GaWx

The 12Z EPS at 228 is similar to the 0Z EPS at 240 with <20% of members threatening the NE Caribbean, Bahamas, or CONUS. Of those that do, two appear to be cat 4-5 with a 923 and a 933 as of 240 both moving WNW near/toward the Bahamas. Bermuda is still threatened by a good # and there are once again many Hs.

Down to 994 at 192.

Yep although it still is the strongest of the 12Z runs at 144. Also, the 0Z Euro had similar weakening 132-156 followed by restrengthening. I'll be looking to see if the 12Z is similar.

12Z Euro is strongest of the 12Z models as of 120 and pretty close to earlier runs.

The 12Z model consensus is weaker/delayed on TCG. Not only does the GFS have very little early in the run (GFS TCG is delayed til the Caribbean), the CMC is similar. Plus the last few ICONs have been weaker than recent days. Also, the last two UK (goes out 168) have no TC after many in a row with it. Trend? Mean further west if there is ever TCG?

NDJ

The CFS ONI peak has come back way down from +2.15 just a few weeks ago to ~+1.80-1.85.

For this same AEW, I count 27 12Z EPS members (53%) with a H. That is more than the 0Z and is as high as any EPS I can recall for any as yet to develop system in the MDR this season to date. So, at a minimum, this looks likely to be a big ACE producer giving us pretty satellite pics. Edited

From the 12Z UKMET hour 168: Surface showing TC 300 miles E of Leewards (moving WNW at 15): scroll down https://weather.us/model-charts/gbr/2023090112/americas/sea-level-pressure/20230908-1200z.html -------------------------- H5: In addition to being a threat to the NE Caribbean, the 12Z UKMET doesn't yet look like anything close to a guarantee to not hit the CONUS although the best bet this far out is a miss based on 75%+ of E MDR systems doing just that: scroll down https://weather.us/model-charts/gbr/2023090112/americas/geopotential-height-500hpa/20230908-1200z.html

It won't be +1.2. It will be either +1.0 or +1.1. August OISST is ~+1.3, which translates to ERSST likely being slightly cooler than that (~+1.25?). June was 0.84 and July was +1.00. So, if Aug ERSST turns out to be +1.25, that would mean JJA average of +1.03, which would round down to +1.0 for the table. But if Aug ERSST were to come in at +1.31+, that would give a JJA of +1.05+, which would be reported as +1.1 in the table. So, that's why I say +1.0 or +1.1. To get a +1.2, you'd need a +1.6 for Aug, which isn't happening.

-Usually the best bet for an E MDR system since 75%+ don't hit the US although nowhere near a safe bet at least yet imo -12Z UKMET: still has TCG but unlike the prior run having TCG on Sunday, this waits til Wednesday (9/6). With the delay, it is notably weaker but is still in a potentially dangerous position at 168 (9/8) 300 miles E of the Leewards moving WNW at 15 mph: NEW TROPICAL CYCLONE FORECAST TO DEVELOP AFTER 120 HOURS FORECAST POSITION AT T+120 : 14.2N 44.1W LEAD CENTRAL MAXIMUM WIND VERIFYING TIME TIME POSITION PRESSURE (MB) SPEED (KNOTS) -------------- ---- -------- ------------- ------------- 1200UTC 06.09.2023 120 14.2N 44.1W 1010 27 0000UTC 07.09.2023 132 14.1N 46.7W 1009 29 1200UTC 07.09.2023 144 14.9N 49.0W 1007 37 0000UTC 08.09.2023 156 15.4N 51.8W 1007 31 1200UTC 08.09.2023 168 16.1N 54.9W 1007 34

Actually, this new CANSIPS run, despite its -NAO, isn't very cold. Where do you see "very cold"? It is for the bulk of the E US (other than the slightly AN in upper Lakes to NNE) near normal (after adjusting to 1991-2020 instead of 1981-2010 baseline), which is oddly enough slightly warmer than the prior run (which was slightly BN in the SE): Of course, near normal would seem very cold relative to most recent winters. I'll take the colder change although I'm hoping for actual BN here in the SE as El Niño provides the best chance for that.

