Slowly Getting Back to Normal

Current Radar 

The Forecast

Remaining Snow/Ice will refreeze tonight. The overnight low is 30°. Secondary roads are especially vulnerable because they are often shaded and less trafficked.

We don’t know the road conditions, we can only tell you how we think the weather might be impacting them. Consult traffic sources on Twitter (such as @TrafficJamSam) and the TDOT Smartway map/cameras.

Significant melting is expected Monday and Tuesday as temperatures warm into the 40°s. Additional rain is expected Monday night, which will help the melting process.

NWS does not foresee flooding as rain melts the snow. However, some creeks and streams may rise later in the week. NWS is watching it closely.

Even though temps may dip below freezing Tuesday night, we think almost all of the snow/ice should be gone. Warmer temps expected through the work week.

Snowstorm Review

Final Snowfall Totals range from 10″ in North Davidson County to 2″ in South Williamson County:

We got 8″ at the airport, which set a daily highest snowfall record for January 22. The snowstorm itself was the largest in Nashville in 28 years.

For a giant list of only some of the reports received by NWS all across middle Tennessee, click this link.

I went back through the models and forecasts over the last several days. Some takeaways:
  • They all did a pretty good job of distinguishing the existence of a line between big totals and small totals. This is a difficult skill in meteorology, but the models and forecasters pointed it out well. The hard part was figuring out where that line between big snow and “we were robbed!” snow was going to set up. The models looked for it. Some run twice a day, others four times a day — and you can certainly pick out one or two runs from each of the model and praise it, and ignore the other bad ones. For example, Thursday afternoon the Euro had it pretty much nailed:

However, other models, including previous runs of the Euro, did not have this solution.

Here are three model runs from noon on Wednesday. They had the right idea, but none of them put the “deformation band” (area of heaviest snow, in purple) in the exact right spot:

NAM12

GFS

Such is the state of the science.

  • ETAs for the change from rain to snow were bad. All guidance let us down on this. On Wednesday, the models were all over the place on the ETA of the changeover to snow. They differed by 12 hours – some models arguing it would changeover around noon, others very late Friday night. A developing consensus arrived Thursday saying the changeover would occur sometime around 10 AM to noon.

In a snow event in Middle Tennessee, the changeover is the most important thing, because once it starts, the impacts are felt immediately. ETAs of a change from rain to snow are very difficult to nail. This is why the weather community expresses uncertainty and margin of error, and discusses information bias, in these forecasts. We want you to know what we think, but we don’t want you adopt a certain ETA as if it’s truth, plan accordingly, then get stuck out in it without you knowing you’re taking a risk. Some of you are risk takers, some of you are risk averse, and we all act on a risk/reward basis. We don’t want to tell you whether to take a trip or not (for several reasons), we just want you need to understand the level of risk and uncertainty built into your decision, which you will make based on your risk tolerance and reward potential.

No one wants the ETA to “miss.” Ours missed by about 4 or 5 hours, depending on exactly where you were. The ETA for Joelton was bad, but it was pretty good for SE Williamson County. Such is life here in a mid-latitude city. The line is always going to get drawn somewhere. In Davidson and Williamson Counties, that line usually bisects us. That was certainly true with this event.

  • Your CrapApp is OK on a sunny day, but when you really need severe or winter weather information, it sucks. It’s clip-art of a sun or snowflake or cloud, a temperature, and a percentage the user doesn’t really understand. All of it is spit out by a computer (we saw how those did) and uncurated by a person with a brain. Your app treats snow forecasting as if it’s a hard science and easily knowable, when it’s everything and anything but. These apps train us to think weather is exactly knowable down to the hour (some apps claim they have it down to the minute), so it’s no wonder you’re demanding from us the same level of precision and accuracy.

It should be no wonder to the weather community when an exasperated public tosses up their collective hands to say a forecast was “wrong” or meteorologists are “bread and milk conspirators” or they’re just frustrated because they’re stuck on I-40 at 8:30 AM because the snow started before their app said it would.

The weather community should be providing you an invaluable discussion and layout of uncertainty and risk during severe/winter weather, not slapping down a snowflake and a number in the palm of your hand and calling it a day. Reviewing weather apps during severe/winter weather is like you going to the doctor with a gaping, bleeding wound, and the doctor looks at it and says, “rub dirt on it, then eat this bag of Doritos, good luck.”

You want an answer, a specific, definite answer, about when the rain will change to snow.

But,

And the truth is we aren’t there yet. Don’t let apps – or, for that matter, any winter forecast here or elsewhere – suggest the exact start time of a snow event is knowable and reducible to an exact time via a sunglasses-wearing sun and a mysterious percentage. Investigate. Embrace uncertainty. Weigh risk/reward. Build some margin into your day to account for the uncertainty. Understand what it is you’re dealing with. Profit.

This website supplements @NashSevereWx on Twitter, which you can find here.