Friday, January 27, 2012

Weather and the Human Condition...Meteorologists Can't Win This Battle

Weather predictions have an ingrained psychological component. The weather conditions might say one thing, Yet we will characterize the weather differently depending on who we are and where we live. Just looking outside at the current weather in our corner of the world in Northeastern Ohio can shape our way of thinking even if the overall weather data says otherwise. This winter’s lack of cold, small snow falls  and "milder” stretches are a prime example of how psychology plays a role in how we perceive the weather. First, let’s set the stage meteorologically then we'll examine the psych.

This early winter weather has been stuck in a fall pattern. A ridge of milder air has sat stagnant over the eastern US. The persistent low over the southwest and a general trough with colder air over the western US and Canada hasn’t budged. The persistent southwestlow has spawned many Texas panhandle/southern systems rich with moisture fromthe gulf of Mexico; a pattern that is not indicative of a typical December or January.    The tropical La Nina pattern (cooler Pacific Ocean sea surface temperatures) and an historically strong arctic pattern over the last several winter say Canadian clipper systems would dominate the early winter with frequent 2-4 inch snows.  Unfortunately, that same arctic component—our cold air driver--has stayed conspicuously stable and quiet.

One computer projection since late November has relentlessly pointed at clippers developing in Canada. A quick snow usually accompanies these clippers across northeast Ohio followed by colder air. This year, once we get inside 48 hours of the forecasted clipper event, the computer projection would quickly push the system back north allowing milder conditions to prevail!  Crazy stuff to be sure. In this atypical pattern where computer projections offer little consistent help in forecasting, we more than ever want a forecast that's definitive.  This winter has been anything but definitive.

At the heart of these computer projections are equations that have no exact solutions, just increasingly better approximations. I tell high school students to imagine math without numbers and no calculators. That description usually follows with a look of horror. Any thoughts of a student in the room becoming a mathematics major are immediately wisped out the window.

Yet weather forecasting is just that: An approximation. Throw in a splash of day-to-day randomness and you have a very tough recipe to replicate over increasingly long time periods.  Computers will get faster and faster. The amount of weather data from satellites will increase.  But the equations that are used in these simulations will always yield approximations…ALWAYS! No one wants to hear this but all simulations are highly detailed shades of "grey" of varying degrees. So are the weather forecasts that we present on television each day.

In a Facebook post, I've explained the "approximate" nature of weather forecasts in the context of this early winter wackiness. I hope that most people are receptive to them.  Many are judging by people’s comments. Yet in times like this where weather isn't matching the Decembers and Januaries of the past (snow and cold), the human condition takes over. Our mind gets thrown off. Our preconceived notion of Decembers and Januaries featuring snow and cold have been replaced with rain and milder air. These changes don't sit well. It makes us feel uneasy holding onto these conflicting ideas.  Psychologists call this "Cognitive Dissonance". How many times recently have you had a conversation with someone and they said, “What is the deal with this 'mild' winter…what is going on here?” The uneasiness in the question is palpable.

No one likes to feel uncertain or conflicted.  Weather most times exists in a perpetual "grey" area. It’s this built in randomness that causes frustration and conflict. Most of the time, we grossly underestimate its significance. We all have a built in motivation to reduce conflicting ideas by altering the existing conditions in our mind to create consistency. In the case of understanding the weather, we do this by 1) either believing the weather information which best fits our comfort level or 2) we alter its importance in our mind or 3) we just plain criticize it. Sometimes, it’s a blend of all three. This inclination to favor information that reinforces our comfort level is called a "Confirmation Bias". The problem is that by creating "consistency" through favoring information , we create a new false interpretation of the weather which we believe to be true. Rather than looking objectively at the reasons for the change scientifically (science scares people), most people tend to use an overly simplified and often inaccurate scientific explanation of the weather to ultimately confirm their predispositions. 

For example, I go on the air and say we’ll see rain to wet snow with 1-3 inches of snow by midnight. Many of us are already preconditioned to believe that this snow forecast will either be too much or too little. The reasons can vary from a disbelief in meteorologists in general--some have the "they never get it right!" mentality—to believing “the mild weather will stay” or that "the big snow is coming!"  Regardless, the preconception of inaccuracy is set from the get-go. 

