In mid-December, 1944, the powerful U.S. Third Fleet under Admiral had just won the Battle of Leyte and was steaming east to get out of range of Japanese bombers. 500 miles east of Luzon, U.S. oil tankers were waiting to refuel the fleet. On the morning of the 17th, Halsey had received a weather report that there was a typhoon 400 miles to his east, or so forecasters thought. The powerful storm was actually just 120 miles away. Halsey believed his course to the refueling area would avoid the dangerous typhoon. As weather conditions deteriorated during the day, the Admiral ordered his fleet, which covered a thirty mile diameter area, to move northwest, thinking it would take them out of harm’s way. Even as the fueling operation began , powerful Typhoon Cobra was beginning to strike the task force. By nightfall, the fueling operation was abandoned and the fleet was trying to outrun the typhoon. Over the next 24 hours, Halsey made several maneuvers that managed to put much of his fleet right into the middle of the storm. There was mass confusion as the ships rolled on monstrous waves as high as sixty feet and dealt with shrieking winds 145 mph and gusts to 185 mph in the Philippine Sea. The central pressure in the storm was estimated to be 26.30 inches.

By the time the heavy weather subsided on the 18th, three destroyers and numerous other ships had been sunk by the storm. Nearly eight hundred sailors and marines died and 146 aircraft were destroyed. Winds and high seas tore life vests from the backs of some survivors. The encounter with the typhoon was more damaging to the Fleet than many battles with the Japanese. A court of inquiry found that responsibility for the disaster did lie with Halsey, but that no offense had been committed. The court made ten recommendations to improve weather forecasting services. But the recommendations were not implemented in time to prevent a second disaster with Halsey’s fleet later.

Lots of chatter around the weather office tonight about the possibilities for some significant winter weather during the next 10 days. In fact, the 18z (afternoon) run of the GFS model painted a very snowy scene across a wide area of Alabama, with a bullseye over East Central parts of the state starting during the day on Christmas Eve.

We find ourselves watching the latest model output anxiously every six hours. In fact that is what I am doing as I write this late on Friday evening.

Let's go through the latest ideas...

An Arctic front will bring another shot of very cold air to the state starting late tomorrow night. Sunday will be a miserable, cold and windy day with temperatures in the 30s. The cold airmass will also squeeze out a few snow flurries over North Alabama. There could be a light dusting on some of the ridges of Northeast Alabama.

Folks, by Monday morning, temperatures will average 17 degrees across the area. Some places could be as low as 13 degrees. I would not be surprised to see temperatures remaining in the 20s Monday!

But things flip-flop quickly on Tuesday. Our winds shift around to the south and a rain event starts by Tuesday afternoon. As low pressure forms and starts across the state Tuesday night, that rain will increase and continue into Wednesday before the low moves east. Some storms could form over southern sections of the state.

Another blast of really cold air will invade the South starting Wednesday afternoon. Then things get interesting.

Before the front is even through here very good, low presure will be forming over the western Gulf of Mexico. As this low moves across the Gulf, it will throw quite a bit of moisture up over and into the cold air mass in place. The rain/snow line will be over extreme South Alabama by Christmas Eve morning. We will see temperatures in the lower to middle 20s that morning.

Snow would begin early Christmas Eve morning as far south as Houston, Texas. This snow would spread into South Alabama and expand during the day. The 18z run of the GFS shows snow accumuluations in excess of six inches in a zone from South Alabama up into East Central Alabama. By Christmas morning, a large six inch zone would extend from near Gadsden southwest to Birmingham to Clanton to Troy and into East Central Georgia.

A strip of 1-3 inches would cover much of West Central, Northwest, North and Northeast Alabama.

This would be the first traditional White Christmas in recorded history for the Birmingham area as well as much of Alabama.

We emphasize that this is not a forecast at this time. It is simply too far in advance to be specific. But it is a possibility, albeit one that we will have plenty of time to hash and rehash as we head into the Holiday week.

**The late run of the GFS is out through Friday as I post this. The model output is further south and not as organize with the Gulf low. This would reduce the possibilities for snow over North Alabama. Once again, another flip flop in the models, so you see what we are up against.

Sorry for the long post, but it is an interesting situation. Everyone in the weather office will be contributing their thoughts all weekend long right here...

OK... Couldn't resist this... look:

http://grib2.wxcaster.com/CONUS_GFS_SFC_SNOW-DEPTH_180HR.gif

That is the accumulated snow chart from the 18Z GFS valid for midnight Christmas eve.

6 to 8 inches in Dothan!!!??!!

NOT A FORECAST. Just model output. But, that sure is fun to see.

