Find us on Google+

Category: Meteorology 101

Why has the winter been so warm, and will it end cold? UPDATED

| 4:05 pm February 9, 2012

I have updated this blog with pictures sent to me by Troy and Jen Gibbons, who have a friend serving in Siberia.  The blog discusses how, until the last couple of weeks, the cold air had been bottled up in the Arctic.  Things are changing now though, as the AO is negative and the NAO is headed down.

This blog will attempt to explain, in fairly plain language, some of the material on atmospheric dynamics that I have taught in graduate courses at UAH.  However, I think weather enthusiasts, and even those who just want to know the forecast, will get something out of it.

It has clearly been a warm winter so far in Alabama, and actually over the eastern half of the US.  By my calculations (using data from Alabama state climatologist John Christy’s website), looking back at winters (Nov-Mar) since 1897, we’ve had the 21st warmest December, the 15th warmest January, and overall, if temperatures stay on the current track, we’ll have the 12th warmest winter on record.  Normally, for Nov 1 through Feb 6, the average temperature at BHM is 46.9.  Since Nov 1 2011, it has been 52.2, or 5.3 degrees above normal.  That’s a lot for a 3+ month period.

Normally in an average late fall/winter/early spring, we have 52 nights at or below freezing in BHM.  So far this year, we have only had 25.


1.  The general atmospheric circulation, maintaining balance, and how we get our cold air

To begin the discussion of all this, I will start with the basic, large-scale dynamics of the atmosphere over the Northern Hemisphere.  There is a lot more sun energy warming the earth in the tropics than in the arctic, and the difference is much larger in winter (they get practically no sun energy at all in the Arctic in winter).  If there were no weather systems to balance things out, temperatures would drop to -100 F or lower in the Arctic, and stay in the 70s or so in places like Alabama all winter.

But, we do have weather systems, in the form of areas of surface low pressure, associated cold fronts, and upper-level troughs and ridges.  When a low pressure area forms and intensifies along a cold front (lows tend to stay near fronts most of the time), a counterclockwise flow develops around it.  As shown in the example below (solid blue lines represent pressure, and dashed green lines are temperature), this pushes warm air north and brings cold air south.

These systems are part of the general circulation of the atmosphere, that redistributes heat northward from the tropics to the Arctic, and angular momentum and other stuff.  The cold air is generally associated with upper-level troughs (since it is more dense and pressure changes rapidly with height in cold air), and warm air is associated with upper-level ridges.  The atmosphere has to develop these low pressure areas and associated upper waves (called Rossby waves) to maintain balance.

Below is a computer model from Naval Research Lab showing a simplified Northerm Hemisphere, with the north pole in the middle, that is coldest in the middle (temperature on the right), and what happens when that unstable situation is present.  Lows and highs (left) form and redistribute the heat north and cold south.


Some winters, the atmosphere is more efficient in moving cold air south and movjng warm air north.  One way to explain the overall northern hemispheric weather pattern is through the Arctic Oscillation (AO).  When the extremely cold air stays bottled up in the Arctic regions (Alaska, northern Canada, Siberia), and does not get flushed out very often into midlatitudes (US, Europe, Asia) by cold fronts and low pressure, it causes the surface pressures to be high in the Arctic relative to the midlatitudes (since colder air is heavier and often causes higher pressure).  This is a positive AO.  When the cold Arctic air frequently gets pushed southward into the US/Europe, pressures are higher here and lower in the Arctic…this is a negative AO.

The past two winters (2009-10 when it dropped into the teens so many times and the edges of the Warrior River froze, and 2010-11 when we had several snow storms) were very negative AO…the cold air got moved out of the Arctic often, making us cold but the Arctic warmer than normal.

2.  Why has it been so mild this winter?

A similar indicator (and one James points out a lot on the blog) is the North Atlantic Oscillation (NAO).  It measures a pressure difference, but only over the North Atlantic (comparing Iceland to Portugal/The Azores).  A positive NAO is generally connected with a strong jet stream over the north Atlantic, and southwesterly upper flow over the eastern US that keeps Arctic air from being able to come south.

This year the NAO has been positive so far, so cold air can’t come south as easy over the eastern US and it doesn’t stay as long.  The AO, representing the northern hemisphere, was also positive until late January.  This means the cold air has stayed in the Arctic (it has been -20 to -60 degrees F over a large area this winter in the Arctic, abnormal for them, and there is debate about a possible record US low temperature, see  But, the AO has gone negative.  This is associated with bitterly cold air over Europe.

