Should the Saffir-Simpson Scale Be Abandoned?

[brunota2003]

The Houston/Galveston NWS shared this on Facebook tonight.
https://www2.ucar.edu/atmosnews/opinion/7834/hurricane-storm-surge-category-its-own

From that article:

In this spirit, NHC is developing a new product: storm surge warnings (see PDF). To be introduced in 2014 or 2015, according to Rhome, these could bridge the gap between current practice and the unique issues related to high water. Storm surge warnings wouldn’t replace hurricane warnings but would supplement them. They would be produced through collaboration between the NHC and local NWS offices, in much the same way that tornado watches are now issued. When a given amount of surge above ground level is expected—perhaps 4 feet, though the criterion is still to be finalized—a storm surge warning would go out for those locations.

“The threshold would be consistent across the board,” says Rhome. “Four feet of water above ground is the same everywhere, from Texas to Maine.”

I totally disagree with that last statement. Four feet of surge in areas where the land slopes up rapidly is totally different from surge that is four feet in an area where the ground remains near sea level and is flat. The 4 feet would push further inland, with big waves to boot, than areas where it isn't as flat. In one spot, it might only push in 20 or 30 feet, whereas the other it might push in 300 feet.

If we can create flood zones based off of that exact river and the lay of the land, why can't we produce surge zones for coastal areas and rivers? So when they say "10 foot surge is expected" we can look and go "well, this house will have water in it"

[mitchell]

If we can create flood zones based off of that exact river and the lay of the land, why can't we produce surge zones for coastal areas and rivers? So when they say "10 foot surge is expected" we can look and go "well, this house will have water in it"

I work in floodplain map production. The time and expense to produce a floodplain map would present significant challenges to transition this technology to storm surge mapping. To give you an idea of coast, figure about $1,000 per mile of stream or mile of cost for a large scale flood mapping project. Every road, bridge, dune line, culvert, dam, levee, (on and on) in the inundation area would have to be included in the model to properly map surge. Its expensive and time consuming and needs to be updated as conditions change.

Also, the minute you produce a official government map depicting which properties, and houses would be inundated by a 4, 5..8...10 , 12...etc. foot surge you instantly have a demand by property owners to correct real or percieved errors in the map, or to get the map changed when you change topography. Many subdivisions in low lying areas are bringing in several feet of fill to elevate land to meet local floodplain regulations.

FEMA has several HUGE contracts with consulting engineers to handle the thousands of requests for floodplain map changes. As a practioner, i dont think we're there yet.

[brunota2003]

On the plus side...it'd at least open a lot of jobs up it seems

[vbhoutex]

The Houston/Galveston NWS shared this on Facebook tonight.
https://www2.ucar.edu/atmosnews/opinion/7834/hurricane-storm-surge-category-its-own

From that article:

In this spirit, NHC is developing a new product: storm surge warnings (see PDF). To be introduced in 2014 or 2015, according to Rhome, these could bridge the gap between current practice and the unique issues related to high water. Storm surge warnings wouldn’t replace hurricane warnings but would supplement them. They would be produced through collaboration between the NHC and local NWS offices, in much the same way that tornado watches are now issued. When a given amount of surge above ground level is expected—perhaps 4 feet, though the criterion is still to be finalized—a storm surge warning would go out for those locations.

“The threshold would be consistent across the board,” says Rhome. “Four feet of water above ground is the same everywhere, from Texas to Maine.”

I totally disagree with that last statement. Four feet of surge in areas where the land slopes up rapidly is totally different from surge that is four feet in an area where the ground remains near sea level and is flat. The 4 feet would push further inland, with big waves to boot, than areas where it isn't as flat. In one spot, it might only push in 20 or 30 feet, whereas the other it might push in 300 feet.

If we can create flood zones based off of that exact river and the lay of the land, why can't we produce surge zones for coastal areas and rivers? So when they say "10 foot surge is expected" we can look and go "well, this house will have water in it"

In some areas of LA and TX that 4 feet of surge would literally go inland miles, not just hundreds of feet. During Ike the much higher surge went inland as much as 17 miles in some areas. Of course some of these areas are uninhabited swamp land. I understand your disagreement with Rhome's statement, but I think "technically" it is correct in that 4 feet of surge as they are describing it(above ground) would be the same everywhere, however the effects and the run in would be VASTLY DIFFERENT in some cases.

[TropicalAnalystwx13]

I do not think it should. The Saffir-Simpson Hurricane WIND Scale is for wind only, if you could not tell be its name. I suggest we make a totally different scale for storm surge and rainfall, if we even need one.

