Forecasting Snow...


Meteorologist Trent Okerson

So, more snow anyone?  I didn’t think so.  Old Man Winter has unloaded on our area this week, and reinforced the idea that our weather can make drastic changes in a very short period of time.  I wanted to take a shot at explaining a couple of the things that we have dealt with this week, and hopefully give you an appreciation of the various factors that give meteorologists a headache when it comes to forecasting winter weather and snow amounts.


Snow-to-Liquid Ratio:

Let’s begin with an easy concept.  Snowflakes and raindrops are very different.  A snowflake, with its crystalline structure, is much less dense than a raindrop, where all of the water molecules are closed packed together.  This means that the average snowflake will occupy much more space than a raindrop…meaning that a small amount of liquid precipitation can yield much higher totals if it falls as snow.  Based on climate records, the average snow to liquid ratio for our area is roughly 10:1.  This means that if you melted down 10” of snow, you’d end up with only 1” of water.  It also means that the margin for error when forecasting snow amounts is greatly magnified!  Think about this...if the forecast was for 0.1” rain, but a band of heavy rain developed and we ended up with 0.6” of rain instead, no big deal, right?  However, during a snow event, that would be the equivalent of forecasting 1” of snow, and instead receiving 6”….very much a big deal!


And there is more to the snow-to-liquid conversation.  Whereas the size of a raindrop doesn’t change based on the temperature, a snowflake does.  The snow-to-liquid ratio changes based on atmospheric and surface temperatures.  A 10:1 ratio works well for a wet snow with surface temperatures in the neighborhood of freezing.  But as temperatures get colder, the snow to liquid ratio increases.  Whereas Monday morning’s snow fell at more of a 10:1 or 12:1 ratio, Wednesday’s snow fell with temperatures in the upper teens, leading to more of a 20:1 ratio.  Using Murray as an example, at this ratio, the whopping 5” of snow was the equivalent of only 0.25” of rain!  Here at the weather center, we saw 3” of snow….the equivalent of 0.15” of rain.  Needless to say, the difference between 0.15” and 0.25” is a small margin…sort of like splitting hairs.  Oh, and to add a final thought to this topic…the snow-to-liquid ratio can change as the temperatures change during a snowfall event, making this even more complicated.


Mesoscale versus Synoptic Scale:

When we create the forecast, we look multiple times a day at a number of different forecast models that crank out charts for the entire U.S.  These models are typically quite reliable for giving us the “big picture” outlook for our weather patterns.  This big picture view is called the “synoptic scale”…an area of several hundred miles.  But the forecast models don’t do so well at analyzing things at the “mesoscale”, or small scale.  This is where we get into trouble sometimes by relying too heavily on models.  This Monday proved to be a great example of where synoptic scale models missed out on the smaller scale details that ultimately were responsible for producing some extreme snowfall amounts. 


The reason we went from a forecast of 1-2” of snow to parts of the area actually seeing anywhere from 4” to locally 10” of snow was because of mesoscale snow banding, a small area of snow just a county or two wide, where there is increased lift and additional moisture, and thus heavier snowfall.  Whereas forecast models can let us know days ahead of time that an area of low pressure will pass to our south and result in a snow chance, they cannot let us know that a band of heavy snow that is 50 miles wide will set up just south of the Ohio River.  With situations like this, about the best meteorologists are able to do is to examine radar trends, and adjust the forecast accordingly.


Deformation Zone:

The reason that mesoscale snow bands develop is because of a feature called a deformation zone.  Not only is a deformation zone very difficult to forecast, it is also difficult to explain.  The NWS defines a deformation zone asAn area in the atmosphere where winds converge along one axis and diverge along another”.  Here’s my best shot at a simple explanation….a deformation zone sets up when winds at the surface converge (blow toward the same region), and winds aloft above that same area diverge (blow in different directions).  This creates an enhanced area of lifting….and a concentrated band of heavy precipitation.  This was what was responsible for the heavy band of snow over western KY on Monday, and the heavy band that set up from Hopkinsville to Murray and into TN on Wednesday.  A deformation zone can take a perfectly good forecast and turn it into a blown forecast, with burst of snowfall that can create snowfall rates of 2-3” per hour.  Unfortunately, deformation zones are extremely difficult to forecast more than about an hour or two ahead of time.


Hopefully this gives you a little more insight as to what happened this week.  If you would like to learn more about all of the intricacies of snow, and you aren’t scared of some hardcore scientific wording, check out this presentation from the NWS Louisville.

Some good news....a big pattern change is on the way!  A steady & significant warming trend will take us back into the 50's by next week, and the latest 8-14 day outlook includes a continuation of those above average temperatures!