The map below shows the temperature anomaly (difference from normal) for December 1 through January 31.  Note the abnormally cold air in Alaska and Siberia, and the abnormally warm air from western Canada down into the eastern US.

3.  How about the rest of the winter?

This is hard to say.  As James and I discussed early this week, with cold air moving into Europe the hemispheric pattern is changing, and perhaps this could kick the NAO to negative later in Feb or Mar, etc.  The large scale upper-level waves have a lot to do with how all this plays out, but at least things are shifting now.  What happens in the Pacific (El Nino/La Nina), what happens in Europe, etc. all can have an effect on our weather here over a longer period of time.

We are already back to near normal temperatures for early February now, and the models do indicate several pushes of cold air coming south over the next 2 weeks.  But, as we get later into the winter, the sun is getting stronger over Alabama, so it makes long-term cold like we had the past two years much more unlikely.  This is not to say it won’t get bitterly cold before the winter is over.  A sudden pattern change in late February or early March could still get us down into the teens, and remember the Blizzard of 1993 was on March 13.


OK, here are the pictures from Siberia.  Temperatures there -30 to -50 have been common this winter.  That road must have buckled or something due to extreme cold/ice expansion?

Snow, cold update

| 7:04 pm January 12, 2012

Light snow showers continue over north Alabama this evening, mainly north of TCL-BHM-ANB.  However, a few reports of snow in NE Jefferson County (Clay/Pinson/Trussville areas)!  The snow has been a little heavier to the north, with slick spots developing, mainly on bridges, in HSV metro and over near Florence.  I’m glad I’m not driving home over the I-65 bridge over the TV River today!  Be careful tonight and tomorrow morning and check road conditions before heading out in case we get a little more snow than originally thought, mainly north US-278, but even a few elevated roads in the BHM and ANB areas could get icy.

As the big upper-level low and associated divergence moves through tonight, more light snow can be expected.   The snow is a little like summer showers, as the air is unstable.  It is 29 at the ground at UAH right now, but the MPR indicates it is 3 degrees at 4,000 ft., so the air is unstable!

Our air is coming from far up north and fast.   The trajectory model below shows that the air in Alabama tomorrow at 3 pm will have come from South Dakota.

It won’t be surprising to see temperatures drop more tonight, as it is in the teens and 20s upstream in MS, TN, AR, and MO.  Wind chill in the morning may be near 15 at times as far south as BHM, so bundle the kids up!  At least we are in a progressive pattern, and after a cold day tomorrow (highs in the 30s) and a cold night tomorrow night (lows 20-25 mainly), it will start to warm up by Saturday.


Big time fog, gravity wave

| 11:41 pm December 25, 2011

With the wet ground from the rain earlier today, and then skies clearing, dense fog has formed over Alabama tonight.

KBHM 260453Z 00000KT 1/4SM R06/1000V1400FT FG VV001 06/06 A3036 RMK AO2 SLP282 T00610061=

The visibility at BHM airport is 1/4 mile, and in some places it is lower than that.   The vertical visibility (VV001) is 100 feet, so landing a plane is nearly impossible.  With calm winds and radiative cooling (the skies are clear just above the shallow fog layer at the ground), visibilities will stay low through early morning, and may drop to 1/8 mile or lower in some locations, especially valleys.  Avoid driving if possible, and if you have to, use common sense.

A weak gravity wave moved over BHM over the past couple of hours too.  Check out the pressure trace at Helena.

Severe weather analysis – 146 pm

| 1:46 pm December 22, 2011

It’s pretty clear from the strong temperature and dewpoint gradient in the analysis above where the front is…from near Demopolis to BHM area to Gadsden.  Research has shown that on the near and just south of a front like this, atmospheric shear is enhanced.  This explains the tornado warnings and rotating storms generally concentrated from south of Tuscaloosa earlier to Shelby county now.  Given the location of the front, expect the tornado threat to stay south of I-59 for the most part, so people in Jefferson, Walker, Tuscaloosa, and Cullman counties can probably breathe easier.  Lots of heavy rain coming.

South of the front, in places like Anniston, Springville, Alexander City, Jacksonville, and Roanoke, the threat for damaging winds and tornadoes will continue.  However, the threat is not nearly what it was during the major outbreaks in the spring.  That day, energy-helicity index was near 10, today it is  only 1, and that is south of the front.

Kelvin-Helmholtz Waves – Analysis

| 4:00 pm December 19, 2011

As most of you know, some of the most amazing clouds most of us have ever seen moved across downtown Birmingham on Friday.  See James’ main post with more pictures here.

