Re: Congratulations to Jim too by the way
Posted by Chris in Tampa on 4/20/2019, 11:43 am
Here is the ob where that SFMR estimated wind occurred:
http://tropicalatlantic.com/recon/recon.cgi?basin=al&year=2018&product=hdob&storm=Michael&mission=15&agency=AF&ob=10-10-170430-25-917.6-124-108

Here's another look at that 138 knot SFMR wind noted as suspect:



Sometimes really high winds might get marked as suspect. Also sometimes if the wind is in shallow water. Or sometime for another reason I guess. Sometime they don't release it, such as over land or if they have some kind of issue. And sometimes it isn't suspect and they happen to go over really shallow water, or even just onshore, and they get a bad reading.

In this case, in the next 30 seconds after that 10 second estimated 138 knot surface wind, there was no reading released.

To me at the time this reading looked valid enough.

The Bob Henson/Jeff Masters blog covers some about how they got to the upgrade. Here was one reason:

"The maximum flight-level wind measured directly by U.S. Air Force hurricane hunters was 152 knots, which would correspond to a surface wind of 137 knots (158 mph) using the standard formula that relates flight-level to surface winds. This allows for the chance that the aircraft was not sampling the most intense part of Michael. Some of the data gathered by the Stepped Frequency Microwave Radiometer (SFMR) were not available at the time of landfall; when reconstructing the full SFMR dataset, the NHC team found that the maximum 10-second estimate of SFMR-indicated winds were 152 knots (175 mph) at 1707Z (1:07 pm EDT). This lends support to a 1-minute landfall wind in the Category 5 range, although NHC notes that there is some evidence from Hurricane Maria for a high bias in SFMR data when winds are this strong."

Assuming SFMR is close to correct, I thought the reading was valid because it was in the location you would expect to find these winds.

At the time I disagreed with the NHC's reasoning for not upgrading it as I discussed in this post:
https://canetalk.com/2018/10/1539295892_1539287470.shtml

Here is the reasoning from the NHC, at the time, for not upgrading:

"There were SFMR measurements of 132-138 kt,
but the validity of those observations are questionable since they
occurred in shallow water and were flagged."

I didn't think the water was that shallow. I had no idea what the depth was that far out, at the time I thought there could be some kind of shallower water I don't know about, but going from how far offshore it was, I disagreed with their reasoning.

The NHC usually uses SFMR reading, which are 10 second estimated surface wind, not directly sampled of course like flight level wind is. Assuming they haven't exactly found the highest wind, they go with the 10 second estimate to represent the 1 minute sustained winds. The plane is constantly moving fast so by the time it sampled the highest 10 seconds of surface wind in a 30 second period, you may already be averaging really high winds with lower winds. I think going with the SFMR, until we find out differently, is the best we have.

So the 138 knots to me is enough to say at the time, it's reasonable to believe that was the intensity. And that was very near landfall. As you can see from that single HDOB message, the core didn't have a lot of readings. They fly right through from one side to the other, not flying around and around the eye. So we have only limited recon. We didn't have that extra reading at the time.

Here is the paragraph from the NHC's tropical cyclone report:

"The maximum flight-level wind measured in Michael near the time of landfall was 152 kt by a USAFR aircraft at 700 mb (approximately 8,000 ft) in the southeast eyewall at 1723 UTC 10 October. This flight-level wind would yield an estimated surface wind of 137 kt using the standard NHC flight-level to surface adjustments, which account for the possibility that the aircraft did not sample the maximum flight-level wind. The maximum real-time surface wind estimate from the SFMR was 138 kt in the south eyewall at 1706 UTC that day. However, there were missing SFMR data in the real-time transmission during that penetration of the eyewall. Re-construction of the instruments raw brightness temperatures during the dropout period by the NOAA AOC indicates that the maximum 10-second SFMR wind estimate was 152 kt near 1707 UTC. The SFMR winds support an intensity greater than 135 kt, especially if the 152-kt value is correct and uncontaminated by wave shoaling in water about 89 ft deep. However, there is a significant caveat regarding the SFMR data, as experience during Hurricanes Irma, Jose, and Maria in 2017 suggests the possibility that the SFMR has a high bias at the wind speeds in question. Research to determine if this is the case is currently underway."

