Katrina's Surge, Part 11

A Weather Underground 16 part series about Hurricane Katrina, by Margie Kieper.

For the remainder of the month, we're traveling the coastline destroyed by Hurricane Katrina's record storm surge. This is something that has never been shown on the news or talked about, either in the overall, or in detail. What you'll be seeing here is what people on the Gulf Coast have been calling the “Invisible Coastline” for almost a year now.

Today we look at the coastline of the cities of Long Beach and Gulfport. From Google Maps, the location of this area along the Gulf of Mexico coastline impacted by Katrina:

Long Beach, Harrison County, MS

Image courtesy of Google Maps

Looking at the above map, we've traveled such a long way since starting at Grand Isle, LA. It's difficult to keep looking at so much destruction, day after day. Incredibly, I've had to try to think of ways to keep such total destruction from being boring. And each day the few images that you see, represent only one of hundreds, or thousands, of other similar images that could be shown for that same area. Once again today we'll see slabs wiped clean, even of plumbing fixtures, with only the details of the floor to show where separate rooms had been. We'll see again piles of debris, two-by-fours from multitudes of houses, looking from afar like a pile of sawdust and toothpicks.

It's difficult to get your mind around the size of just how many square miles were destroyed, and how many lives were affected by each of those square miles. Today I read that there are over 36,000 FEMA trailers in the state of Mississippi. I know there are a lot more families that for one reason or another are not living in those FEMA trailers, so how many people does that make, total?

Unlike all the other locations we have seen thus far, surge did not reach very far inland here; Harrison County has a lot more high ground than Hancock or Jackson, especially close to the shoreline. The limited extent of the surge did save many residences that were inland from flooding. However the surge was high, and did hit a narrow strip of coastline very hard, including most of the critical infrastructure of these cities -- and they experienced the full force of the eastern eyewall and probably the highest winds of the storm, for the Mississippi landfall. For this reason, and because of the quick access inland to the interstate via Hwy 49, Gulfport was the location of choice for most stormchasers.

This Mapquest image shows more detail about the cities. The surge here mainly occurred along the narrow strip of land between W. Railroad Street (and the railroad embankment), down to the shoreline and the coast road, Hwy 90.

Mapquest Long Beach and Gulfport, MS

Image courtesy of Mapquest

Here are the NOAA aerial images, which give a very clear indication of the surge in this area; first, Long Beach, then two of Gulfport:

NOAA aerial Long Beach

Image courtesy of NOAA

NOAA aerial Gulfport

Image courtesy of NOAA

NOAA aerial Gulfport east

Image courtesy of NOAA

Here are a couple photos of the by-now familiar images of debris along Long Beach:

Long Beach debris

Image courtesy of brsilvo

Long Beach debris 2

Image courtesy of brsilvo

And here is a photo that conveys the sense of desolation of being surrounded by so much debris. All along the MS coast today, while major debris cleanup is almost done (and in record time), all the small debris remains, in every patch of grass or dirt, and hanging from trees.

Long Beach desolation

Image courtesy of brsilvo

At the port of Gulfport, hundreds of shipping containers (the kind that fit on railway cars), were scattered about like mah jong pieces:

Long Beach desolation

Image courtesy of katrinadestruction

The only thing left of these Long Beach homes is the different styles of wood flooring selected by each homeowner, still attached to the concrete slab.

Long Beach desolation

Image courtesy of katrinadestruction

All the debris that we've been seeing -- how much is it? From an AP article in mid-January:

Mississippi has cleaned up 27 million cubic yards of debris – about two-thirds of the total, Powell said. He estimated that Katrina left Hancock, Harrison and Jackson counties in Mississippi with more debris than the combined total created by Hurricane Andrew in 1992 and the terrorist attack on the World Trade Center in 2001.

Yesterday, Governor Haley Barbour was quoted in the Sun Herald regarding debris removal:

Mississippi has set records in terms of both volume and time. We cleaned up well over 40 million cubic yards of debris in 10 months, which was about the length of time it took Florida to clean up 23 million cubic yards after Hurricane Andrew, the previous record holder for leaving debris in its wake.

Wind vs. water: Since Long Beach and Gulfport received the highest winds of the Mississippi landfall, today seems like an appropriate time to talk about the issue of wind damage vs. surge damage. Normally this wouldn't be a sensitive issue; it would be a scientific curiosity to break down the damage by how it occurred. There would also be little controversy over whether surge caused damage, or winds. FEMA flood zoning, hand in hand with the approach insurance companies took towards policies and coverage, both before Katrina and after, changed that. My opinion: I don't believe in working backwards from the fine print in insurance policies to try to rewrite established science and fact. But I also think insurance companies should pay for everything -- that they are hiding behind deceptive practices and legalese to get around the job they are supposed to do: provide coverage. What is the point of insurance if it doesn't provide coverage? Enough said about that.