Lots of e-mail today about the terms I use in the video and on the blog. Great questions... here are a few answers that might help:

GFS: Global Forecast System. The GFS is one of the main computer models we use for the short term, and into the medium range time period beyond 48 hours into the future. The GFS has output to 384 hours.

UKMET/ECMWF: We also use two European models to forecast in the medium range time period. One of models, known as the UKMET model, was developed in the United Kingdom. The other model, known as the ECMWF, is based in the European Meteorological Center.

ETA: A higher resolution model that has output for the U.S. only. The ETA has output to 84 hours.

540 line: Where the distance between a pressure of 1000 millibars and 500 millibars in the atmosphere is 5,400 meters. This is often used to define the rain-snow line in winter weather events.

Zero at 850 line: The freezing line (0 degrees C) where the pressure is 850 millibars, or about 5,000 feet off the ground. Another good way to determine the rain-snow line.

12Z run, 00Z run, etc: The computer models in America are run four times each day: 00Z, 06Z, 12Z, and 18Z... They are available around 10:00 a.m., 4:00 p.m., 10:00 p.m., and 4:00 a.m.

See model output from NCEP here:

http://www.nco.ncep.noaa.gov/pmb/nwprod/analysis/

Will add to this later...

The afternoon video is on the server:

http://beta.abc3340.com/weather/video.hrb

Once again, not much need to change anything. Get lots of e-mail from people who are focused on individual computer model runs. Try not to focus on individual runs, but the TRENDS they show. And, focus on the pattern. The pattern favors somebody across the deep south getting snow sometime between December 24 and 28. Maybe twice!

*Very cold and windy on Sunday with 30s all day, sub-freezing wind chills, and a few snow flurries. Monday morning we go into the 13 to 23 degree range. Most places will be in the upper teens.

*The pre-Christmas storm around the middle of next week will feature rain. Maybe a few flurries on the back side as cold air rushes in on Thursday morning, but nothing significant.

*A major storm will develop in the Gulf of Mexico Christmas eve, and will bring the risk of snow to parts of the deep south late December 24 into Christmas day. TOO EARLY FOR SPECIFICS. But, everything will be in place. It all has to come together at the right place at the right time. This will be our best shot at a white Christmas in a long time. Yeah, I know, the 12Z run of the GFS has the system too far south in the Gulf. I don't really care. The pattern and the trend favors a snow storm somewhere across the southern states. Don't focus on just one or two runs!

*Another storm in the Gulf forms in the December 27-28 time frame which could bring another risk of snow to Alabama. Needless to say, we won't even think about specifics this far out.

And, if you like some real voodoo forecasting, the 12Z run of the GFS spits out yet another snow threat around New Years eve.

Brian Peters will be doing the web video updates this weekend (Brian is a very good meteorologist; he really knows his stuff)... and J.B. Elliott and Bill Murray will be doing tag team forecasting on the web. I will be hanging around this weekend and will probably post some notes here on the blog.

Spend some time with your family this weekend; have some fun and some good laughs!

Dreams can come true, you know. But, not always.

The morning video is on the server:

http://beta.abc3340.com/weather/video.hrb

Lets look at our three coming attractions:

*Cold blast Sunday and Monday. Not much change here, although the guidance is not quite as cold now. Really doesn't make any difference, the wind will drive wind chill values down in the single digits and teens on Sunday with temperatures holding in the 30s. Monday morning will be the coldest with teens likely. Yep, we are still mentioning snow flurries on Sunday, but no big deal. If the ground gets white, it should be up in the high terrain of northeast Alabama.

*Pre-Christmas storm. Looks like mostly rain Tuesday and Wednesday. Possibly a quick change to light snow or flurries before it ends early Thursday, but doesn't look like a big deal. Another shot of very cold air follows for Thursday and Friday. This IS a specific forecast. Of course, it is subject to change, but confidence is fairly high.

*Christmas storm. A few days ago, we called this the post-Christmas storm, but the GFS now wants to bring this thing in here on December 24-25. This is the one to watch, and has the greatest potential to produce snow here. TOO EARLY for a specific forecast; lets wait until Monday for that. All of the figs will be in the basket; polar air, deepening storm in the Gulf of Mexico, favorable jst structure. But remember, everything has to come together at the right place at the right time. Just too early to call. But, for the snow lovers, model output is really looking good. But please remember, lots can change between now and then.

On today's video, check out the upper air charts sent to me by Josh Johnson, a former intern and now a graduate student in meteorology at the University of South Alabama. Interesting comparison to a rather historic event here. HMMMMMM.