The rain slowly changes over to snow but only for some areas. It takes a few hours of rain/wet snow before finally going to all snow. The new weather conditions highlighting a slower, back and forth transition to rain and snow present new information that favors the preconception of inaccuracy already present in many peoples' minds. I go on the air and explain that there is no well-defined line where rain goes to snow. I explain that it takes a while for cold air to overwhelm the milder air so the rain/wet snow mix would exist for a longer period of time.

Those who are preconditioned to believe that the forecast would be inaccurate dismiss the scientific explanation, ignore the random changes and replace them with their own simplified, non-scientific explanation while criticizing the real explanation from the meteorologist as hogwash. The countless emails and phone calls are strong evidence. All of this stacks the deck confirming their bias that weather forecasts and meteorologists are always wrong.  For a meteorologist, you can't win even if you present objective information to the contrary.

The psychology happens involuntarily: We struggle with the randomness of the changing weather conditions. We feel conflicted.  We feel frustrated.  We dismiss the weather information that we deem unnecessary to ease our conflict.  We might blame Lake Erie.  We often say "Its Cleveland." We criticize. We simplify.  We use “weather myths” to explain weather events. We come to a new conclusion and now believe we fully grasp the nature of the weather. The false interpretation we just created we believe to be very accurate.  We feel much better about ourselves. Case closed.

This inconsistency doesn't mesh well with every one's already highly simplified view of the weather. A crazy, changeable 8day forecast, negative connotations of weather forecasts and forecasters in general, coupled with a lack of general understanding is a highly volatile psychological mixture which further drives more false conclusions and irrational beliefs about the weather. All of which makes people even more uncomfortable and irritable.  The vicious circle is hard to break.

Weather prediction is just as much art and psychology as it is science. We try to tailor the 8day forecast to match the viewers’ perceptions by smoothing out some of the randomness so that it fits nicely on the 8day but it doesn’t always work out. Let’s hope that the weather pattern returns to "normal" soon.  As a television meteorologist, the forecasts daily would become easier to handle.  As a television viewer, the forecast could become easier to digest perhaps quelling some of our cognitive discomfort. 

I realize that shifting the weather back to “normal” won't change perceptions. It won't change the human condition. It probably won't alter beliefs about weather forecasts, meteorologists and weather patterns as much as I wish it would.  I can guarantee this:  Cold weather in early February will make most of us feel more at ease about the winter. That is until the first snow over spring break!

Tuesday, January 24, 2012

Cleveland Snow This Winter vs. Least Snowiest Winters

Temperatures have been well above normal since November 1st.  Snowfall hasn't stuck around for more than 4 or 5 days at a time. The reasons for the mild winter outlined a few weeks back. You can read about it HERE.

So far this winter, we've received:

22.8" through January 25th

How does this winter's seasonal snowfall stack up  through the end of January when compared to the least snowiest winters of years gone by?

This winter is 33rd all-time on the "least snowiest winters by January 31st" list! Here are the top 32 winters on the list working backward from the 2001-2002 winter.

Breaking down the list by decade:

2000s:  1
1990s:  1
1980s:  2
1970s:  4
1960s:  4
1950s:  3
1940s:  4
1930s:  4
1920s:  7
1910s:  3

Thursday, January 19, 2012

Does Warming Over the North Pole Lead to Colder Temps in Ohio?

Technical weather discussions aren't exactly front page material. When I try to explain the weather to my wife, the conversation abruptly ends of changes subjects.  I get it.  Its geeky.

Yet sometimes, we need the complex nature of the atmosphere to be explained so that the variables in a winter like this one can be handled a bit easier with ultimately better short and long term forecasts.

So indulge the geeky, scientist in me for a bit...the late winter forecast lurks at the end! :)


Most of our weather occurs in the lower 10 miles of atmosphere.  This layer is called the troposphere.  However, the behavior of the atmospheric layer above it--called the Stratosphere--plays a big role in major cold air outbreaks like the one in the winter of 2009.

