March 20, 2014

When Is the Air Coldest Aloft? The Answer Might Surprise You!

When do you think the air in the Northwest is coldest above the surface, say at 5000 ft above sea level?

(1)  December 21st when the sun is weakest.
(2)  First week of January when the surface temperatures are lowest on average.
(3)  Early February
(4)  Now, mid-March
(5)  April 15th.
(6)  Late November when Northwest weather is most wet and windy.
(7) Mid-June, when the Northwest is the midst of lots of low clouds.

Time's up!  The correct answer is (4)...right now.   Bizarre isn't it?  Our coldest temperatures aloft are nearly coincident with the beginning of Spring (today!), when the sun is far stronger and daylight much longer than 3 months ago.

Don't believe me?   Here is a plot the climatological 850 hPa (roughly 5000 ft)  temperatures over western Washington by month.  The data is based on upper air data from 1948 to 2013 at Quillayute on the NW Washington Coast, nearby Tatoosh Island, and Seattle.  Take a look at the 50% line or the median temperature value.  The coldest month on average?  March! The extreme cold temperatures are in November through February (black line), but ON AVERAGE temperatures aloft are colder in March.  Who would of thought?
If you think about it, it all makes sense.   The sun strengthens after December 21st, and  a shallow layer at the land's surface starts to warm from the greater sun.  To put it more technically, the land's surface has limited heat capacity, the ability to store heat.   So it warms up rapidly as the sun's rays strengthen. That is why a sandy beach warms up so readily when the sun comes up.   And keep in mind that there is always cooling from infrared radiation emitted to space; warming only occurs when the solar radiation coming in is greater than the infrared radiation going out.

But now consider the atmosphere.  The air has a substantial mass and volume.  It has more heat capacity than the surface or what we call thermal inertia.  Think of huge flywheel....hard to spin it up and later to slow it down.

So the ground warms up relatively quickly, but it takes some time for the atmospheric flywheel to rev up from the sun's rays (and from the infrared radiation from the warmed surface below).   And around here there is another factor, during March and April we have less strong storms with powerful southerly winds bringing warm air up from the subtropics.  The flow is more westerly in the lower atmosphere, with a greater tendency to pull air from more northern climes.  The 850 hPa  wind maps for April and December illustrate this.

Now there are big implications for the differing heating rates of the surface and the air aloft during the spring....and you experienced the effects today if you were here in the Northwest.   With the surface warming up more rapidly than the air above, the difference in temperature between these two levels increases.  To put it another way, the rate of change of temperature with height...known as the lapse rate...becomes very large in spring around here.   And big changes in temperature with height leads to vertical instability and convection (cumulus and cumulonimbus clouds).  We have lots of instability and convective showers the last few days (like the picture below) as a result of the large springtime change in temperature aloft.

You can see all the spring instability clouds from the MODIS satellite.  Most of the white clouds in the image are convective clouds.

All of you have lots of experience with big changes in temperature causing convection:  in your hot cereal pot or the convecting water in your water for pasta. All examples of vertical instability produced by large changes of temperature in the vertical.

So because of the lag of warming aloft, spring is the most unstable time of the year, with the most change in temperature with height, and the most cumulus convection.  It is also the least foggy time of the year, since fog does not do well with large cooling in the vertical and vertical mixing.

And autumn is just the opposite.  The atmospheric flywheel keep the atmosphere relatively warm, while the land surface cools as the sun weakens.  The atmosphere becomes very stable, with lots of fog and the potential for air pollution.  Far fewer thunderstorms.


  1. Great post, Cliff! Very informative!
    To your point, we had several rounds of hail Wednesday afternoon in ne seattle. Fortunately, no supercells like they get in the Midwest (I am thankful for dewpoints under 60!)

  2. > The air has a substantial
    > mass and volume. It has more
    > heat capacity

    That changes dramatically with humidity, right?

    So upper level air that's lost its moisture because the water froze out is more transparent to outgoing infrared and so it can stay cool by losing heat to space, compared to moist air in which infrared is absorbed by water vapor and heat passed around to surrounding molecules by collision?

    Correction appreciated, this is hard stuff to follow.

  3. Why is there such a huge seasonal disconnect between the cold surface temps and the cold 850mb temps? Nearly 2 1/2 months, and by mid-March the average high temps at the surface are 8-15 degrees F above their early January levels.

    How do the 850-vs-surface seasonal temperature curves of OTHER midlatitude climates compare to ours? Is there a general categorical difference between oceanic and continental seasonal temperature curves?

  4. Appreciate the illustration, "mini" lesson, Professor.

  5. I grew up in Vermont, where statistically the coldest days of the year are in mid- to late-January. In the Northwest, the coldest days are near December 31st. Why is there a difference in the lag-time between these two locations? I thought perhaps it could be a result of snow cover that could reflect much of the sun's energy. However, in Winthrop, where there is a high probability of snow from December into March, the coldest day of the year is also near December 31st. It is even more puzzling after I've learned about the upper air pattern. Do you have an explanation?


Please make sure your comments are civil. Name calling and personal attacks are not appropriate.

How can we predict the climate 50 years from now if we can't forecast the weather next week?

This is a question asked by a commenter, Skeptic517, on my blog yesterday. It is a good question , one that I suspect most climate activist...