Discussion of Intense Tropical Cyclones

[Hurricane Jed]

Please tell me how Tip is obsolete if it had recon. Several stronger typhoons you say? Based off what? Dvorak estimates? Satellite estimates? Seems like you're contradicting yourself. And does pressure and data really matter if they use such techniques to ballpark intensity and save lives?

[1900hurricane]

Something that hasn't stopped bugging me since it happened almost two years ago is Irma '17's warm CDO on September 5th, especially the western side of it. It bugs me so much that I even made a meme about it on Twitter that people seemed to quite enjoy.

 https://twitter.com/1900hurricane/status/1152737091270381571




But seriously, it strikes me as very strange. The warm region on the western side of the CDO even managed to break the black ring on BD imagery for a time, dropping instantaneous DTs to as low as 5.5. It just so happens that was the period of time when the data first began to suggest the 155 kt peak intensity. The warm is particularly apparent on the METOP-B pass I have below.



In spite of this all, microwave intensity estimates preformed very well with Irma. So, what gives? Visible imagery from around the same time doesn't show much appreciable difference across the CDO, warm side or cold side, a stark contrast to the IR imagery. There might be a slight texture difference just immediately on the western side of the eye, but nothing to explain the strange warm region. However, I have managed to come up with a working hypothesis: what if the warm region is similar to the warm spot on an Enhanced-V signature seen with an overshooting top in a severe thunderstorm?

What the warm region exactly is in an Enhanced-V is still up for debate, but there are a few ideas that include things like different amounts of water content in the updraft or the presence of lower stratospheric cirrus that is warmer than tropopause temperatures. Updraft characteristics are usually pretty different between severe thunderstorms and tropical cyclone eyewalls, with severe thunderstorms having much fatter CAPE and higher updraft speeds. With that said, updraft speeds fast enough to produce an overshoot cannot be ruled out in extreme cases. It's also worth noting that while extremely rare, a similar type of signature appeared a year later in Typhoon Mangkhut, which was another system where microwave intensity estimates were significantly higher than conventional IR methods.



I don't know, this is just a very preliminary thought of mine and could be way off base, but perhaps it's something worth considering.

[supercane4867]

The one thing Irma bugs me the most is the fact that she's been assigned the same peak wind speed of 155kt for 24 consecutive hours despite the drastic change in central pressure and satellite presentation. Looking at recon history and 1min interval IR loop on September 5th, we can see that the first occurrence of 160+kt flight level winds was found just after 12z, at the time when there's a warm dent on the western CDO as mentioned in previous post. Recon recorded an unflagged SFMR measurement of 160kt later at 17z when the satellite presentation was just beginning to improve. Judging by satellite data alone, I'd argue that Irma had her intensification phase during daytime(which the irregular eye and lopsided CDO suggest. I think she was still trying to sort out any outflow/dry air issue) and reached peak intensity later around 00z when the center is nearing Barbuda. I remember everyone was quite surprised when recon did not find any stronger winds while the pressure bottomed out at 914mb and satellite presentation continued to improve. With no evidence of unfavorable structural changes on radar, It's really hard to convince me that the 12mb pressure drop between 18z and 00z did not result in any increase in wind speed. While SFMR may had a high bias I think Irma could have attained 160kt intensity before landfall on the islands. I would also re-exam some ground observation, including the unofficial site that reported a 173kt gust on St. Barthelemy and 10-minute sustained wind of 165kt observed at Anguilla international airport (30m elevation).

That said, I don't think the CDO of Irma is particularly warm during the actual peak compared to other Atlantic storms of similar intensity. I remember Mitch and Allen both resembled a W ring around peak intensity.

[HurricaneEnzo]

It just shows that cloud temperature isn't the end all when it comes to cyclone intensity. The environment a cyclone is embedded in can play a big role in how intense the storm is or how low the pressure gets. It is easier for WPAC systems to attain such cold cloud tops cause the atmosphere is generally cooler in that area. Does this make them so much stronger than ATL storms as Euro likes to constantly imply? I don't necessarily think so as I said intensity is relative to the environment the storm is in. It might take the same amount of energy to produce -80 degree cloud tops in the ATL as it does -90 in the WPAC because of the difference in atmospheric temps so perhaps it equals out more than we believe.

[supercane4867]

It just shows that cloud temperature isn't the end all when it comes to cyclone intensity. The environment a cyclone is embedded in can play a big role in how intense the storm is or how low the pressure gets. It is easier for WPAC systems to attain such cold cloud tops cause the atmosphere is generally cooler in that area. Does this make them so much stronger than ATL storms as Euro likes to constantly imply? I don't necessarily think so as I said intensity is relative to the environment the storm is in. It might take the same amount of energy to produce -80 degree cloud tops in the ATL as it does -90 in the WPAC because of the difference in atmospheric temps so perhaps it equals out more than we believe.