The profile of the atmosphere that you may remember from middle school science class
Notice how the temperatures in the Stratosphere rise whereas the temperatures in the troposphere drop as you increase in height.  When the temperatures in the stratosphere rise suddenly over the high latitude regions close to the north pole, often time this rise can result in a sharp drop in surface temperatures which can propagate south across the continent and into Ohio.  Remember some of the below zero periods in our past winters? These warming events often enhance the cold outbreaks. I'll have examples of this later.

This sudden rise in temperature is called a SUDDEN STRATOSPHERIC WARMING EVENT.

There are several key components that go into identifying SSW events.  1. A sharp temperature rise of at least 50-70 degrees in a short time interval. And 2) is an abrupt reversal of the jet stream from west to east away from the pole. Since 1958, 31 Sudden Stratospheric Events have occurred according to this paper published in 2010.

Why does the stratosphere warm?  The physics get very complicated and so do the equations that describe it. Here is the technical description from a paper I just read:

"The SSW is caused by a rapid amplification of planetary waves propagating upward from the troposphere. Planetary waves deposit westward momentum and create a strong meridional circulation which produces a large warming in the polar stratosphere due to adiabatic heating (e.g., McIntyre 1982)"

To visualize these "planetary waves", imagine a stream with random ripples on the surface caused by leaves, rock, sticks or any random object that permiates the surface of the water. 

Imagine the atmosphere is like that stream filled with ripples. The only difference--and its a big difference--is that the atmospheric ripples are created by the earth's rotation and changes in heat from the oceans and land. These waves propagate like ripples throughout the atmosphere.

Courtesy: The COMET Program

When they travel up into the stratosphere (100,000 feet above the ground), they dissipate and give off heat as the flow decreases. This causes the warming described above.

Notice the warming that has taken place since December 15th

Here is where we are now...notice the sharp warming in the high latitude regions over the last week or so. The temperature graph is in KELVIN shown in the purple line. Last year's levels are in blue; the 30 year average is in black.

Also the change in the jet stream has begun but its not sharp...yet. If the purple line drops below zero, that will indicate sharp shift in the jet stream which might be enough to start pushing temperatures south.

I looked at several warming events of different intensities (not necessarily major events) to see how the different stratospheric warming events effects temperatures for us in northern Ohio and surrounding areas. The first year, 1987, didn't result in any cooling; the 2003 warming event lead to a much colder February.

2009 was a MAJOR warming event.  Look at the sharp temperature jump and the abrupt "180 turns" in the jet stream. Did it result in a major cold air outbreak across the eastern US? Not really...cold air developed over the extreme southeast US

Each SSW event is very different and can yield different temperatures across the US and Ohio. Often times, the cold air that develops over the North Pole sinks over Europe or Asia. While cold air outbreaks aren't a foregone conclusion over Canada and the Great Lakes when a stratospheric warming occurs, the chances increase by more than 50% during and after an event according to the scientific paper I sighted earlier.

Can we use the ARCTIC OSCILLATION (a measure of the arctic's stability) to determine cold air outbreaks during a SSW event?

I plotted the AO levels for 1987 noting the SSW time period to see whether the AO reacted to it.

How about 2003?

And finally 2009, one of the largest warming events in recent memory...

While Stratospheric warming events do lead to drops in the ARCTIC LEVELS, the are not always sharp and sustainable.  SSWs can lead to major cold air outbreaks but not always as indicated by the 2009 February temps

What does this technical weather stuff mean for our long term trends in Ohio?

The warming over the North Pole continues but is forecasted to bounce around a bit.

Take a look at the projections through January 21st.. Notice the colder than normal air building in Canada trying to push south and the above normal temperatures staying out west.

While I'm not jumping on the ARCTIC COLD bandwagon just yet, the trend given even a slight warming event would be for more frequent periods of colder temperatures in northern Ohio and more frequent Alberta Clipper type snows and more lake effect in February!

Thursday, January 12, 2012

What is Driving Our Not-so-Cold Winter?

We've grown accustomed to cold and snowy starts to winter.  The last two have featured rounds of cold and snow with steady snow cover. This winter hasn't behaved like her predecessors.  Why the change?  wasn't La Nina supposed to drive this winter's cold?  Not so fast.

