Wednesday, February 20, 2013

Weather columnist Kevin Myatt: Meteorology and weather don't always mesh

Kevin Myatt is The Roanoke Times' weather columnist.



Weather with Kevin Myatt

Recent columns

Read the Weather Journal blog

#swvawx on Twitter


The 2012-13 winter has been an exercise in mesoscale madness.

"Mesoscale" is defined this way in the Random House dictionary: "In meteorology, the scale of phenomena that range in size from 1-100 km (0.7-67 miles), intermediate between small (single storm storms) and synoptic scales."

Much of meteorology is focused on the synoptic scale, or the tracking of large weather systems covering hundreds or even thousands of miles, such as high- and low-pressure systems and cold and warm fronts.

Forecasters have become very good at tracking and predicting the movements of weather systems on the synoptic scale.

But the weather we experience is determined in the mesoscale. This winter has shown repeatedly it can be very different within half an hour's drive.

We saw mesoscale differences exemplified to a small degree on Tuesday morning, when some locations saw snow fall while others didn't.

Some of it could be explained by typical elevation factors — Blacksburg at more than 2,000 feet above sea level seeing snow while Roanoke near 1,000 feet got little or none. But not all of it could be. Lynchburg had heavy snow for a time, even though it is 400feet lower than Roanoke, and several miles farther east than snow was expected to fall.

Saturday brought an even more extreme example of mesoscale madness. For the most part, forecasts were accurate in projecting a broad area of light snow south and east of our region from central Virginia into the Carolinas, and a second area of streaky light snow to the north and west from typical upslope wind effects in the mountains.

But there was a streak, about 20 miles wide, in Appomattox, Buckingham and Charlotte counties east of Lynchburg that got 4 to 8 inches of snow, well above any forecasts.

Roanoke was the recipient of a mesoscale band of heavier snow on Feb. 7 and 8. Forecasts of rain, sleet and snow mix with less than an inch accumulation were generally correct over much of Southwest Virginia, but the Roanoke Valley got 3 to 5 inches, when forecasts projected neither enough atmospheric cooling nor enough moisture to support that much snow accumulation.

The reverse occurred Jan. 17, when most of our region saw anywhere from 2 to 13 inches of snow, but a portion of Bedford County saw little or no accumulation.

Mesoscale differences have greater implications for life and property in severe storms forecasting. Meteorologists have become very good at projecting, days ahead of time, where a tornado outbreak is likely. But picking out a single storm or even a county or two where a tornado is likely to occur is difficult even just a few hours ahead of time, especially when it is a single event (like the Pulaski tornado of April 9, 2011) rather than an outbreak.

The limits of available data and the resolution of forecasting tools are much of the reason forecasts become more difficult in the mesoscale even when synoptic scale features are accurately predicted.

Weather balloons launched twice a day from sites hundreds of miles apart, supplemented by measurements from satellites and aircraft, make up much of the data that goes into computer forecast models. But that can't sample every cubic inch of the atmosphere.

When 500 vertical feet of temperatures half a degree warmer or colder than projected can have enormous impacts for precipitation at ground level, those missing bits of information add up quickly. Much of meteorology — even that done by computers — is educated guesswork to fill in missing data.

Also, computer models don't really see the atmosphere as an infinite number of small points, but rather as various squares and cubes with the same conditions throughout those blocks. Resolution has improved greatly over the years, to a few miles, but there can still be fairly large differences in weather even within those blocks.

Meteorologists study the mesoscale exceptions to synoptic weather patterns to learn how to fine-tune forecasts in the future. And computer technology offers the hope of tightening the squares ever more finely on data and forecasts.

For now, weather forecasters can get all the meteorology right, but still get the weather outside your window wrong.

Weather Journal appears on Wednesdays.

Weather Journal

News tips, photos and feedback?
Sign up for free daily news by email