I'm trying to understand the terminology better. At 12:12Z on 8/30, the linked graph for Tampa shows a 6.64 foot tide (red line) that appears to be using MLLW as a reference. It also shows a 1.433 predicted (blue line, which is the astronomical) tide then. In addition, when one clicks on "observed - predicted", it then displays a purple line that was at its max of 5.21 feet at 12:12Z. That's 6.64 ft minus 1.433 ft. So, that appears to be the maximum extra water level caused by the storm. Are you saying that the maximum "observed - predicted" is not the storm surge? If not, what is that called? https://tidesandcurrents.noaa.gov/waterlevels.html?id=8726607&bdate=20230829&edate=20230831&units=standard&timezone=GMT&interval=6

From NWS CHS: "A MAXIMUM TIDE LEVEL OF 3.47 FT MHHW (9.23 FT MLLW) WAS OBSERVED AT THE CHARLESTON HARBOR TIDE GAGE." From the first image linked below, there was a maximum surge of 2.38' at CHS at 8:12 PM on August 30th that coincided exactly with high tide to produce the 9.23 ft tide (in red)/CHS battery flooding: https://tidesandcurrents.noaa.gov/waterlevels.html?id=8665530&units=standard&bdate=20230829&edate=20230831&timezone=GMT&datum=MLLW&interval=6&action= This was made worse by the center being ~50 miles closer to CHS at high tide vs where models had projected it then. Thus the strongest perpendicular SE winds came in and peaked just before and at high tide instead of being much lighter then, which would have been the case had the center been 50 miles further away. OTOH, it being 50 miles closer to CHS at high tide meant 50 miles N of SAV meaning W (offshore) winds at Ft. Pulaski instead of very close to SAV. That result was a major factor that lead to the maximum surge at Ft. Pulaski of 2.7' occurring much earlier at 1:36 PM on August 30th, which was luckily near LOW tide when winds were nearly perpendicular to the coast (ESE winds). Soon after that, winds shifted to SSE (no longer perpendicular and thus not pushing as much water in) and eventually to WSW at the time of high tide literally pushing the high water out. The result was that at the 8:24 PM high tide, the storm surge had shrunk to nearly nothing as per this linked image, which meant at Ft. Pulaski only an 8.95 ft tide vs a 9.4 ft tide (both in red) at the high tide the evening before. So, while CHS was having their worst flooding, Ft. Pulaski and the nearby islands/lowlands around SAV were already back to normal! https://tidesandcurrents.noaa.gov/waterlevels.html?id=8670870&bdate=20230829&edate=20230831&units=standard&timezone=GMT&interval=6 ---------- Edit: My power was out ~10 hours (5PM 8/30-3AM 8/31). I measured under 1" of rain. Closer to the coast got less than well inland, which was well predicted by the models. The airport got ~1.5". Just 50 miles inland from SAV despite the fast storm movement (20 mph), Candler, Bulloch, and Screven counties had flooding from very heavy rainfall as the remnants of the very wet NW eyewall hit that area. Statesboro, Metter, and Newington were deluged with 7-8"! That extended into Allendale, Hampton, and Colleton counties. Some of that scraped CHS (airport well inland), which got 3". There were 3 confirmed EF0 tornadoes, two in the CHS metro (including the one linked to earlier ITT that caused that car to lift in the air) and one in Liberty County, GA. No injuries and minimal damage, fortunately.

Well, I looked at the site linked below and it appears that Tampa may have had a 5.2 ft storm surge at 8:12 AM on August 30th! If this were to verified as accurate, that may be the highest surge at least since the 5 ft from Gladys of 1968! That's pretty amazing considering how far offshore it was and that it landfalled way up in the Big Bend! Can anyone verify this data? https://tidesandcurrents.noaa.gov/waterlevels.html?id=8726607&bdate=20230829&edate=20230831&units=standard&timezone=GMT&interval=6