[wxman57]

The Houston/Galveston NWS shared this on Facebook tonight.
https://www2.ucar.edu/atmosnews/opinion/7834/hurricane-storm-surge-category-its-own

From that article:

In this spirit, NHC is developing a new product: storm surge warnings (see PDF). To be introduced in 2014 or 2015, according to Rhome, these could bridge the gap between current practice and the unique issues related to high water. Storm surge warnings wouldn’t replace hurricane warnings but would supplement them. They would be produced through collaboration between the NHC and local NWS offices, in much the same way that tornado watches are now issued. When a given amount of surge above ground level is expected—perhaps 4 feet, though the criterion is still to be finalized—a storm surge warning would go out for those locations.

“The threshold would be consistent across the board,” says Rhome. “Four feet of water above ground is the same everywhere, from Texas to Maine.”

I totally disagree with that last statement. Four feet of surge in areas where the land slopes up rapidly is totally different from surge that is four feet in an area where the ground remains near sea level and is flat. The 4 feet would push further inland, with big waves to boot, than areas where it isn't as flat. In one spot, it might only push in 20 or 30 feet, whereas the other it might push in 300 feet.

If we can create flood zones based off of that exact river and the lay of the land, why can't we produce surge zones for coastal areas and rivers? So when they say "10 foot surge is expected" we can look and go "well, this house will have water in it"

There are a few problems with issuing surge warnings, and that's why it's taking years to come up with a plan. What the NHC is talking about is issuing a warning to a specific location for a storm surge that could produce water "X" feet deep. The slope of the coastline would determine the inland extent of the warning. Graphics would certainly help the general public to visualize the extent of water coverage.

Another issue is that these surge warnings may be issued 24-36 hours ahead of any tropical storm or hurricane warning, as setup tides ahead of a storm can cut off evacuations a day or more ahead of the arrival of winds, as was the case with Ike in 2008 when the Bolivar Peninsula went underwater about 30 hours before the arrival of TS winds, stranding hundreds of people.

[MGC]

The problem with the current warning system......During Katrina everyone around here was focused on the winds....I can understand that. When you have a hurricane with 175mph wind bearing down on you kinda can't really digest all the threats you are facing. The surge forecast was burried in the bottom third of the public advisory. The NHC forecast up to 28 feet in places and they were pretty accurate. Honestly, I kinda quit reading the advisory taht Sunday Morning about a third of the way through it....i was numb just thinking what those kind of winds were capable....visions of Homestead raced through my head. Nothing like waking up to a Cat-5 that will start hitting later that afternoon. Sadly, it was the surge and the resulting flooding that killed nearly all the 1800 or so people that died from Katrina. For this area surge is the primary threat from a tropical cyclone. Warnings should be customized for a particular area......MGC

[brunota2003]

Water kills the majority of people regardless of where the TC hits. Inland fresh water flooding is the #1 killer in TCs, followed by surge (if I remember correctly).

[Laplacian]

There are a few problems with issuing surge warnings, and that's why it's taking years to come up with a plan. What the NHC is talking about is issuing a warning to a specific location for a storm surge that could produce water "X" feet deep. The slope of the coastline would determine the inland extent of the warning. Graphics would certainly help the general public to visualize the extent of water coverage.

Another issue is that these surge warnings may be issued 24-36 hours ahead of any tropical storm or hurricane warning, as setup tides ahead of a storm can cut off evacuations a day or more ahead of the arrival of winds, as was the case with Ike in 2008 when the Bolivar Peninsula went underwater about 30 hours before the arrival of TS winds, stranding hundreds of people.

If I'm understanding you correctly, the graphics to which you refer are already available for the North Atlantic basin at

http://www.nws.noaa.gov/mdl/psurge/

They were developed by MDL, the Meteorological Development Lab. I recommend clicking on "Archive" just to see how everything works (there are no current Atlantic storms slated for U.S. landfall, so you can only look at past storm data).

[wxman57]

There are a few problems with issuing surge warnings, and that's why it's taking years to come up with a plan. What the NHC is talking about is issuing a warning to a specific location for a storm surge that could produce water "X" feet deep. The slope of the coastline would determine the inland extent of the warning. Graphics would certainly help the general public to visualize the extent of water coverage.