I'll stick to the technical side of things, and explain a bit about the hurricane wind field, winds over water vs. winds over land, and how such a large surge was possible with a weakening storm at landfall.

Before we answer the question of how the surge came to be so large, let's look a little closer at the wind field of a hurricane.

Because the winds in a hurricane rotate around the center in a counterclockwise motion, the winds on the east side of the storm are southerly winds (i.e, moving northward). Since Katrina was also moving northward 10-15mph, that speed can be added onto the speed of the rotating winds on the east side of the storm, in effect making that side of the storm the most powerful in terms of windspeed relative to the ground. Similarly, the western side of the storm is weaker in terms of relative windspeed. This is true generally speaking for all northward-moving hurricanes in the northern hemisphere.

But because of the rotation of the convection around the center, the strongest winds tend to be found in the northeast quadrant of the storm rather than simply to the east of the center.

Now when storms in the GOM approach land that is to the north, the northern part of the storm is no longer over water, and so it cannot obtain the moisture needed to fuel the convection, and the section of the storm that comes off of the land and rotates around the west and to the south tends to have fallen apart a bit, and this is what happened to Katrina. Dry air came into the storm and ate away the convection on that side.

And the strongest winds are only in the area of the eyewall, surrounding the eye, which is not a very wide area. However, even though everybody knows that, there is a tendency to assume that you might have experienced the highest winds even if the eyewall never came close to you.

How are winds measured? How do we know the wind is at a certain category? And what does it mean to be in a specific category? Well, the answer may surprise you. A plane flies around in the hurricane, back and forth, in a large X pattern, making one pass through each quadrant of the hurricane. It may repeat two or three times. And it flies at a safe enough level to avoid the most intense turbulence, which is closer to the surface, and to be high enough to handle the turbulence that they do encounter (which can still be quite a lot). In strong storms this may be about 10,000 feet. And all this time it will be recording the flight-level winds that it finds. Occasionally it will release a "dropsonde," which will record the winds all the way down to the ocean surface.

Now, in all this time, the highest wind speed that it comes across, in the thousands of measurements that are taken on the flight, will be the windspeed used to designate the category of the hurricane. In other words, it is the maximum windspeed that is found. It is the upper bound of what that storm is capable of generating. And since it is the maximum, it is only found once, in one location, at one point in time.

I cannot stress how important this is. Most people take this number to be in some way representative of winds all across the storm. It is not.

And what happens if they don't happen to be in the right place at the right time to measure the maximum wind? Too bad; only actual data can be used to quantify the intensity of the storm, if aircraft recon is being used. For storms beyond the reach of a recon flight, out in the ocean, that will not hit land, satellite estimates, along with other data and the forecaster's experience are used to estimate intensity.

So to give an example for a hypothetical hurricane, if only one time in the hours of flying through the entire hurricane, recon finds a windspeed of 111 mph (96 kt) at the surface either from a dropsonde, or, at flight level, finds a higher wind that can be extrapolated to that same ground speed, using one of several standard formulas, then that storm is labeled a Category 3 with the highest windspeed of 111 mph. Remember that Cat 3 storms have a range of winds from 111-130 mph (96-113 kt).

This won't prevent the news from saying, "Hurricane X roared ashore with 130 mph winds," even if those winds were never measured. But since each category is a range, you can have a weak Cat 3, a strong Cat 3, and one that is a solid Cat 3.

So first of all the general public may be hearing about winds that are higher than any that were actually measured, and then they are thinking more along the lines of this highest measured windspeed as being, well, everywhere, when it is hardly anywhere. And the distinction of whether the eyewall passed over a location -- like your house -- or didn't, is, well, forgotten. Not only that, I have seen, for example, on the weather channel, a meteorologist talk about a storm that is a Cat 3, but then, quickly gloss over that, and say, it “could” be a Cat 4, and then spend a great deal of time talking about Cat 4 and Cat 5 windspeeds.

Not only that, but when winds are extrapolated from flight level, then whether they are reaching the ground level or not is not easy to judge. That is where dropsonde and buoy readings are very valuable. And in some cases the winds at flight level are not making it to the ground level, and unless you happen to have a dropsonde or ground level reading, you don't know for certain.

And not only that, but the structure of a hurricane changes all the time, but especially as it makes landfall, and readings from a hurricane hunter aircraft even an hour earlier may not apply to what is actually happening on the ground as the landfall is occurring. Once even a small part of the storm is not over water, then the dynamics of what make a hurricane to begin with, are disrupted, and this starts to impact the storm immediately.

And not only that, but friction with the ground tends to reduce the sustained windspeed upon landfall. Over water, wind observations are referred to as “marine” obs. But for a location, say, right on the beach, before friction with land can reduce the sustained windspeed, that location may experience winds more characteristic of a marine observation.