It's simple. WPAC does produce more CAT5s than any other basins in the world. However, the frequency of occurrence does not directly correlates with the maximum potential intensity that a basin can achieve. For example, the SPAC basin had many exceptionally lame seasons with no majors or even no hurricanes (2008-09 for the most recent one - only 6 TS), but it also have no problem producing some of the most intense TC ever observed when conditions are favorable enough. Are there many overlooked sub-900mb typhoons in the WPAC since the end of recon era? Sure. Does the WPAC frequently produces ultra intense monsters with intensity far exceeding Patricia? HELL NO.

[Chris90]

One storm that I'm surprised hasn't received more attention in this thread is Hurricane Andrew. It's only been mentioned briefly a couple of times. In the NHC report, they mention that recon measured 162kts at the standard 700mb flight level during the flight that occurred directly before and during landfall. Considering Andrew was deepening right up to and just a little after landfall, combined with how vigorous the convection was, I believe Andrew made landfall with surface winds of 155-160kt, and I think, especially based on damage photos, there was localized extreme gusts in the 190-200kt range, especially in the form of tornado-scale vortices, also referred to as mini-swirls by Dr. Fujita. I shared a link earlier in this thread, I think sometime around April 2018 (maybe page 11-13 somewhere?) to a study that was done during Harvey using DOW and other instruments, where they found evidence of tornado-scale vortices embedded in Harvey's eyewall.
Below is a link to a YouTube video that shows some different radar loops of Andrew. It's still one of the most impressive eyewalls on radar in my opinion, just a thick, circular doughnut with intense convection rotating around the entire eyewall. You can see the mesovortices rotating around the inner edge of the eyewall as well.
https://youtu.be/ZgoqyumbXLE

Also to note with Andrew, on the 23rd of August, recon actually measured an even higher flight level wind at 700mb during his first stint as a Cat 5, as they measured 170kts, directly followed by a 165kt reading.

If I had to pick one storm from the past to get SFMR and eyewall dropsonde data from, it would be Andrew. I would especially love if eyewall dropsondes were used back in 1992, because I would really have loved to get data on the vertical wind profile of Andrew's eyewall.

[SconnieCane]

One storm that I'm surprised hasn't received more attention in this thread is Hurricane Andrew. It's only been mentioned briefly a couple of times. In the NHC report, they mention that recon measured 162kts at the standard 700mb flight level during the flight that occurred directly before and during landfall. Considering Andrew was deepening right up to and just a little after landfall, combined with how vigorous the convection was, I believe Andrew made landfall with surface winds of 155-160kt, and I think, especially based on damage photos, there was localized extreme gusts in the 190-200kt range, especially in the form of tornado-scale vortices, also referred to as mini-swirls by Dr. Fujita. I shared a link earlier in this thread, I think sometime around April 2018 (maybe page 11-13 somewhere?) to a study that was done during Harvey using DOW and other instruments, where they found evidence of tornado-scale vortices embedded in Harvey's eyewall.
Below is a link to a YouTube video that shows some different radar loops of Andrew. It's still one of the most impressive eyewalls on radar in my opinion, just a thick, circular doughnut with intense convection rotating around the entire eyewall. You can see the mesovortices rotating around the inner edge of the eyewall as well.
https://youtu.be/ZgoqyumbXLE

Also to note with Andrew, on the 23rd of August, recon actually measured an even higher flight level wind at 700mb during his first stint as a Cat 5, as they measured 170kts, directly followed by a 165kt reading.

If I had to pick one storm from the past to get SFMR and eyewall dropsonde data from, it would be Andrew. I would especially love if eyewall dropsondes were used back in 1992, because I would really have loved to get data on the vertical wind profile of Andrew's eyewall.

Really makes me wonder how Andrew (If I recall correctly) was operationally assessed as "only" a 120 or 125kt Cat 4 on the last advisory before landfall. A 5kt bump one way or the other in reanalysis (as with Michael) is understandable, but Andrew may have been underestimated by as much as 30 kt.

[Chris90]

Really makes me wonder how Andrew (If I recall correctly) was operationally assessed as "only" a 120 or 125kt Cat 4 on the last advisory before landfall. A 5kt bump one way or the other in reanalysis (as with Michael) is understandable, but Andrew may have been underestimated by as much as 30 kt.

