EPAC: OTIS - Aftermath - Discussion

[tolakram]

The NHC Otis archive is here: https://www.nhc.noaa.gov/archive/2023/OTIS.shtml?

Only 4 days of advisories
Hurricane watch posted at 4pm Oct 23rd
Hurricane warning posted 4am Oct 24th
Near major at 2pm Oct 24th
Landfall at 1:25am Oct 25th

Hurricane watches can be posted 48 hours in advance of expected tropical storm conditions. Otis had a watch 33 hours prior to landfall, with a warning posted 21 hours prior to landfall. (approx times). Warnings can be posted 36 hours prior to expected hurricane conditions.

https://oceanservice.noaa.gov/facts/watch-warning.html

I am pessimistic that hurricane models can be improved much until the GFS has a significant overhaul.

[tolakram]

EPS, 12Z Oct 22nd run for 1AM Oct 25th (local)


GEFS, same run time, same local time


GEPS


I don't have access to the high rez, but clearly the Euro did better here, much better. Would that have caught anything more than a tropical storm though? I don't think the euro showed any clue of a major hitting the coast.

[cycloneye]

Wow.

 https://twitter.com/WxNB_/status/1717437000339358089

[kevin]

Death toll 'at least 27' and 4 people remain missing. I'm afraid the true number of casualties will be way higher.

https://www.washingtonpost.com/world/2023/10/26/mexico-hurricane-otis-acapulco/cad62e40-73be-11ee-936d-7a16ee667359_story.html

[cycloneye]

As kevin said, I also fear more casualties.

https://weather.com/news/news/2023-10-2 ... 83103501=1

[cycloneye]

Extensive damage outside the city of Acapulco, in the rural areas.

 https://twitter.com/CycloforumsPR/status/1717595180587426241

[cycloneye]

A big faliure of the models.

 https://twitter.com/yconsor/status/1717610518196703612

[Yellow Evan]

What was not captured by models is the alignment of the LLC and MLC’s within a few hours early on 10/24 once a low level core had developed (despite being exposed at the time) and convection overcame a caping inversion. This set the stage for the infamous RI.

On another note, Mexico needs to give permission for the USAF to station Recon planes much like Barbados.

Mexico also really needs to up the number of radiosonde launches and install Doppler radars throughout. Same goes for a lot of countries in the tropics tbf most of which aren’t quite as developed.

Considering that about10% of homes in Mexico still do not have plumbing or running water, one might conclude that these modern meteorological "necessities" are not high on their yearly budget wish list.

I do recall when frequenting the NHC decades ago, the NOAA recon. folks would get budgeted in some Octobers to station in Acapulco for a couple of weeks. That would happen when the yearly apportioned expenses were not used up by the dearth of Atlantic basin missions in those years.

Maybe but India, Philippines, China are of similar economic status and have radar with velocity and I don’t think giving another country permission to use their airports as a base is something that is monetary dependent.

[NDG]

I believe the lack of data may have something to do with models failing with Otis.
This was the worst case scenario for Acapulco to get hit with a Cat 5 not forecasted by models less than 24 hrs before landfall.
Is going to take a long time for Acapulco to be like it was before.
I also think that the NHC needs to make a better job when a hurricane hits a town with people living on high hills where the winds are even stronger than at sea level.

[cycloneye]

Even Josh was caught by surprise and didn't go to chase.

 https://twitter.com/iCyclone/status/1717032379431202834




 https://twitter.com/iCyclone/status/1717032489506508974

[sikkar]

https://twitter.com/zakkumec/status/171 ... QOVfA&s=19
Speechless.

[Teban54]

 https://twitter.com/WxNB_/status/1717594733592064356

[WAcyclone]

Obs from an AWS in Roqueta Island (107m elevation). At 06:45Z, the station recorded an MSLP of 957.4 mb in the eyewall with sustained winds of 70.6 kt, gusting to 116 kt. The station continuously reported gusts above 100kt from 06:30Z to 07:30Z, with a maximum of 117.2 kt!

https://i.imgur.com/MgFrQD3.png
https://i.imgur.com/f7Vj951.png

That supports the thought the eye went a bit farther east than the satellite angle suggests. The eye probably went right over the city looking at the limited reports I have found, and the pressure gradient was likely razor-sharp.

It looks like the eye may have went west of this station. During the period of minimum pressure, the winds shifted from NE at 6:45 to SE at 7:00 to SSW at 7:15.

[ljmac75]

https://twitter.com/osocom123/status/1717733723687305442

With all these high-rises being gutted, I have to imagine the damage toll will be exceptional.

[mmmmsnouts]

https://twitter.com/osocom123/status/1717733723687305442

With all these high-rises being gutted, I have to imagine the damage toll will be exceptional.

I wonder if they’re all still structurally sound. That first before and after GIF, the one with the two apartment buildings, that one on the right looked like it was leaning. We rarely see high-rises tested by powerful hurricanes, or tornadoes for that matter.

[aspen]

https://twitter.com/osocom123/status/1717733723687305442

With all these high-rises being gutted, I have to imagine the damage toll will be exceptional.

I think Otis has a high chance of becoming the costliest EPac hurricane by a huge margin. If I remember correctly, Manuel ‘13 was the costliest at $4.3B.

[Kingarabian]

I understand tbe lack of local obs and recon played a role in this deadly bust by the global and hurricane models. But they supposedly have satellite assimilation. GFS/ICON/Euro all runs 4x a day. How did the models not see the favorable conditions before landfall?