So we know that the "max" windspeed is hardly ever occurring in any location, but, in addition, all these factors tend towards reducing the windspeed, not increasing it.

Is there any factor that will increase windspeed? No, but there is a factor that will increase wind gusts. Up until now we have talked only about sustained winds (winds sustained for at least 1 minute). But at landfall, while friction with the ground reduces the windspeed, the turbulence can increase gustiness.

Now there are a couple more things to consider regarding winds. Many smaller gusts can create the same effect as one large gust. There is a good example of this on the web, showing an auto repair shop in a series of images. At first, the large garage doors are easily blown off by the wind, leaving most of the front wall open. Now that the structure is, in effect, a tent, successive gusts blow into the building and pull the remaining structure off the foundation, especially in the front. Once the building has completely detached in the front, it simply blows over, backwards. But the cars, sitting right in front, are undamaged. Afterwards people might speculate it was a tornado or a large gust of wind, when actually it was several smaller wind gusts.

Also, people overestimate windspeeds by quite a bit. Many people think that windspeeds are up to hurricane force (74 mph) when they are just getting to tropical storm force (39 mph). This was true with my brother, who was out in the storm as he was responsible for evacuations on the western side of Jackson County. He thought they had twelve hours of hurricane force winds when they had about twelve hours of tropical storm force winds, with maybe five to six hours of hurricane force winds in the middle. I also initially overestimated the winds by about the same amount, but the more I studied the data and talked to meteorologists, the better I understood the windfield of the hurricane and the windspeeds that were measured during the storm. Josh Wurman notes that 60 mph feels like a lot of wind on a body -- that it feels more like 100 mph. Regardless of how the actual windspeeds were perceived, Katrina's winds were a much longer event than anyone east of the landfall anticipated. Coupled with the surge, which was mostly unexpected, the effect was very overwhelming.

There were no Cat 3 sustained wind measurements for Katrina, but there was failure of anemometers in key areas (and the failure was more commonly a power failure than a failure of the instrument due to high winds). Based on aircraft and dropsonde measurements, NHC determined that the MS landfall had a possible maximum sustained windspeed of 105 kt (121 mph), which would be associated with the eyewall, not everywhere affected by the storm.

So now we can put together a mental picture of the windspeeds of Katrina at landfall. The strongest sustained winds, in the eastern eyewall, would have been felt in Long Beach and Gulfport, and overall the winds would have been felt to the east of the eye. Winds would have been stronger over the water than over land. Here is a radar image from Mobile at the time of Katrina's landfall:

Mobile NWS radar velocity

Image courtesy of NOAA

What about the high unofficial wind gusts measured by the Poplarville and Pascagoula EOCs? These were 117 kt (135 mph) and 108 kts (124 mph), respectively. First of all, gusts are always higher than sustained winds, and roughly about 1.5 times greater than the sustained winds. Applying this range yields approximate sustained winds of 90 mph for Poplarvile EOC, and 83 mph for Pascagoula EOC at the time the gusts were measured – very believable. Pascagoula was about 40 miles to the east of the eyewall at landfall. Poplarville was in the path of the eyewall to the northeast of the landfall location.

The Pascagoula EOC sits right next to a multi-story building, the county courthouse, which happened to be downwind of the EOC for Katrina, and so wind likely was eddied or funneled (Bernoulli effect) off of the larger building, and effective windspeed increased at the anemometer. A bad location, as it turned out, as the wind did cause the EOC roof to peel off in mid-storm, forcing everyone to wade across the street to the courthouse. Also that EOC was very close to the water, both the gulf and the wide river estuary, so it is possible that the wind was more representative of marine obs at that location. Inland, the FIU wind tower data, nine miles inland at Trent Lott airport in Escatawpa, located in an open area, measured a max sustained one minute wind of only 64 kt (74 mph), which is the bottom of the scale of hurricane force winds – 63 kt would be considered a tropical storm, not a hurricane. That location was far enough inland for friction with land to have reduced the higher sustained winds from just offshore.

As far as the Poplarville reading (that EOC also lost its roof by the way), without knowing the specifics of the topography around the anemometer it is hard to say how representative the reading was, but Mobile radar did show a wind maximum in that area to the NE of the eyewall: “…the National Weather Service (NWS) WSR-88D single-Doppler radar in Mobile (KMOB WSR-88D) measured winds as high as 132 mph between 3000 and 4000 feet above ground level during the morning hours. It is estimated that eighty to ninety percent (approximately 104-119 mph) of the latter maximum wind speed value reached the ground.” Image below. Because both EOCs had roof failures, the wind measurements were incomplete.