The typically conservative UKMET (0Z) once again is pretty strong with this E MDR lemon AEW that about all models now develop: NEW TROPICAL CYCLONE FORECAST TO DEVELOP AFTER 54 HOURS FORECAST POSITION AT T+ 54 : 10.9N 22.9W LEAD CENTRAL MAXIMUM WIND VERIFYING TIME TIME POSITION PRESSURE (MB) SPEED (KNOTS) -------------- ---- -------- ------------- ------------- 1200UTC 03.09.2023 60 11.4N 24.8W 1008 37 0000UTC 04.09.2023 72 10.8N 27.6W 1008 32 1200UTC 04.09.2023 84 11.6N 31.6W 1008 31 0000UTC 05.09.2023 96 12.3N 35.1W 1006 34 1200UTC 05.09.2023 108 13.0N 39.3W 1005 35 0000UTC 06.09.2023 120 13.7N 42.6W 1005 31 1200UTC 06.09.2023 132 13.9N 45.5W 1001 37 0000UTC 07.09.2023 144 14.5N 48.1W 1000 39 1200UTC 07.09.2023 156 15.3N 50.6W 997 51 0000UTC 08.09.2023 168 16.2N 53.3W 997 47

Due mainly to Harold and Idalia, much of the Gulf has cooled ~1C over the last two weeks: Aug 17th near record warmth dominated by 31-32C (~88-90F): August 30th: still very warm but much of it cooled ~1C to mainly 30-31C (86-88F) with a small area affected by both storms cooling nearly 2C (32 to 30) near 25-26N, 84-85W:

Great point. Does anyone know what the official Tampa Bay storm surge was?

-Based on the normalized SLP anomaly TT maps of the last two CANSIPS runs, it has ~-0.35 NAO averaged out over DJF on this run vs ~-0.30 on the prior run. A -0.35 NAO would qualify as a -NAO per my requirement of sub -0.25. -If that were to verify, it would be only the 2nd -NAO winter of the last 13 winters and only the 7th of the last 45. -I'd love for this to verify, but I still don't believe the NAO will average that low as I'm predicting either neutral or +NAO. But keeping in mind that neutral goes as low as -0.25, it could end up as neutral and still be close to this -0.35. -The combo of two things is keeping me from forecasting a -NAO DJF. First, we've been in an +NAO winter era the last 44 winters with only 14% of these winters having a -NAO. Second, an approaching solar max and associated expected very high sunspot activity (150+) tells me that this is not a winter to go out on a limb (i.e., going with the 14%) and predict a sub -0.25 NAO. The highest DJF averaged sunspot activity during these pretty rare -NAO winters within the current +NAO era (last 44 winters) was only 33. -CANSIPS of 11/30/22 forecasted a -0.40 NAO for DJF 2022-3 vs the actual +0.68. So, the CANSIPS can be far off and this one was only one day before met winter: -I do think that one sub -0.50 NAO month of DJF is quite possible though (perhaps in Feb). That wouldn't even be that far off from you since you think Dec will suck. And then there's the chance for that also in March, which I'm not even considering since it isn't part of met winter.

PUBLIC INFORMATION STATEMENT NATIONAL WEATHER SERVICE CHARLESTON SC 1209 AM EDT THU AUG 31 2023 ..HIGHEST WIND REPORTS FROM IDALIA LOCATION SPEED TIME/DATE PROVIDER SOUTH TYBEE ISLAND 69 MPH 0439 PM 08/30 WXFLOW BEAUFORT 66 MPH 0619 PM 08/30 WXFLOW CALIBOGUE SOUND 64 MPH 0536 PM 08/30 WXFLOW Based on the above from NWS CHS, I'd think 60+ on parts of HH is supported. This data was pulled from here: https://kamala.cod.edu/offs/KCHS/2308310409.nous42.html

This map of the newest DJF Greenland SLP anomaly shows a 2-4 mb anomaly vs 1993-2016 climo. Per TT, the prior CANSIPS run had a 2+ mb anomaly (vs 1981-2010 climo) only in SE Greenland. So, if the different climo bases as well as any other differences of mapping parameters for WCS vs TT aren't making a big difference, I'd then expect the updated TT to show a significantly stronger DJF Greenland SLP anomaly in its update compared to the prior run on TT.

