Tropical Atlantic
Hurricane hunter reconnaissance data and models.
Website: tropicalatlantic.com
Backup systems hosted at: @hurricanecity.com
Our site is not affiliated with any governmental entity.
- Current administration's destruction of NCAR will be an unmitigated disaster that will kill lots of people. When they destroy something that saves lives, that's what they're doing. They won't move important parts somewhere else. Entire point is destruction & mayhem like every other action they take.
- @bhensonweather.bsky.social spent most of his career working at NCAR. He writes, "Working at NCAR and UCAR was an enormous privilege and one of the great joys of my life. I gained a humbling appreciation for the intricacy of the atmosphere." yaleclimateconnections.org/2025/12/trum...
- The Air Force & NOAA hurricane hunters conducted 33 missions into, and around, #Melissa. Additional information appears in the alternate text for the images.
- Regarding the sonde from yesterday morning with the 219-kt spot wind, highest dropsonde wind ever seen in a TC, AOML/HRD shared the raw data and I took a look this morning. They’ll have the final call, but I see nothing wrong with the ob and I suspect it’s going to hold up.
-
View full threadI thought the AI response was on the mark, except maybe the part about 20 samples near the beginning - I no longer remember exactly how the filter is coded. The 3-bullet summary you have here seems good to me.
- Thanks. I'll add the disclaimer to all the decoded dropsonde data in my archive when I have a chance. People too often get confused by that and I wanted to make it more clear.
- Found an old email thread from my archives. A number of us looked at it in 2006 and concluded it was likely false. That agrees with my assessment now. The Katrina peak isn't valid.
- Thanks for checking into it. I'll work on adding a note about that to that sonde in my archive.
- Part 2 of 3: "Before transmission, the data is processed through a low-pass filter whose half-power amplitude occurs at a 5-second wavelength. This means that wind fluctuations occurring faster than about every five seconds are increasingly damped, reducing short-term noise and turbulence."
- Part 3 of 3: "Near the beginning and end of the descent, the effective averaging period shortens as the filter tapers to zero. Earlier dropsondes used a 10-second filter." Thanks.
- I was trying to put together a disclaimer for my website to add to dropsonde data. I'm not a meteorologist, instead I code things, so I inputted what you had said to ChatGPT to help me understand it some. That long conversation is here & probably has errors: tropicalatlantic.com/reconnaissan...
- But the conversation isn't important. I was wondering what the result sounded like to add. I have to ask scientists when I add something like it: Part 1 of 3: "A dropsonde's reported winds represent roughly a five-second smoothed average derived from measurements taken about every 0.25 seconds."
- Hopefully I can find the raw file. It was obviously of interest at the time and so must be in the archive.
- I was able to find raw sonde. Link is commented out on this page: www.aoml.noaa.gov/hrd/Storm_pa... So that the link doesn't appear. But file still is: www.aoml.noaa.gov/ftp/hrd/data... With launch time of 2005/08/28, 14:21:03.92 I included some screenshots of that section. (truncating part of it)
- From presentation about Katrina sonde: "A dropsonde released into the inner edge of that northeast eyewall descended into a mesovortex that contained a boundary layer (BL) wind peak of 120 meters per second (m/s), or 234 knots, at the 866 mb level (approximately 600 meters above the Gulf surface)."
- From presentation about Katrina sonde (continued): "If valid, these would be the highest winds recorded to date in the eight years that GPS sondes have been sampling these BL maxima." Thanks.
- I’m sure I’ve worked that sonde up before, but after 20 years I just don’t remember.
- I couldn't really find much about that Katrina sonde. I found one mention of it here from a presentation in 2006: ams.confex.com/ams/27Hurric... "The Intensity of Wind Gust Underneath Areas of Deep Eyewall Convection in Hurricanes Katrina and Dennis at Landfall" PDF: ams.confex.com/ams/pdfpaper...
- The raw wind data are taken every 0.25 s. Before transmission they’re run through a filter with a 5-s wavelength. Until a couple years ago it was a 10-s filter.
- Thanks. Does that mean it currently takes 2.5 seconds of data above the pressure altitude and 2.5 seconds below it and averages it all together for the reported pressure altitude? And then for surface wind is it less than 5 seconds or the last five seconds before the sonde terminates?