Another issue is that these surge warnings may be issued 24-36 hours ahead of any tropical storm or hurricane warning, as setup tides ahead of a storm can cut off evacuations a day or more ahead of the arrival of winds, as was the case with Ike in 2008 when the Bolivar Peninsula went underwater about 30 hours before the arrival of TS winds, stranding hundreds of people.

If I'm understanding you correctly, the graphics to which you refer are already available for the North Atlantic basin at

http://www.nws.noaa.gov/mdl/psurge/

They were developed by MDL, the Meteorological Development Lab. I recommend clicking on "Archive" just to see how everything works (there are no current Atlantic storms slated for U.S. landfall, so you can only look at past storm data).

I looked at those maps during Isaac found those surge probability maps practically useless. People don't want to know that there could be a 20% chance of a 10ft surge in their area. Doesn't mean anything to them. They want to know what kind of surge to expect at their locations. How deep could the water get if the storm moves inland nearby? When i emailed Jamie Rhome and asked when they might issue some deterministic SLOSH maps, as they said they would 36 hrs prior to landfall, he said they weren't making them anymore and pointed me to those surge probability graphics. I told him they weren't any help to the general public, as it is difficult to decipher what the probabilities mean - too complex. I'm sure I'll see him and others from the MDL at the January AMS meeting and will talk to them about it.

While the SLOSH maps have their problems (highly-dependent on a perfect tack forecast and they're not high-res enough), they convey the potential threat much better as far as the general public are concerned.

[Laplacian]

There are a few problems with issuing surge warnings, and that's why it's taking years to come up with a plan. What the NHC is talking about is issuing a warning to a specific location for a storm surge that could produce water "X" feet deep. The slope of the coastline would determine the inland extent of the warning. Graphics would certainly help the general public to visualize the extent of water coverage.

Another issue is that these surge warnings may be issued 24-36 hours ahead of any tropical storm or hurricane warning, as setup tides ahead of a storm can cut off evacuations a day or more ahead of the arrival of winds, as was the case with Ike in 2008 when the Bolivar Peninsula went underwater about 30 hours before the arrival of TS winds, stranding hundreds of people.

If I'm understanding you correctly, the graphics to which you refer are already available for the North Atlantic basin at

http://www.nws.noaa.gov/mdl/psurge/

They were developed by MDL, the Meteorological Development Lab. I recommend clicking on "Archive" just to see how everything works (there are no current Atlantic storms slated for U.S. landfall, so you can only look at past storm data).

I looked at those maps during Isaac found those surge probability maps practically useless. People don't want to know that there could be a 20% chance of a 10ft surge in their area. Doesn't mean anything to them. They want to know what kind of surge to expect at their locations. How deep could the water get if the storm moves inland nearby? When i emailed Jamie Rhome and asked when they might issue some deterministic SLOSH maps, as they said they would 36 hrs prior to landfall, he said they weren't making them anymore and pointed me to those surge probability graphics. I told him they weren't any help to the general public, as it is difficult to decipher what the probabilities mean - too complex. I'm sure I'll see him and others from the MDL at the January AMS meeting and will talk to them about it.

While the SLOSH maps have their problems (highly-dependent on a perfect tack forecast and they're not high-res enough), they convey the potential threat much better as far as the general public are concerned.

So I see you're a strictly a deterministic forecaster. Are you the kind of forecaster who chooses "the model of the day" (GFS, NEMS-NMM-B, etc)? You apparently take the SLOSH model as the last word deterministic model?? If you do, you and I probably don't have much in common with regard to forecasting strategies. In situations where there's uncertainty, I am a firm believer in probabilistic forecasts.

Thus, I respectfully disagree with what you said. You act as though the SLOSH model is infallible. It is not. I'd much rather have a forecast based on probability on which to base my decisions. I grant you that the general public ALWAYS wants a deterministic forecast. "Tell me yes or no," they say. Sometimes we just shouldn't fall into that trap.

I wrote an article in BAMS arguing for probabilistic medium-range forecasts instead of deterministic forecasts. I got shouted down by the television people and my opinion undermined by the editor of BAMS. Nonetheless, I hold steadfast and continue to believe that we as a professional community fail the general public when we always give them exactly what they want, no matter what the uncertainty of the forecast. And the general public always wants deterministic forecasts.

I won't say anything else on this topic because arguing with people who always think that deterministic forecasts are always the answer, regardless of the forecast uncertainty, really don't want to hear what I have to say.

[Laplacian]

P.S. I'm always puzzled by people who denigrate probabilities in forecast situations such as storm surge, etc. ("they're just too complicated"), but who don't think twice about using probabilities of precipitation (a MOS forecasting tool with some statistical subtleties).

