are reading the coverage, the useful questions are not about the category. They are these. Which side of the track am I on. What does the coastline do here, does it funnel. What is the tide doing when the storm arrives. And how high above sea level am I actually sitting. Those four answers tell you more about your risk than any wind speed in a headline, and unlike the wind speed, most of them can be answered a day or two in advance.
NOAA storm surge overview and diagrams: https://www.nhc.noaa.gov/surge/
Haiyan at Tacloban, taken apart
The argument above is that the water matters more than the wind, and that the shape of the coast matters more than the category. Both of those can be shown with numbers from the peer-reviewed reconstruction of Typhoon Haiyan's surge in Leyte Gulf, and the result is more extreme than the general argument implies.
Where the water came from
The best-performing simulation put the maximum water level at Tacloban at 5.15 metres, matching watermarks surveyed on land afterwards. The study separates out the contributions.
| Component | Height | Share of the 5.15 m |
|---|---|---|
| Pressure-driven surge, limited to within 50 km of the eye | About 1.1 m | 21% |
| Wave radiation stress over shallow bathymetry near Tacloban | Up to 0.3 m | 6% |
| Wind-driven surge and resonance inside the gulf | About 3.75 m | 73% |
The pressure-driven component is the mechanism that every explanation of storm surge leads with: low pressure at the centre of the storm lets the sea rise beneath it. At Tacloban it accounted for about a fifth of the water.
And the figure that settles the argument
The study found that only about 1.2 per cent of the total water volume entered the gulf from the open Pacific. Nearly all of the water that destroyed Tacloban was already inside Leyte Gulf. The storm did not push an ocean in. It shook a basin.
Specifically, the gulf resonated. Peak oscillations occurred at a period of 4.6 hours, corresponding to the first mode of a 112-kilometre seiche in the gulf. A seiche is a standing wave in an enclosed or semi-enclosed body of water, the same effect as water sloshing in a bath at the one rhythm that makes it build.
Haiyan happened to drive Leyte Gulf at close to its natural period. That is why the same coastline produced a comparable disaster in 1897 from a comparable track, and the paper draws exactly that link: the structure of the surge follows the seafloor bathymetry of Leyte Gulf and the Philippine Trench, which is why two events 116 years apart look so alike.
The practical translation is uncomfortable but useful. The question "how strong is the storm" is secondary to "what is the shape and depth of the water in front of me, and does this track drive it at its own rhythm". The second question is not one a forecast answers, and it does not change from storm to storm. It is a property of the place.
A second lesson: the remembered numbers were wrong
Haiyan's surge is widely remembered at around nine metres, because that is what downtown Tacloban eyewitnesses reported at the time.
| Source of figure | Surge height |
|---|---|
| Eyewitness reports, downtown Tacloban | 8.5 – 9.1 m |
| Early estimates | Up to 7.5 m |
| Numerical analysis | 5 – 6 m |
The eyewitness range runs roughly 1.4 to 1.8 times the modelled maximum. That is not a criticism of anybody who was there; judging water height while it is happening to you is close to impossible. But it matters for how you calibrate your own risk, because the number people carry around is the eyewitness one.
Five metres of fast-moving water in a built-up coastal town is already catastrophic. If you assume the threshold for catastrophe is nine metres, you will dismiss a forecast that should have moved you.
The third lesson, and the strongest: category predicts almost nothing about deaths
Haiyan was a Category 5 equivalent with gusts exceeding 305 km/h at first landfall, the strongest storm ever to make landfall at the time. It killed or left missing about 7,350 people, making it the deadliest Philippine typhoon on record.
The record it broke belonged to Tropical Storm Thelma, in November 1991. Thelma was not a typhoon. It was not even close to one. It killed at least 5,081 people, with roughly 3,500 of those in the single coastal town of Ormoc, through flash floods and landslides, worsened by deforestation and inadequate warnings.
| Haiyan, Nov 2013 | Thelma, Nov 1991 | |
|---|---|---|
| Classification | Category 5 equivalent | Tropical storm |
| Gusts at landfall | Over 305 km/h | Not notable |
| Deaths | About 7,350 dead or missing | At least 5,081 |
| Killing mechanism | Storm surge | Flash flood and landslide |
A tropical storm killed 69 per cent as many people as the strongest storm ever to hit land. Between them, the two deadliest events in Philippine history sit at opposite ends of the intensity scale, and neither was killed by wind.
What to actually watch, in order
| Priority | Question | Why, from the above |
|---|---|---|
| 1 | Is there a bay, gulf or funnel in front of me? | 73% of Haiyan's surge at Tacloban was wind and gulf resonance |
| 2 | How high above sea level am I, in metres? | 5 m was catastrophic; the remembered 9 m was wrong |
| 3 | Has this coast flooded before from a similar track? | Leyte Gulf did it in 1897 and again in 2013 |
| 4 | What is the rainfall forecast upstream of me? | Thelma killed 5,081 with rain, as a tropical storm |
| 5 | What is the category? | Last, and it is already covered by the questions above |
Sources for this section
- Local amplification of storm surge by Super Typhoon Haiyan in Leyte Gulf, PMC4373162 (5.15 m simulated maximum, 1.1 m pressure component, 0.3 m wave radiation stress, 1.2% of volume from offshore, 4.6 h period and the 112 km seiche)
- Repeat Storm Surge Disasters of Typhoon Haiyan and Its 1897 Predecessor in the Philippines, Bulletin of the American Meteorological Society, vol. 97 no. 1
- NOAA Climate.gov, 2013 State of the Climate: Record-breaking Super Typhoon Haiyan (surge estimates, death toll, comparison with Thelma)
- Tropical Storm Thelma (at least 5,081 deaths, Ormoc, deforestation and warning failures)
- Inquirer, IN THE KNOW: Deadliest typhoons in the Philippines
The percentage shares, the 3.75 m residual, the eyewitness-to-model ratio and the 69 per cent comparison are arithmetic on the published figures rather than separately published results.
Comments
Post a Comment