The answers lie within several regions of the northern hemisphere and the indices that describe their behavior. Here are the regions and their parent indices in no particular order.  I'll explain what each index measures and describe the impact each index has on our weather here in northern Ohio.


We'll describe each teleconnect and show how each plays a role in our weather patterns across the Great Lakes and northern Ohio. It gets complicated so I've tried to keep the technical explanations to a minimum.


The Arctic Oscillation (AO) is a measure of the opposing pressure patterns of the mid-latitudes (across Canada and US) and the polar north.

When the index is POSITIVE, high pressure over the US and Canada drive storm tracks farther north keeping temperatures above normal across the eastern US (see left side of graphic above). When it is NEGATIVE (see right), the pressure tends to be lower across the US and Canada with higher pressure over the polar region. This causes storm tracks to come from the northwest ushering in colder than normal air.

The North Atlantic Oscillation (NAO), the AO's cousin, describes the pressure changes between low pressure over Iceland and the semi-permanent high over the Azores (eastern Atlantic).

Both of these teleconnections play direct role in how much and to what degree the cold air/storm track and intensity becomes in winter across the eastern US and Ohio.  The lower or more negative the AO and NAO levels, the more frequent and deeper the cold air.

The problem is that these pressure variations are incredibly random and very difficult to predict much further than a week ahead of time. This is why we refer to these as "wild cards".

I plotted the average AO and NAO levels for the Novembers and Decembers since 1991.

The last 3 winters had AO/NAO levels well below zero. See how the NAO and AO run hand in hand. When the AO goes down, the NAO follows. Notice below how many days (graph below) over the last two early winters the AO/NAO were well below zero

The pressures over the arctic were alot higher, the pressure/storm tracks across the Great Lakes were much lower, the cold air much colder. The shades of blue show temperatures below normal across Ohio and the  eastern US.

The pattern resulted in snows like  this...

This year, we count only 9 days with AO levels below zero! Not many at all.

Yet this year, the levels were very high. The highest levels in decades! What did this do? It kept the polar jet stream strong and consistant which keeps the cold air locked up north.  The pressure tendencies in the north Atlantic were such that it steered the storm tracks well north of Ohio.  All of this kept unseasonable temperatures for November and December. 
I plotted the average November and December combined temperatures for each year since 1991 on the AO/NAO plot. Notice that the warmer periods match up well with the higher levels of the
AO/NAO; colder periods for the lower AO/NAO levels.

The Pacific Ocean has also played a pivotal role in our pattern this year. TWO OTHER INDICES were very good guides on this milder pattern. The first is the PACIFIC-NORTH AMERICAN OSCILLATION. Its the general pressure pattern over this area shaded in red.

The higher the pressure pattern, the stronger the ridge over the western US and Canada, the stronger the trough/cold air over the US. This, unlike the other indices, means that lower than average pressure was centered over the western US/Canada and higher than average pressure was over the eastern US.

This winter, the PNA has been mainly negative.

The ocean temperatures in the tropical Pacific have been below average since last fall. This below normal ocean sea surface temperature pattern is a classic LA NINA PATTERN.

This change in ocean temperatures buckels the jet stream across North America which alters the storm tracks so that they come from the northwest. Last winter, we had more than 9 cold front-type snows from Canada. This year, ONLY TWO!

How does this LA NINA compare to others that share similar characteristics (time of peak/length of duration)? Use like to use the MEI Index because it factors in PRESSURE DIFFERENCE, WINDS, AIR TEMPERATURE, CLOUD COVER and SEA SURFACE TEMPERATURES in the tropical Pacific Ocean. Here is a graph that plots this LA NINA event to the others that compare favorably. Notice that it is similar to the three year LA NINA back in the early/mid 1970s.

How can we summarize what has happened so far?

  • While the LA NINA pattern was present last winter allowing for clipper snows from Canada, the ARCTIC/NORTH ATLANTIC PATTERN continued to drive the cold air south reinforcing the La Nina storm track.  