- Was 234 knot sonde at the 866mb level in Katrina unreliable? tropicalatlantic.com/recon/recon.... Nothing really matches close to it for that storm though, expect one with 182 knots on another mission at the 839mb level. Melissa had another mission with one with 210 knots & other high readings.
- Also, do you know what the wind reported in the dropsonde message is averaged over? (The ones released in WMO messages) Such as a half second perhaps? Or 0.25, or 1? I would like to put that information on my site. Maybe it has varied over the years and I can put a range if so.
- 1:00pm EDT on Oct 28, 2025: #Melissa has made landfall near New Hope, Jamaica with 185mph wind & pressure of 892mb. For wind, ties for strongest landfalling Atlantic hurricane matching Labor Day hurricane in 1935 in Florida Keys & Dorian in 2019 in Bahamas. For pressure, ties for Labor Day hurricane
- Per NOAA/AOML/HRD’s Sim Aberson, the 219 kt reading will be the fastest wind ever measured by a dropsonde if it passes final quality-control checks.
- Katrina had one a bit higher, but I don't if it is valid (234 knots, 269 mph): tropicalatlantic.com/recon/recon....
- At 11:15 am EDT, Air Force hurricane hunters posted about returning to Curacao "after encountering heavy turbulence today while entering the eye of Hurricane Melissa". "During the event, the aircraft briefly experienced forces stronger than normal due to turbulence" www.facebook.com/hurricanehun...
- A catastrophic, historical event is about to unfold in Jamaica. Hurricane warnings are also in effect for parts of eastern Cuba and parts of the Bahamas which will also experience the impact of #Melissa.
- 9:05am EDT Monday on #Melissa: Air Force dropsonde has measured a pressure of 893 millibars. (with 13 knots of surface wind this may indicate a pressure of around 892 millibars) If 893mb, it makes it 4th strongest ever recoded in the Atlantic in terms of pressure. If 892mb, it makes it tied for 3rd.
- If #Melissa makes landfall with a pressure of 892 millibars, it would tie as the lowest pressure on record for a landfalling Atlantic hurricane. The "1935 Labor Day" hurricane in 1935 made landfall in the Florida Keys with a pressure of 892mb.
- 9am EDT Tuesday update from NHC on #Melissa: "An Air Force Reserve Hurricane Hunter aircraft has found that Melissa is strengthening with maximum sustained winds of 180 mph (290 km/h). Estimated minimum central pressure based on aircraft data has fallen to 896 mb (26.47 inches)." www.nhc.noaa.gov
- Air Force hurricane hunters departed at 6:12am EDT from Curacao for #Melissa.
- In regard to high SFMR readings with the current Air Force mission, please see this post from @franklinjamesl.bsky.social. SFMR readings are estimated. Recon does indicate that Melissa is further strengthening and some of the most recent flight level wind data was not available for the 2am advisory.
- A sonde measured a pressure of 925mb at an elevation of 4 meters at 1:54am EDT. While some of the data is unclear from the sonde, it may have had wind of 16 knots at the surface. Please visit the NHC for the latest updates: www.nhc.noaa.gov If needed, they would post an update before 5am EDT.
- A disclaimer about SFMR estimated wind data will now appear in the recon system's archive page for an HDOB in some cases.
- I'm writing code to display wind swath over life of storm using track & wind radii. Want to include exact smoothing NHC uses. "smooth wind field depicted on the graphic is produced by a trigonometric interpolation between the four defined radii" (www.nhc.noaa.gov/aboutwindfie...) Anyone know how?
- Using actual wind radii gives results that do not match NHC due to smoothing not being applied. 1st image is NHC smoothed wind swath. 2nd is wind radii based on forecast advisory history (with interpolation; has bugs) 3rd is a smoothed wind field based on equation from AI. But I want exact NHC code.
- After investigating the cause of why our site was operating slowly we have found that the result was a Python script that accessed our site from the same IP address over 12 million times in the past week. We have blocked that IP address and the Python module used.
-
View full threadDuring times our site is under significant load, some of our site's systems will fail. During earlier mitigation attempts using tools at Cloudflare our site's systems failed when the highest security level was used. This is our final update on this issue.
- If you see significant disruption in connecting to our site, please contact us through Bluesky. As previously mentioned in our last post, certain content has been temporarily removed from our site. Some data will be made available again within a day or so and some should return next year.