Mobile NWS radar velocity

Image courtesy of NOAA

Another very good resource for understanding the wind field of a hurricane is HRD AOML, where an analysis is done of all available wind observations, and a map is developed of wind profiles at the surface for both sustained winds and gusts, for landfalling hurricanes, that gives an idea of the wind at any location. Those analyses also indicate weaker winds than was portrayed by the media.

So now, with weaker winds at landfall than Cat 4 or Cat 5, why did the surge correspond to one generated by winds that high?

What are the components of the wind that create the surge?

Remember in an earlier blog entry, the rotating column of water under Katrina's highest winds, at the center, that were pulled up onto the shallow continental shelf and became surge at landfall. This rotating water was formed earlier, when the hurricane was stronger, off the coast. Also, the total amount of rotating water has a lot to do with the size of the eyewall – a larger eyewall will generate more rotating water. This will increase nonlinearly, as a square (area is proportional to the radius squared). This component of the surge was created earlier, when Katrina was a Cat 5 and Cat 4 hurricane.

Also during the weekend, prior to landfall, the strong winds caused what is called “wave setup.” Strong waves generated by the wind were set in motion and traveled along with the hurricane to the north. These waves arrived on the Gulf Coast, and contributed to the total surge as well. An interesting note: two days before Rita made landfall, parts of Jackson County were affected by an overnight surge generated days earlier from wave setup, when Rita was a strong storm over the GOM loop current, to the south. The wave setup with Katrina was also created earlier, when Katrina was a Cat 5 hurricane in the GOM.

Finally, the winds over the water to the east of the eye, moving directly onshore, created wind-driven surge as Katrina made landfall. Katrina had a very large wind field, and this enhanced the surge far to the east of the landfall point. These winds, while over the water, would have ranged from Cat 3 at the eyewall location, down to tropical storm force far to the east. Katrina's envelope of winds increased in area eight times, in the weekend before landfall. It almost filled the eastern two-thirds of the GOM, by Sunday afternoon; Katrina was a huge storm. Katrina had three separate outflow channels, allowing a dramatic increase in both size and intensity between Saturday morning and Sunday evening, while over the deep warm loop current eddy. So, the wind field extended far from the center of the storm.

And so now we can see how the more powerful winds, days before landfall, contributed to the surge, in spite of weakening prior to landfall.

The surge was a combination of these effects caused by wind, but it is not really possible to quantify how much of the surge at any given location was due to each of these factors. The SLOSH model takes all of these factors into account, and is very accurate in determining what “category” of surge will affect any given area of coastline.

Stormchaser Photos and Videos

Here are some links to different stormchasers that filmed Katrina from the Gulfport-Biloxi area:

Ultimate Chase's Mike Theiss and Jim Reed

Scott Blair's chase account

Beau Rivage video Anonymous (clearly not someone local...they can be heard pronouncing Biloxi, bi-LOCKS-ee)

Storm Video


Hurricane Katrina Storm Surge:

Weather Underground Storm Surge Articles

Storm Surge Safety Actions

  • Minimize the distance you must travel to reach a safe location; the further you drive the higher the likelihood of encountering traffic congestion and other problems on the roadways.

  • Select the nearest possible evacuation destination, preferably within your local area, and map out your route. Do not get on the road without a planned route, or a place to go.

  • Choose the home of the closest friend or relative outside a designated evacuation zone and discuss your plan with them before hurricane season.

  • You may also choose a hotel/motel outside of the vulnerable area.

  • If neither of these options is available, consider the closest possible public shelter, preferably within your local area.

  • Use the evacuation routes designated by authorities and, if possible, become familiar with your route by driving it before an evacuation order is issued.

  • Contact your local emergency management office to register or get information regarding anyone in your household whom may require special assistance in order to evacuate.

  • Prepare a separate pet plan, most public shelters do not accept pets.

  • Prepare your home prior to leaving by boarding up doors and windows, securing or moving indoors all yard objects, and turning off all utilities.

  • Before leaving, fill your car with gas and withdraw extra money from the ATM.

  • Take all prescription medicines and special medical items, such as glasses and diapers.

  • If your family evacuation plan includes an RV, boat or trailer, leave early. Do not wait until the evacuation order or exodus is well underway to start your trip.

  • If you live in an evacuation zone and are ordered to evacuate by state or local officials, do so as quickly as possible. Do not wait or delay your departure, to do so will only increase your chances of being stuck in traffic, or even worse, not being able to get out at all.

  • Expect traffic congestion and delays during evacuations. Expect and plan for significantly longer travel times than normal to reach your family's intended destination.

  • Stay tuned to a local radio or television station and listen carefully for any advisories or specific instructions from local officials. Monitor your NOAA Weather Radio.

Source: NOAA

Hurricane Preparedness

National Hurricane Center

Centers for Disease Control & Prevention