So from that they say the water depth was 89 feet, which I thought was deep enough, but maybe not for these kinds of winds. They are talking about the SFMR possibly having a high bias, so that's notable too. But the SFMR from the missing reading was estimated to be 152 knots. (10 second surface estimate) That would be 175mph. I don't know why in real time that ob was not there. Maybe the wind was so high the software did more than just mark it suspect, it didn't report it. I don't know. While the NHC often goes with the SFMR, a single extreme value wouldn't likely be enough for them to raise it to 175mph 1 minute sustained. I think you would want a few more readings. But 160mph didn't seem like an unusual estimate at the time. I like to see two 10 second estimated surface winds in a row and average them. Every 30 seconds usually there is a line in an HDOB message. The highest 10 second estimated surface wind in that 30 seconds is what is reported. So if you get two in a row, it's as close as you can get to what might be 1 minute sustained. If they are both around the same, not seemingly in shallow water, and seem to be in the part of the storm where you would expect those winds, I usually think the readings are likely valid. So in this case if you averaged the two, it would be over 160mph. But then we get into is it representative? Was that a mesovortex they happen to catch? Was that actually the wind going around the eye south of the center? Or is 89 feet too shallow? I have no idea. Is there a high SFMR wind bias in these kinds of winds? Unfortunately, I guess we have to go what we think we know now with the understanding that it could eventually change if we learn the SFMR isn't as good in these winds.

Then they get into other points about why they upgraded.

"By emulating as best as possible the collection of reconnaissance aircraft flight-level data, the radar velocity estimates should be considered to be comparable with the aircraft winds. For the last aircraft pass through the southeastern eyewall, this radar technique yielded an estimate of 155 kt winds at 1722 UTC at approximately the location and altitude where the aircraft reported 152 kt. It is notable that the Doppler radar data indicated that stronger velocities existed just to the northeast of the aircrafts flight path, suggesting the aircraft likely did not sample the strongest winds associated with Michael."

So it seems like even at flight level, the aircraft might not have even sampled the highest flight level winds.

It does seem like it was a category 5 based on the understanding of the current science. There is no direct measurement, but until we start investing on some kind of network all along the coast to measure winds, and build it to withstand significant winds, this is all we got.

Consider this image:



In under 5 miles, the SFMR readings went from:

17:05:30Z: 116 knots (134 mph) ... originally noted as suspect
17:06:00Z: 138 knots (159 mph) ... originally noted as suspect
17:06:30Z: 152 knots (175 mph) ... originally was missing
17:07:00Z: 119 knots (137 mph) ... originally noted as suspect

Over two minutes they went from 134mph to perhaps 175mph to 137mph.

The path is noted on the map. Looking at where the storm made landfall by using the plotted circles for vortex messages, you can where that core was in relation to the coast. There may have been very few people who saw cat 5 winds. Between the main part of Tyndall AFB and Mexico Beach is where the worst may have been, where there is very little there. I haven't read the whole tropical cyclone report, and I don't know how much wind damage there was in Mexico Beach compared to surge, but I don't know how high the winds were there. I'm not sure if Mexico Beach was just into the category 5 winds or not. I have not read the full report to see if they get into that kind of estimated detail.

I did see this though:

"It should be noted that future revisions to the Florida landfall intensity are possible, as additional re-assessment is expected once the research on the reliability of the SFMR at these high wind speeds is complete."

As for the Coyote drone, it was a lot earlier in the day. Since the hurricane hunters have to fly near to where the Coyote is dropped to be near the signal until the little drone doesn't fly any longer, it's something they rarely do. Getting regular recon is more important for now I guess. You sample a larger area faster. And of course in winds this high the drone sometimes just doesn't survive long.

And I took a look at the satellite loop of Michael:

https://twitter.com/metofficestorms/status/1050116806667583488


https://twitter.com/NOAASatellites/status/1050441034277228545


And various of radar:

https://twitter.com/CharlesRoopWCTV/status/1050080928775725060


https://twitter.com/CharlesRoopWCTV/status/1050080928775725060


https://twitter.com/BMcNoldy/status/1052949171169349637


https://twitter.com/BMcNoldy/status/1052903261269835776


I'm not good at reading radar images, so even though it looks stronger, I don't know if that was due to the radar not getting to that side as well. But looking at that last tweet, Brian McNoldy says it did seem stronger. Then again, the radar site in the first image is less than 70 km to the NW eyewall. In the second image, the other radar site is 130km to the eastern eyewall. But he's the expert. Recon was more focused south and east of the center. I have no idea why the northwestern eyewall appeared more intense on radar. Can it look more intense but still have weaker winds? I would assume based on the damage, and how hurricane spin, but that's beyond what I know.
59
In this thread:
Hurricane Michael upgraded to Cat 5 - Beachlover, 4/19/2019, 11:00 am
< Return to the front page of the: message board | monthly archive this page is in
Post A Reply
This thread has been archived and can no longer receive replies.