- At times you may see a CAPTCHA on our site that you must solve to proceed. (sometimes it is automatic) In some cases, you may be temporarily blocked if you are using our site too frequently.
- We don't know how the security upgrades might impact the ability to view data from a Google Earth network link. (That is used to access data from our site within a file you previously downloaded from our site that allows you to view model and recon data.)
- Our site is not able to handle that kind of load. After identifying who it was it became clear that the repeated accesses were the result of a misconfigured script and were not meant to be malicious. As part of this process we have made changes to our site's security at Cloudflare.
- Our site is currently being hit with 30,000 requests per hour hitting the same pages in our recon system, mostly the same recon missions on Erin. It may be a Python script doing it. We're trying to mitigate it & you may be asked to solve a CAPTCHA challenge from Cloudflare before accessing the site.
-
View full threadThe recon system will remain removed for now. We might rename the main file used to view data in the next week so that the old location doesn't try to load data again if we add it back and the script that is trying to download hasn't stopped.
- As a note for the future, our site is not designed to be a repository that people check for data with automated scripts. We can't handle that kind of load. We will be redesigning things to help mitigate future incidents, but we still won't have the capability to act as that kind of repository.
- So that we more easily notice issues happening across all of our Tropical Globe sites, all content may be moved & consolidated to that site. That would allow certain resources, like large scripts that are duplicated across sites, to be downloaded only once rather than per site duplicating bandwidth.
- The mitigation is working. The model system is loading as normal rather than experiencing a delay of about 15 to 20 seconds. But again, data before 2005 in the model system will not be added back until the system is redesigned next year. (to reduce file sizes & change the way data is downloaded)
- There was also an unusual amount of traffic downloading large model wind swath files this month in the model system. As a result all older Atlantic content in that system will not return until at least next year when data will be delivered in a different way that hopefully uses a lot less bandwidth.
- Some of our other content may also be removed, at least temporarily. Other sitewide changes may be needed over the next year or so to be make sure the results of something like this has less impact.
- There have been 3.9 million requests in the past 24 hours and 11.5 million in the past 7 days. We have moved most of the content on the site so that it is no longer accessible. Some of the content will hopefully return later in the year while some will likely return in 2026 or 2027.
- Vast swaths of NOAA will be gone in 2026 if 2026 NOAA budget is passed. Hurricane research will be significantly reduced, with some parts of NOAA that do it being completely eliminated. Important tools used by the National Hurricane Center will be gone. There will be less hurricane hunter flights.
-
View full threadTrump, & Republicans that support him, are trying to get Americans to turn on NOAA by ensuring NOAA can't deliver quality forecasts you expect. Then they can privatize the weather service & have their rich donors run it all. Killing 1000s with less accurate forecasts has everything to do with that.
- Some of what was here will no longer be: www.nrlmry.navy.mil/TC.html Remains gone: www.fnmoc.navy.mil/tcweb/cgi-bi... Some of what could be gone in 2026 if NOAA budget passes: tropic.ssec.wisc.edu realearth.ssec.wisc.edu rammb-data.cira.colostate.edu/tc_realtime/ rammb-slider.cira.colostate.edu
- With what's already cut forecasts will be impacted. If 2026 cuts occur it'll seriously undermine ability for anyone to continue to forecast storms. If you end government & university tools everyone uses, from remaining government meteorologists to private ones, public suffers from degraded forecasts
- Data & websites related to weather & climate from US government are already disappearing. It's only going to get worse. It'll take generations to undo some of damage done by these actions. But for the thousands killed due to storms that couldn't be as well forecast, it will be too late for them.
- While Trump administration can't completely reverse progress made over past half century in forecasting, their attempts to do so will nonetheless costs tens of billions more than any money "saved" & will kill thousands. We will have storms intensify unexpectedly & we won't know it as fast as before.
- Imagine ending public release, including to hurricane forecasters, of vital data to see inside storms for fear of data being hacked & released publicly while on its way to being released publicly. That's Trump logic in 2025. If 2026 NOAA budget is passed, there won't be much hurricane research left.
- [Not loaded yet]
- [Not loaded yet]
- If you think being less safe is problematic, we encourage you to contact members of Congress in US. (www.congress.gov/contact-us) The economic & human cost of this devastating action can be reversed. You have the power. If no one wants to listen, then we'll only be less safe until they're voted out.