That's correct, operationally they analyzed his south Florida landfall at 125kts. They also only analyzed his peak at 135kts originally on the 23rd after the recon flight which measured 170kts directly followed by 165kts. These are 10-second averages, but this is typically what they use nowadays, I'm not sure what their practices were back in 1992. Even still, I believe the 30-second average on that same flight was 154kts, which is still good for Cat 5.

Just 3 years previously, they measured 162kts at flight level in Hugo, and gave it Cat 5 classification at 140kts operationally, so I'm not sure what the discrepancy was. Also, the year before that they measured 173 kts in Gilbert and he received a 160kt intensity, so to me the 170kts measured in Andrew would have been good for a 155kt intensity on the 23rd.

The only thing I can think of is maybe they held it back due to pressure, as Andrew had a minimum pressure of 922mb at the time, and it was previously thought that generally 920mb was the threshold for Cat 5. Using the comparisons, Hugo bottomed out at 918mb and Gilbert bottomed out at 888mb, so maybe they were more confident with those storms. Obviously, with better recon data nowadays with the added benefit of SFMR and eyewall dropsondes, we've learned more about the relationship between pressure and wind and have more examples of storms producing Cat 5 winds at pressures above 920mb.

[SconnieCane]

Perhaps worth pointing out that that's three extremely intense and damaging tropical cyclones in the Atlantic during a five-year span that is generally considered a down period for the basin in terms of overall numbers.

[1900hurricane]

It's probably worth mentioning that Andrew '92 and Hugo '89 both occurred several years before the use of dropwindsondes. The data from those that established the standard flight level reductions we use today wasn't even gathered until the 1997 through 1999 seasons. When we were flying in those hurricanes, we didn't have a ratio to fall back on when impressive flight level winds were recorded.

GPS Dropwindsonde Wind Profiles in Hurricanes and Their Operational Implications

[supercane4867]

All the intensity numbers from 20th century needs to be taken with a grain of salt. Many wind speed estimates were off by as much as more than 20kts. That's why we have the ongoing re-analysis project.

[euro6208]

Please tell me how Tip is obsolete if it had recon. Several stronger typhoons you say? Based off what? Dvorak estimates? Satellite estimates? Seems like you're contradicting yourself. And does pressure and data really matter if they use such techniques to ballpark intensity and save lives?

I meant to say that if recon never ended in 1987, Tip would have been dethroned long time ago. During the recon era, we had typhoons back to back to back dethroning each other for strongest in the world. If recon never ended, we will have seen the same wind wise and pressure wise. That's what i meant for saying Tip is obsolete. Still sticking to the same record because there is no validation that something else is stronger.

However though there was a study conducted many years ago which found many typhoons that were stronger based on presentation and dvorak. Not sure what year but i think it doesn't include the monsters past the year 2000. Sad.


Dvorak can be off by a few categories especially for stronger systems hence people will take it more seriously and probrably save more lives if there was recon which almost always find something stronger despite it's less than impressive structure and convection. Look at the atlantic.

[euro6208]

It just shows that cloud temperature isn't the end all when it comes to cyclone intensity. The environment a cyclone is embedded in can play a big role in how intense the storm is or how low the pressure gets. It is easier for WPAC systems to attain such cold cloud tops cause the atmosphere is generally cooler in that area. Does this make them so much stronger than ATL storms as Euro likes to constantly imply? I don't necessarily think so as I said intensity is relative to the environment the storm is in. It might take the same amount of energy to produce -80 degree cloud tops in the ATL as it does -90 in the WPAC because of the difference in atmospheric temps so perhaps it equals out more than we believe.

It's simple. WPAC does produce more CAT5s than any other basins in the world. However, the frequency of occurrence does not directly correlates with the maximum potential intensity that a basin can achieve. For example, the SPAC basin had many exceptionally lame seasons with no majors or even no hurricanes (2008-09 for the most recent one - only 6 TS), but it also have no problem producing some of the most intense TC ever observed when conditions are favorable enough. Are there many overlooked sub-900mb typhoons in the WPAC since the end of recon era? Sure. Does the WPAC frequently produces ultra intense monsters with intensity far exceeding Patricia? HELL NO.

Yeah not just me but i bet alot of people thought as well that a TC with such cold convection would be stronger but hold and behold, the atlantic proves us wrong just because they have recon and better data.

How would you know that? The EPAC got lucky with Patricia for having recon. Not frequent but likely once every 2-6 years.

Exceeding? = Possible

Close to?= Very Possible.

Dvorak can brainwash alot of people.

