How much would a more accurate weather forecast cost? For Kentucky, it was about $3M

Tuesday, February 5, 2013 at 12:11am

It’s often said winter weather in the South is magical.

Snow and ice are not easy to manufacture. There is a precise set of factors required, such a narrow range of Xs and Ys that the science indeed seems like wizardry on the rare occasion when it all comes together to turn the city white.

As hard as weather is to predict generally, predicting frozen precipitation might as well be necromancy.

Twice in the past month, meteorologists hinted that potentially dangerous ice storms would hit Nashville, crippling the city in the way only a glaze of ice can, clinging to trees and power lines, ripping them with its weight.

And twice, Nashville watched as a mere cold rain fell.

Ice is the toughest forecasting nut to crack, as it requires cold air aloft, a pocket of warm air just below that, and then freezing temperatures near the surface. Guessing whether those three ingredients will meet at a precise time in a precise place is not easy. And twice, the weathercasters got it wrong.

But, in reality, they weren’t that far off. The surface temperature in Nashville needed to be, of course, at 32 or below. And it never quite got there — the mercury would dance at 34, creep down to 33 and hold steady. One or two degrees Fahrenheit, in the great scheme of things, isn’t that bad of a miss.

Meanwhile elsewhere, west towards the Tennessee River and then again east towards the Cumberland Plateau, the air hit the magic number and ice fell.

The colder air belted the Nashville area, but never squeezed shut. And all through those two nights, the meteorologists had to guess when — or if — the colder air would arrive.

It never did. And they never really knew how close it was to the city.

Meteorologists can’t be everywhere.

The National Weather Service office in Old Hickory is responsible for predicting the weather for all of the Midstate. They themselves, the experts, can observe the conditions at their office. They can rely, with some degree of confidence, in the observations made by air traffic controllers at the larger airports or at Fort Campbell. And they can count on the TV weather people in Nashville to provide accurate information.

But if they need a temperature on the ground in, for example, Hohenwald, they are relying on a layperson.

And when it comes to events, like those two ice storms, when dead-eye accuracy is the keystone, those observations can fall short.

The weather service also employs expensive, fancy modeling equipment, which given a certain set of variables, can estimate temperatures all through a region. But as with any computer program, the output is only as good as the input — if it’s garbage in, it’s garbage out.

Meanwhile, Nashville sits at 33 degrees and rainy.

While it’s reflex to simply blame your local weatherman when a forecast misses — when plans are changed for a storm that never arrives, when panic befalls for a crisis that doesn’t occur — it’s not that simple. For a more accurate prediction, the meteorologists need better data.

Kentucky’s meteorologists have an advantage: a statewide network of 64 automated weather stations in nearly every county in the commonwealth. It’s called a mesonet.

In short, a mesonet provides real-time — updated every minute — data and has the ability to “detect smaller-scale weather features,” according to WKRN-Channel 2 meteorologist Justin Bruce.

“While mesoscale weather features can be detected on radar, the radar beam is often pointing a few thousand feet high into the storm on the radar’s periphery, and not indicative of actual conditions on the ground. A mesonet gives us physical observations and actual ground truth. This is exceptionally helpful in the world of weather modeling, where you can never have enough observed data to feed the computer’s system of equations,” he said.

Bruce said that comes in handy during winter weather events as well as more traditional severe storms. Meteorologists have the back-end capability of producing highly accurate short-term forecasts, but the process works best with lots of front-end information.

“Short-term high-resolution weather models … run on a set of grid points that are spaced just four kilometers apart. When you feed that sort of a model with observations from only Clarksville, Nashville and Chattanooga, you are theoretically missing out, because it’s equipped to handle so much more,” he said. “It’s like feeding Chris Johnson canned ravioli before a game instead of lean protein, or whatever it is that fast NFL running backs eat.”

And Kentucky has plenty of that lean protein.

Western Kentucky University professor Stuart Foster is Kentucky’s state climatologist and director of the Kentucky Mesonet. The project was funded via earmarks secured by Sen. Mitch McConnell — a total of $3 million. Foster wouldn’t disclose the annual recurring budget except to say it’s not seven figures.

“In order to really leverage your investment and actually derive the benefit from it, you are looking at a budget approaching $1 million. You can run it on less than that, and we are doing that right now. We can operate it and put the data out there, but we don’t have the resources to extract all the value,” he said.

Nevertheless, the mesonet has proven to be a valuable resource, and the feedback on the project is almost universally positive. Information is used by TV forecasters, farmers, county emergency management offices and law enforcement. And while pinpointing snow and ice is fun — especially for school kids — Kentucky’s National Weather Service offices find the extra data especially useful during thunderstorms.

“The wind data is what they like the most. When we are in a severe weather situation, the radar is telling us what’s happening in the storm environment but not on the ground. Before, we were blind at the surface,” Foster said.

But the mesonet provides virtual eyeballs from Paducah to Pikeville, an uncommon wealth of information.

Kentucky worked with the operators of Oklahoma’s mesonet, as well as U.S. government scientists at Oak Ridge National Laboratory to create a network and an infrastructure that best suits the state, its weather and its geography — all similar to Tennessee.

But Kentucky has one advantage Tennessee doesn’t — beyond being home to the U.S. Senate minority leader. Kentucky — and most every other state — has a state climate office able to coordinate the planning and later operation of the system. Tennessee’s climate office was dissolved in 2005.

“There are people working on [reviving a state climate office] and have a vision. There’s a possibility to do what we’ve done here in Kentucky. We would certainly be happy to work with people in Tennessee,” Foster said.

Unless and until that happens, forecasters like Bruce will be stuck trying to serve a steak dinner out of a Chef Boyardee can. And forced to endure the slings and arrows of missing a forecast by just a degree or two.

1 Comment on this post:

By: sonny1024 on 2/5/13 at 6:22

I don't watch the weather man on TV for weather forecasts I watch the rush own at KROGERS and then I know we are going to have snow or Ice I call it the KROGER FORECAST.