Keep in mind that Nino 4 now being at +1.1 on the OISST, especially with it only August, is on a relative basis vs its own historical ranges super strong. The current +1.1 there is a new record for August.@bluewavehas talked about how warm Nino 4 has been vs its past. The old record warm weekly prior to 2023 was only +0.8 (2015). Warmest Aug wkly Nino 4 back to 1982 OISST +1.1: 2023 +0.8: 2015 +0.7: 2019...outside of El Niño +0.7: 2002 +0.6: 2018, 09, 06, 04, 94 +0.5: 1991, 87 +0.4: 2014, 97 +0.2: 1982 However, some caution is needed with the weekly OISST table when comparing to prior years, especially the further back one goes. That's because they are all based on 1991-2020 averages rather than adjusting more for cooler climo the further back one goes like ONI does. For example, the mere +0.2 of 1982 would be equivalent to something a good bit higher if it were reflecting, say, 1966-1995 averages. Similar idea for the +0.5s of 1987/91. Even the +0.4 of 1997 would be warmer. Note that even the weak Nino of 2018 was able to attain +0.6. Even more notable is the +0.7 of 2019, which is very warm for being past the 2018-9 Nino. So, it is important to keep in mind that 2023's +1.1 is likely helped out by overall GW. If there were an adjustment for simultaneous warming of other areas similar to how RONI has recently been adjusting significantly cooler for ONI, Nino 4 would likely be a fair bit cooler than +1.1. Nino 3 and Nino 1+2 would also be in the same boat.

I was comparing Idalia with 2016's Hermine, which peaked at 80 mph/981 mb 5 hours before FL landfall near where Idalia hit. It went steady state through landfall rather than either further strengthening or weakening. From landfall, it took a near identical NE path to a little W of Savannah. Per the archives at closest approach Hermine was at 31.9N, 82.0W..55 mi WSW of Sav and 989 mb/50 mph. (It was moving NE at 18 mph.) To compare, Idalia was at 32.2N, 81.7W or 40 mi W of Sav with 984 mb/70 mph. (It was moving a similar NE at 21 mph.) So, they were at a similar distance away at their closest but Idalia was noted to have 20 mph higher winds (70 vs 50). What's really strange is that Hermine produced 6 mph stronger winds than Idalia! From Hermine wiki: "Hermine weakened while crossing from Florida into Georgia, but still produced sustained winds of 45 mph at Savannah, with gusts to 58 mph." https://en.m.wikipedia.org/wiki/Hurricane_Hermine#:~:text=Hermine weakened while crossing from,(93 km%2Fh). Idalia's highest sustained/gust at Savannah were 6 mph lower with 39 sustained/gusts to 52. I remember clearly that Hermine was significantly stronger. And keep in mind that Idalia landfalled at 125 mph vs Hermine's only 80. I'm thankful but also amazed at the comparison. Any thoughts from anyone about this comparison?

I've had no power for 6 hours. About 1/3 of the area lost power. Otherwise, thank goodness we got pretty lucky compared to how bad I thought the winds might get based on earlier NHC forecasts. The highest official wind gust was near 52 mph and there were only a few near that high. But had those forecasts held, we likely would have had gusts into the 60s, which would have meant more trees down and a bigger mess. I was comparing Idalia with 2016's Hermine, which peaked at 80 mph/981 mb 5 hours before FL landfall near where Idalia hit. It went steady state through landfall rather than either further strengthening or weakening. From landfall, it took a near identical NE path to a little W of Savannah. Per the archives at closest approach Hermine was at 31.9N, 82.0W..55 mi WSW of Sav and 989 mb/50 mph. (It was moving NE at 18 mph.) To compare, Idalia was at 32.2N, 81.7W or 40 mi W of Sav with 984 mb/70 mph. (It was moving a similar NE at 21 mph.) So, they were at a similar distance away at their closest but Idalia was noted to have 20 mph higher winds (70 vs 50). What's really strange is that Hermine produced 6 mph stronger winds than Idalia! From Hermine wiki: "Hermine weakened while crossing from Florida into Georgia, but still produced sustained winds of 45 mph at Savannah, with gusts to 58 mph." https://en.m.wikipedia.org/wiki/Hurricane_Hermine#:~:text=Hermine weakened while crossing from,(93 km%2Fh). Idalia's highest sustained/gust at Savannah were 6 mph lower with 39 sustained/gusts to 52. I remember clearly that Hermine was significantly stronger. And keep in mind that Idalia landfalled at 125 mph vs Hermine's only 80. I'm thankful but also amazed at the comparison. Any thoughts from anyone about this comparison?

That was wild, the car lifting up! Those little tornadoes are sneaky and can pack a heck of a punch even though this was most likely "weak" on the scale. Fortunately no serious injuries despite the car damage.