[brunota2003]

Here is another scale, of sorts, I did. It is inspired by both the IKE and HSI scales, and has an "overall" threat assessment. This is more focused toward landfall threats. For out to sea storms, I would recommend just using winds and waves (or not putting anything at all).

Components:
Winds
Surge
Waves
Rainfall
Inland Flooding

Wind:

0 <39
1 40 - 55
2 55 - 70
3 70 - 85
4 85 - 100
5 100 - 115
6 115 - 130
7 130 - 145
8 145 - 160
9 160 - 175
10 175+ mph

Surge:

1 0 - 3 feet
2 3 - 6 feet
3 6 - 9 feet
4 9 - 12 feet
5 12 - 15 feet
6 15 - 18 feet
7 18 - 21 feet
8 21 - 24 feet
9 24 - 27 feet
10 27 + feet

Waves:

0 0 - 5 feet
1 5 - 10
2 10 - 15
3 15 - 20
4 20 - 25
5 25 - 30
6 30 - 35
7 35 - 40
8 40 - 45
9 45 - 50
10 50+ feet

Rainfall:

0 0.0 - 2.5 inches
1 2.5 - 5.0 inches
2 5.0 - 7.5 inches
3 7.5 - 10.0
4 10.0 - 12.5
5 12.5 - 15.0
6 15.0 - 17.5
7 17.5 - 20.0
8 20.0 - 22.5
9 22.5 - 25.0
10 25.0+ inches

Inland Flooding:

1 Low: Nuisance flooding, small streams flooding, poor drainage areas and street flooding may occur.
4 Medium: Smaller rivers may flood, nuisance flooding likely. Some evacuations possible.
7 High: Major flooding possible, many evacuations likely due to flooding.
10 Extremely High: Flooding of epic proportions likely (record levels).

Overall threat level:
Add numbers together, divide by 5.

Examples:

Hurricane Isabel:
September 17 5 pm EDT advisory
Winds: 105 mph (5)
Surge: 7 to 11 feet (4)
Waves: 25 feet (4)*
Rainfall: 6 to 10 inches (3)
Inland Flooding: Medium (4)

Overall threat assessment:
On a scale of 0 to 10: 4.0

*No wave size is in the TCR or reviews by NWS Newport/Morehead City, so wave size is a guess.

Hurricane Katrina:
August 28 5 pm EDT advisory
Winds: 165 mph (9)
Surge: 18 to 22 feet, locally as high as 28 feet (8)
Waves: 40 feet (7)*
Rainfall: 5 to 10 inches with higher amounts possible (3)
Inland Flooding: Medium (4)

Overall threat assessment:
On a scale of 0 to 10: 6.2

*Significant Wave size out in the Gulf, per a NOAA buoy, measured between 30 and 55 feet, according to the TCR.

Hurricane Issac:
August 28 5 pm EDT advisory
Winds: 80 mph (3)
Surge: Max 6 to 12 feet (4)
Waves: 20 feet (4)*
Rainfall: 7 to 14 inches, up to 20 inches possible (5)
Inland Flooding: High (7)

Overall threat assessment:
On a scale of 0 to 10: 4.6

*No wave size is listed in the advisories, and post storm studies are incomplete at this time.
Interesting note: Overall threat assessment for this Category 1 Hurricane is higher than Category 2 Isabel.

Tropical Storm Ernesto:
Winds: 70 mph (3)*
Surge: 3 to 5 feet (2)
Waves: 15 feet (2)**
Rainfall: 4 to 8 inches, up to 12 inches (3)
Inland Flooding: Medium (4)***

Overall threat assessment:
On a scale of 0 to 10: 2.8

*Increased wind threat to 3 due to possibility of slight strengthening prior to landfall.
**No wave data, 15 feet is an estimate.
***Life threatening flash floods listed as possible, no mention of major river flooding though. Decided Medium threat was appropriate.

Hurricane Ike:
Winds: 105 mph (5)
Surge: Up to 20 feet, 25 feet possible (8)*
Waves: 25 feet (4)**
Rainfall: 5 to 10 inches, 15 possible (3)
Inland Flooding: Medium (4)

Overall threat assessment:
On a scale of 0 to 10: 4.8

*Raised risk one level, in effort to increase evacuation efforts. Level is within possible surge range.
**Estimate.

EDIT:
I edited the surge scale. I got rid of the 0. All examples have been updated to reflect this change.

[somethingfunny]

SSHS will probably kill again this week.