Saturday, January 26, 2013

A Mix on the Way? (And What Was KARE11 Thinking?)

NWS forecast as of Sunday mid-afternoon.
We haven't done many forecast assessments in a while because, for one, there hasn't been much snow in the forecast. Indeed, there's been a string of measure-by-the-millimeter snowfalls lately.

For much of the day on Thursday, forecasters were expecting the possibility of an inch or snow that could affect the Friday morning commute. However, by mid evening, most outlets had backed off that forecast, with many forecasting either nothing or just a dusting. And even a casual observer of the radar had not to notice that most of the snow had seemed to slide past to the southeast and northeast.

Notably, however, Belinda Jensen on KARE11 warned during the early weather tease on the 10 p.m. news  that up to an inch of snow was possible that would likely affect the morning commute. This was at the same time when all other outlets saw the storm's handwriting on the wall.

We can only wonder what KARE11, typically among the most conservative Twin Cities stations when it comes to forecasting snow, was thinking. Was it a steadfast belief that snow would materialize into something more than 10 flakes? Was it laziness in not accessing the weather data used by other outlets to produce a "call off the wolves" forecast? Who knows... but KARE11 viewers were decidedly not well served by the 10 p.m. weather forecast on January 24.

Switching gears.... As of Sunday mid afternoon, the NWS issued a Winter Weather Advisory for 9 a.m. Sunday to Sunday at midnight. The NWS had earlier issued a Winter Storm Watch that seemed rather puzzling to our uneducated eyes.

Beyond that, virtually all forecasters were predicting a return of the arctic cold by Wednesday night.

Monday, January 21, 2013

Going Negative

The much-advertised cold is here, and in our view it's been a very well-predicted cold wave. There's some talk among forecasters of a little light snow later this week, but clearly the word of the week is C-O-L-D.

A blog note: I understand there have been some problems accessing this blog and/or making comments. I'm note sure what's going on as I'm simply using Blogger as I always have. I've heard that there may be problems viewing it with Explorer so you may wish to try other browsers.

Stay warm!

Saturday, January 12, 2013

When it Comes to Snow Efficiency, Twin Cities a Dud

I’ve long suspected that the Twin Cities endures a tremendous amount of cold weather for a relatively paltry amount of snow. But the question always remained: how can you measure cold weather “efficiency” in producing snow?

By tossing around a few numbers and making some general assumptions, I developed an efficiency rating that measures “bang for the cold weather buck” when it comes to producing snow. The statistics confirm that when it comes to snow making and cold weather, the Twin Cities is like a souped-up car that struggles to reach 50 mph.

First, some background on my approach. I decided to measure “cold hours” as every hour of temperatures at 32 or below. That’s also the temperature at which legitimate snowmaking becomes a possibility. I divided the number of cold hours by the average annual snowfall (obtained here). This results in a snow-making efficiency factor expressed as cold hours per inch of snow produced.

I developed the numerator by reviewing daily averages for each of the cities detailed below. Here’s how I measured that. For a given day, if a city had an average high of 40 and an average low of 24 (a daily average of 32), I assumed that half the day’s hours were at or below 32. Using the same general assumption, if a city had an average high of 40 and an average low of 28 (a daily average of 34), I determined that 10 of those 24 hours were at or below 32.

I came up with a scale of cold hours per day based on the theory above. If a city’s daily average was 44 or higher, I assumed they had no “cold hours.” If the daily average was 20 or lower, I assumed that had 24 cold hours. It’s not a particularly scientific approach and I’m sure there are inherent statistical biases, but given my limited abilities and the desire to come up with a ballpark estimate, I think it works for the purpose.

The results were generally as I expected and probably won’t surprise most weather enthusiasts, yet I think it’s fun to see an actual number assigned. As you can see from the table, Erie, Pennsylvania must only endure 17.7 cold hours to receive an inch of snow. Rocky Mountain locations also score well in the snow efficiency rating. Aside from cities bordering the Great Lakes, the Midwest must endure the coldest temperatures for relatively little snow. The Twin Cities requires nearly 54 hours of sub-freezing temperatures to receive one inch of snow. For comparison, Burlington, Vermont, has almost as many cold hours as the Twin Cities, but receives 60 percent more snow. East coast cities are more “snow efficient” with cold weather than Midwestern cities, though obviously not as prolific as lake-effect locations.

Hardly surprising data to be sure, but hopefully rather entertaining for weather enthusiasts. The Twin Cities climate may produce one of the better theatre of seasons in the country, but when it comes to snow production, the long-duration winter – which consists of an estimated 2,675 hours of sub-freezing temperatures – is a relative dud.

Thursday, January 3, 2013

GFS Weather Model Eyes Super Arctic Outbreak

How much value is there in weather models, particularly one notoriously nicknamed "Good For Shit," that predict temperatures more than a week out? We're about to find out.

Two recent entries in blogs by the Twin Cities' "weather Pauls" noted that the GFS model predicts some mighty cold temperatures arriving in the middle of the month.

The Updraft MPR weather blog, authored by Paul Huttner, included this depiction of the GFS forecast in its edition yesterday (Wednesday) morning. Note the model includes a full week of all-day sub-zero temperatures with a low of -31F on Wednesday, January 16.

For more complete context, including Paul's thoughts on how likely it is that the arctic cold will materialize, refer to the full blog entry here.

This Thursday morning, Paul Douglas's Star Tribune weather blog included this GFS capture:

It's interesting to note the considerable difference in the GFS in just 24 hours. The coldest sub-zero temperature forecast is "only" 12 below compared to the 31 below in the previous day's forecast. Daytime high temperatures are similarly warmer in the GFS forecast from this morning.

As an interested but essentially uninformed weather enthusiast, I'd like to know how accurate the GFS model is for a forecast that many days out. While temperatures undoubtedly flip flop, does the overall trend remain? Or is it a case of the long-range GFS and $3 will get you a beer?

Your thoughts, dear Minnesota weather enthusiasts?