[1900hurricane]

More confirmation of the NHC's uncertainty of the SFMR estimates at the higher intensities, this time with Dorian.

 https://twitter.com/EricBlake12/status/1167825365122846720

[CyclonicFury]

I still find it interesting NHC upgraded Matthew to a Category 5 in 2016, when flight-level winds at the time were only around 138 kt. The maximum SFMR was 143 kt. If NHC determines through more research that SFMR is high biased at these intensities, it's possible Matthew's peak intensity may be revised in the future.

On the flip side, if research discovers these SFMR readings to be reliable, marginal cases like Jose 2017 and possibly Dorian (If it fails to intensify further) may also be revisited.

[1900hurricane]

I still find it interesting NHC upgraded Matthew to a Category 5 in 2016, when flight-level winds at the time were only around 138 kt. The maximum SFMR was 143 kt. If NHC determines through more research that SFMR is high biased at these intensities, it's possible Matthew's peak intensity may be revised in the future.

On the flip side, if research discovers these SFMR readings to be reliable, marginal cases like Jose 2017 and possibly Dorian (If it fails to intensify further) may also be revisited.

The issue with Matthew is that the possible high bias at high intensities after the SFMR algorithm change wasn't called into question until 2017, after Matthew's data was already finalized into best track. I'd argue both Jose '17 and Dorian '19 are stronger cases for category 5 all else being equal. Some of the SFMR questions have been discussed earlier in the thread, but most of them remain unanswered thus far. We just need more data. Maybe we could run some research missions in the WPac to get high intensity data?

Speaking of getting more data, I'm really digging all the eyewall sondes we've been getting with Dorian that are staying in the RMW. All the dropsonde data from the stronger storms the past four years or so could end up going a long way verifying the validity of high end SFMR data or not. A similar study to the original Black, Franklin, and Valde, which established the standard reductions we use today, but with only the intense systems could reveal if these assumptions remains constant the further up in intensity a storm gets.

[Kingarabian]

First Irma and now Dorian. Dvorak has failed in matching the same intensity that Recon finds.

[WeatherEmperor]

If SFMR continues to have issues like this why are they still using it? Why dont they just use dropsondes exclusively? Is it a cost issue?


Sent from my iPhone using Tapatalk

[Kingarabian]

Well yeah. Biggest issue before Dvorak has to be SFMR. Since 2017, the NHC has trended away from trusting it. Regardless of basin, It's rare to get intense storms (other than the WPAC) and even if they do form, it's even more rare to have recon available. So to test the full scope of SFMR and determine its fate, there should be more research recon missions on Typhoons in the WPAC.

The other thing is Dvorak. It's less of an issue in the Atlantic because recon is available, but in the EPAC and WPAC, it's likely underestimating TC intensities by a good margin.

[CyclonicFury]

I still find it interesting NHC upgraded Matthew to a Category 5 in 2016, when flight-level winds at the time were only around 138 kt. The maximum SFMR was 143 kt. If NHC determines through more research that SFMR is high biased at these intensities, it's possible Matthew's peak intensity may be revised in the future.

On the flip side, if research discovers these SFMR readings to be reliable, marginal cases like Jose 2017 and possibly Dorian (If it fails to intensify further) may also be revisited.

The issue with Matthew is that the possible high bias at high intensities after the SFMR algorithm change wasn't called into question until 2017, after Matthew's data was already finalized into best track. I'd argue both Jose '17 and Dorian '19 are stronger cases for category 5 all else being equal. Some of the SFMR questions have been discussed earlier in the thread, but most of them remain unanswered thus far. We just need more data. Maybe we could run some research missions in the WPac to get high intensity data?

Speaking of getting more data, I'm really digging all the eyewall sondes we've been getting with Dorian that are staying in the RMW. All the dropsonde data from the stronger storms the past four years or so could end up going a long way verifying the validity of high end SFMR data or not. A similar study to the original Black, Franklin, and Valde, which established the standard reductions we use today, but with only the intense systems could reveal if these assumptions remains constant the further up in intensity a storm gets.

That's interesting. Since Matthew is already finalized into the best track, I doubt we would see any changes to the official intensity any time soon. But overall, its case for Category 5 is weaker than the others (I agree that Irma, Maria, Michael and Dorian are all solid Category 5s). In addition to the 143 kt SFMR reading, there was also a 142 kt SFMR reading, but it was marked as suspect. During the same recon pass, the maximum flight-level winds were only 138 kt, which supports an intensity of only 115-120 kt. I find it interesting NHC didn't use the blending method here, and decided to base the peak intensity off a single SFMR reading. Last night when NHC kept Dorian at 130 kt, there were both stronger flight-level and SFMR wind readings than Matthew had at its peak intensity.

If NHC followed the same method they did for Matthew with Dorian, they would have gone with an intensity of 175 kt.