Given that one of the areas we have been studying at UAH the past 5 years is atmospheric waves, these photos were of particular interest to us.  Contrary to some comments, these clouds are not government operations or anything weird like that.  They are simply caused by Kelvin-Helmholtz (K-H) instability being in the perfect spot at the perfect time.

K-H instability occurs when there is very large wind shear (wind changing speed with height), and the air is unstable enough to allow the air in the waves to rise and fall.  A computer simulation of K-H waves is shown below.

(Click on picture for animation)

The air in the upper (blue) layer is moving faster than the air in the lower (red) layer, and with enough instability (temperature decreasing with height), these waves form.  They happen fairly often, but they are usually either a) high up in the atmosphere and may only be seen in clouds, like the ones I photographed over the Warrior River below, or b) occurring where there is not enough moisture for a cloud to form, so the waves are there but you can’t see them, or c) within a rain area (as shown in the radar picture from Springfield, MO).

Everything worked out just perfectly for the large waves near the ground in BHM on Friday.  First of all, as shown by a temperature and wind profile taken by an aircraft landing at the BHM airport at 1701 GMT (11:01 am CST), the temperature decreased quickly with height behind the cold front over a shallow layer (likely due to the cold air rushing in over warm ground), and the wind shear was extreme (a change of 11 mph over 700 feet).  The combination of these two numbers provides a Richardson number of 0.12.  As long as it is below 0.25, KH waves can occur.

Another example of KH waves is shown below.  The satellite picture shows the waves over south-central Texas, while the animation is a time-lapse from some one’s backyard in San Antonio.

Frost at 37 degrees?

| 9:00 am December 13, 2011


(Photo by Ronald Hughes, Coker, AL)

Why is it that we differentiate between a “frost” and a “freeze”?  How can ice form on the ground (frost is not frozen dew, but ice that deposits directly onto the surface from water vapor in the air) when the temperature is above freezing?  It has to do with the big temperature changes on clear, calm nights between the ground and where the official temperature is taken, 2 meters (about 6 feet) off the ground.

When it cools off at night (except when the cold air is blowing in behind a cold front, etc.), it is actually transferring heat to the ground.  The ground radiates heat into space, and then the air conducts heat to the cold ground.   On a rainy or windy night (even light wind), the air is fairly well-mixed, so the temperature is uniform in the lowest 100 feet or so.

But, on a clear, calm night, the coldest air (formed by heat conducting into the cold ground) may stay right near the surface, within 1 foot or closer.  Cold air is more dense than warm air, so it forms a pool of dense fluid, almost like water, near the ground.  The cold, dense air will slosh around sometimes, flow downhill, etc., causing changes in the temperature and helping to mix it up some, but the coldest air is nearest the ground.  Take a look at the night of Nov 30-Dec 1.

The top graph shows temperatures (every 5 seconds) from 6 pm Nov 30 through 8 am Dec 1, at 0.5 m, 2 m, and 10 m.  Notice how temperatures drop almost constantly until just after 6 am, when the sun comes up.  However, the temperature at 0.5 meters (about 1.5 feet) is the coldest, the 2 meter temperature is in the middle (this is the official temperature), and the 10 meter temperature is the warmest.  The bottom graph shows the change in temperature between 0.5 m and 2 m.  It is 1 to 3 degrees colder at 0.5 meters than at 2 meters most of the night.  There was a little wind increase around 130 am that mixed things up (see top graph), causing the ground temperature to rise some and the 2 m and 10 m temperatures to cool.  This shows up as the zero temperature change at 130 on the lower graph.  The “official low” was 28 degrees (at 2 m), but nearer the ground, it reached 25.  Sometimes we see temperature differences between 0.5 and 10 m as large as 6 degrees (one reason helicopters are used in late spring freezes to keep peaches warm).

By contrast, take a look at the night of Dec 3-4.  The wind was blowing 5 to 15 mph that night, keeping the air mixed up, not allowing the cold, dense air to stay settled near the ground.  Notice that there is little difference between the temperature at 0.5 m, 2 m, and even 10 m.

So, on a calm night, it is typically colder, sometimes several degrees colder, right near the ground than the official temperature indicates.  This is why frost can form when the official low is only 37.  Why no frost under the car?  Because the ground can not radiate heat into space under there!

Also, keep this in mind with pets, etc. They are low to the ground, so it is colder for them than for a person.  A doghouse, preferably up off the ground a few inches, will protect them somewhat from radiative cooling.