  • This year, the LA NINA is present yet the ARCTIC/NORTH ATLANTIC CONNECTION is noticably absent.  So instead of reinforcing the colder air/northwesterly storm tracks, it shifted the storm track more south featuring above normal rain (7+ inches for Cleveland in November and December).The ridge over Ohio and the east coast pushed snow producing CLIPPER SYSTEMS north and the temperatures higher.   
  • The cooler than normal Pacific water temps off of the coast of California and Alaska kept the cold and storminess further west (NEGATIVE PNA SHOWN EARLIER) while further enhancing theridge of dry/milder weather across the Great Lakes

How about snow totals comparing this year through January 12th to the last two winter? MUCH LOWER

How will these TELECONNECTIONS change in the weeks ahead?  More on that later after this rare round of snow moves through. :)

Wednesday, January 04, 2012

What About this 2011-12 La Nina...Is it Different from Last Year?

I created an animated Sea Surface Temperature map comparing last year from early October 2010 to early January 2011. Notice the streak of blue and purple in the middle of the map. That's the classic La Nina signature. Also, check out the warm pocket in the northwestern Pacific and the cooler water to the east of it. THE WARM POCKET SEEMS WEAK... 

This year, the La Nina signature doesn't seem as strong. Perhaps more importantly, THE WARM POCKET IN THE WESTERN PACIFIC IS STRONGER although it is showing signs of fluctuating.

The weaker La Nina and the strength and position of the warm pool in the western Pacific, I believe, have played a large role in the storm paths across the US and over Ohio.  Last year, we had prolonged cold.  This year, only one day below freezing in December. 

La Nina is peaking which means that other factors will come into play in the months ahead. Will we be able to get the Christmas decoration down due to the snow and cold? Will sledding at Virginia Kendall Park in Peninsula we possible?  Will we see more of the grass than snow cover? More later this week

Tuesday, January 03, 2012

Where is the Lake Erie Ice? Will it freeze over this winter?

Living in Cleveland, Ohio, the lake just to our north plays a major role in our wintry weather. Lake effect snow especially early on in the season can produce more than 24 inches of snow in a very short time. The relatively warm water of Lake Erie (temps between 40 and 50 degrees) in November and early December provide a surplus of moisture necessary for the creation of full blown lake effect events. All that is needed is cold air.  This winter trough the end of December, we haven't had much.  The absence of cold air keeps the lake free of ice. 

All of this begs the question:  Will Lake Erie freeze over this year?

Go back to last winter when we had 18 of 31 days in December with temps in the 20s. The lake water temperature went from 45 degrees on December 1st to 33 degrees on December 24th. The ice cover for December 20th looked like this: 

RED indicated 80%+ ice cover. Notice the open water in white.

January of 2011 continued with temperatures below averages with 20 of 31 days in the 20s or colder. Ice development on the lakes tends to lag several weeks after the cold air starts.  Here is the ice cover map on January 2nd.
The ice started to develop further east
Now the ice cover map for January 17th, only 15 days later.

80% ice cover spread across most of the lake

Finally, the ice map for January 27th.

80% to 100% ice cover over much of the lake 
How about this year? Temperatures in December of 2011 stayed well above normal. Only one day had a high BELOW FREEZING! The lack of ice on lake erie isn't that surprising.
Ice cover for December 20th

How about January 2nd? A little bit of ice over the western basin where the water is shallow indicated with the light blue color.

Even the ice coverage is less than 10%
How about the rest of the month? We are running way behind in ice development. Below is the actual satellite photo of the open water of Lake Erie on December 26th.  The white on the right is cloud cover

A two week arctic outbreak would be needed to drop the lake water temperature to 33 degrees. The current temp is 39 degrees!  Remember last year, we fell below 39 on December 9th! It would take some time for the water temp to fall to 32-33 degrees before widespread ice would form.  I just don't see this happening. Maybe 50% coverage by early February. If it doesn't happen by early February, the water temps will start to climb. Even last year, the water temp was in the upper 30s by the end of February and above 40 by March 13th.

All of this means that when it does get cold even for a day or two, lake effect snow will be more common and might last into February and March. Whereas last year, lake effect snow ended in mid January. 

Just our luck.