[kevin]

An estimate of peak intensity using the Knaff-Zehr-Courtney wind-pressure relationship.

Inputs
Vmax = 145 kt
Translation speed = 9.2 kt
R34 = 61 nm
Latitude = 16.8N
Background pressure = 1008 mb

KZC has a different equation below 18N and above 18N with slightly different results. Even though Otis was still below 18N, NHC often uses the >18N version for very intense systems close to the boundary. For example the >18 N equation is how the 872 mb value for Patricia was verified. Since 16.8N is kind of an edge case I'll show a range of both equations (the <18N equation results in a slightly more intense system).

Output
Peak intensity = 917 - 920 mb

So Otis was most likely a little deeper than NHC indicated in its discussion, but that will be something for post-season analysis and could also benefit from surface observations near Acapulco. If NHC underestimated the winds the pressure could've been even lower. Here are estimates of the minimum pressure of Otis depending on the maximum wind speed. An upgrade to 150 kt in post-season could thus cause an even lower pressure estimate.

140 kt = 922 - 925 mb
145 kt = 917 - 920 mb
150 kt = 911 - 915 mb
155 kt = 905 - 910 mb

Link to the latest version of the KZC relationship that NHC uses to my knowledge: https://rammb.cira.colostate.edu/resources/docs/Courtney&Knaff_2009.pdf

Based on some further analysis of the available data (unfortunately I pretty much can't find any surface observations, they were probably destroyed by Otis) I think Otis peaked at 150 kt before 'weakening' to 140 kt at landfall. I also assess Otis' R34 to perhaps be a little larger than what NHC had (for now I put it 75 nm which is still tiny for a cat 5), but this is more of a guesstimate using sat images since I couldn't find ground observations. Based on these things and using KZC with a 50/50 blend of the <18N and >18N equations I'd put peak intensity at 150 kt / 912 mb (which would make Otis the 4th most intense EPAC storm) and landfall intensity at 140 kt / 923 mb. Very interested to read NHC's analysis on Otis including its unexpected ERI.

[cycloneye]

An estimate of peak intensity using the Knaff-Zehr-Courtney wind-pressure relationship.

Inputs
Vmax = 145 kt
Translation speed = 9.2 kt
R34 = 61 nm
Latitude = 16.8N
Background pressure = 1008 mb

KZC has a different equation below 18N and above 18N with slightly different results. Even though Otis was still below 18N, NHC often uses the >18N version for very intense systems close to the boundary. For example the >18 N equation is how the 872 mb value for Patricia was verified. Since 16.8N is kind of an edge case I'll show a range of both equations (the <18N equation results in a slightly more intense system).

Output
Peak intensity = 917 - 920 mb

So Otis was most likely a little deeper than NHC indicated in its discussion, but that will be something for post-season analysis and could also benefit from surface observations near Acapulco. If NHC underestimated the winds the pressure could've been even lower. Here are estimates of the minimum pressure of Otis depending on the maximum wind speed. An upgrade to 150 kt in post-season could thus cause an even lower pressure estimate.

140 kt = 922 - 925 mb
145 kt = 917 - 920 mb
150 kt = 911 - 915 mb
155 kt = 905 - 910 mb

Link to the latest version of the KZC relationship that NHC uses to my knowledge: https://rammb.cira.colostate.edu/resources/docs/Courtney&Knaff_2009.pdf

Based on some further analysis of the available data (unfortunately I pretty much can't find any surface observations, they were probably destroyed by Otis) I think Otis peaked at 150 kt before 'weakening' to 140 kt at landfall. I also assess Otis' R34 to perhaps be a little larger than what NHC had (for now I put it 75 nm which is still tiny for a cat 5), but this is more of a guesstimate using sat images since I couldn't find ground observations. Based on these things and using KZC with a 50/50 blend of the <18N and >18N equations I'd put peak intensity at 150 kt / 912 mb (which would make Otis the 4th most intense EPAC storm) and landfall intensity at 140 kt / 923 mb. Very interested to read NHC's analysis on Otis including its unexpected ERI.

Agreed. The Otis TCR will be the most anticipated from the rest of 2023.

[ksgWXfan]

A question from someone who is currently out of practice observing model output and has done so relatively infrequently:

Have past model intensity forecasts shown skill in being reliably accurate or accurate forecasting RI?

If one of the models did prog a 36-hr period of RI of +100kts and a 70mb drop, wouldn't it have been considered an (absurd) outlier?

Interestingly RI is mentioned in this discussion at 03Z on the 24th referencing a 37Ghz ring.

Probably this imagery:
https://www.nrlmry.navy.mil/tcdat/tc23/EPAC/18E.OTIS/gmi/color36/20231024.005713.EP182023.gmi.GPM.color37.49kts.61p2.res1p0-cr100.jpg


https://www.nrlmry.navy.mil/tcdat/tc23/EPAC/18E.OTIS/gmi/36v/20231024.005713.EP182023.gmi.GPM.37V.45kts.52p1.1p0.jpg


In that discussion, it is mentioned that "none of the models show that intensity change occurring, seemingly due to persistent moderate shear and some dry air in the mid-levels" but with the forecast "to stay on the high side of the guidance given
the structure and very warm SSTs in the path".

18 hours later it was a major hurricane. So is it common for intensity guidance to be weighted too heavily? It's also unfortunate that microwave passes aren't more frequent.