Snow analysis – 1019 pm

| 10:32 pm November 27, 2011 | Comments (4)

Above is a loop of 500 mb heights (lines) and vorticity (colors) over the next 36 hours.  You can see the upper low cutting off from the main flow, or forming a full circle, to our west tonight.  Satellite imagery indicates this may already be happening.  At the same time, much colder air is moving into AL at low levels.  Temperatures will drop to near 40 by sunrise, and then hold steady or fall into the 30s by afternoon.

With the dynamic cooling associated with the upper low and the colder air coming in anyway, temperatures aloft will drop very quickly.  For it to be cold enough for snow, we typically look for 850 mb temperatures below freezing (that starts in BHM around noon tomorrow), and 1000 to 850 mb thickness of 1310 or lower (cold air is more dense, and this number gives us an idea of the average temperature over the lowest 4,000 feet).   That number drops below 1310 around 2 or 3 pm tomorrow, and will stay there through Tuesday.  Temperatures will be VERY COLD just off the surface, in the 20s as low as 3,000 feet or lower.  So, by tomorrow afternoon in west Alabama and tomorrow night in east Alabama, it will be cold enough for snow to fall.

The 2 main things that create lift in the atmosphere and precip are warm air advection (warm air taking over cold air) and positive vorticity advection at upper levels (more red moving in in the loop, aka upper level disturbance).   We will have light precip tomorrow afternoon and night, but the heaviest precip should come through north and central Alabama early Tuesday morning.  We’ll have plenty of disturbances rotating around the cold air, but we reach the middle of the cold core (and, therefore, start having warm air advection) by then.  The best combo of warm air advection and positive vorticity advection is indicated by models over NW Alabama, but the models are always suspect in upper level lows.  That also means confidence in the forecast is lower than normal!

(Note by Tuesday morning we are at center of cold air and have warm advection)

Bottom line…expect rain to start mixing with a little snow west to east tomorrow afternoon and evening, then some snow tomorrow night.  How much and where is hard to pin down…NW Alabama most likely, but that could change.  With that much cold air aloft, we could even get a few convective snow showers that would be heavier, temporarily causing accumulation on grassy areas, cars, etc.  But, the ground is warm after all the 70s last week, and any significant accumulations are unlikely, as are any travel problems.  But, with this system being so cold and intense, the forecast could change, and somebody could have travel issues Monday night and Tuesday…especially at higher elevations and/or in NW Alabama (Florence, etc.)  It will be fun to watch…

NOTE:  Don’t forget to bundle up the family tomorrow also.  We’re going from 70s Saturday to wind chills in the 20s at times during the day tomorrow and Tuesday.

Snow analysis

| 11:47 am November 27, 2011 | Comments (5)

Wow, what a complicated forecast!  Two things are  pretty sure…it will be damp and occasionally precipitate from now through Tuesday morning…and it is about to get very cold.  It was in the 70s yesterday afternoon…same time tomorrow it will be in the 40s.

The big question is whether snow will fall.  With an upper-level low like the one shown above (forecast at midnight Monday night), due to dynamical cooling the air gets very cold aloft.  One of the main indicators we look for in terms of the atmosphere being cold enough for snow to fall instead of rain is the 1000-850 mb thickness (in cold denser air the pressure changes quickly with height.  According to the latest 12Z models, the 1000-850 mb thickness will be below 1310 m (cold enough for snow) from Monday at 6 pm through Tuesday at 6 pm.  Another indicator is a sub-freezing temperature at 850 mb (about 4,000 feet)…this occurs from 3 pm Monday to 6 pm Tuesday.  So, this means that any precip that falls Monday night or Tuesday will mainly be in the form of snow over north and central Alabama.

It is nearly impossible to pinpoint where the precipitation will fall, as the upper low is a complicated blob of upper-level disturbances, and with the air getting so cold with height, the air will unstable enough for a few bursts of heavy snow (almost like summer thundershowers) at times. 

All of this is to say that many Alabamians will see snow fall sometime Monday night or Tuesday.  However, it has been a warm 2 weeks overall, and the ground is very warm.  Also, precipitation amounts will not be that heavy.  If it snows hard, it won’t last but 10 or 15 minutes.  So, any accumulations will be light and probably temporary, since surface temperatures will stay above freezing (in the mid to upper 30s) during the time frame for snow.  We do not expect travel problems at this time, but it will be a neat event to follow.  If it had been down in the 20s last week, it might be a different story.

I’ll have another update tonight around 930 pm or so when 00Z models come in, but models, other than showing the overall picture, aren’t much good at this point in showing exactly where the heavier snow will fall.