Five years ago, we were in the meteorological pits.

Western New York was headed into the final stretch of not just its coldest February on record, but also its coldest month on record. (Buffalo records go back to 1871.)

We can take a sampling from this time of the month in 2015. On Feb. 15, we had a high of 2 and a low of -10; on Feb. 16, a high of 5 and low of -10; on Feb. 17, a high of 10 and a low of -6; on Feb. 18, a high of 19 and low of 8; on Feb. 19, a high of 12 and a low of -6; and on Feb. 20, a high of 1 and a low of -8.

Daytime highs hit 32 on Feb. 4 and Feb. 7, but they never went above freezing. The average monthly temperature was an all-time miserable record of 10.9 degrees. Snow depth at the Buffalo Airport Observatory ranged from 21 inches to 26 inches between Feb. 5 and Feb. 28. Only 10 of the 28 days had light winds, averaging under 10 mph. With the prevailing temperatures in this brutal month, the once-in-a-lifetime Blizzard of ’77 could have been about duplicated, had an extended high-wind event occurred. All the snow on the ground and the Lake Erie ice was low-density and could have been easily sent windborne.

The upper-air pattern that made such incredibly harsh conditions prevail typically looked like this (from Feb. 18, 2015).

A lobe of the weakened polar vortex had dropped down to the northern Great Lakes. The shape of the contours and the wind barbs at 18,000 feet tell most of the story, with little embellishment from me necessary. This was a byproduct of a blocking pattern that has been absent this winter. The polar vortex has remained strong and fixed closer to the north pole most of this winter since December. Colder than average near-surface temperatures since Dec. 1 have been hard to find, as seen in this Alaskan/NOAA climate research graphic. Much of Alaska and a part of the Northwest Territories have been the singular exception:

In the near future, following a very cold Thursday in the low 20s, readings will edge up to the low 30s Friday and move back to and over 40 this weekend into early next week. Current indications keep the mild air in place from Saturday all the way into Wednesday next week.

However, there are some strong hints a colder pattern will be setting up to finish the month and start March. The European/ECMWF model is showing readings returning to much below average by late next week.

Of potentially greater significance looking this far out is the American GFS ensemble mean, which is a blend of 21 separate runs of the GFS model, each with slightly different starting conditions. This blend becomes more useful because it contains far more sampling, and that’s very important looking beyond seven to 10 days. Here is an ensemble projection for March 3.

The Canadian ensemble mean also supports much of the northern two-thirds of the United States’ being dominated by cold air once again, so there is good agreement between it and the American ensemble.

Long-range specialist Dr. Judah Cohen also points out an American model called the CFS v2 is projecting a cold trough over the northeast United States in March.

I have to note that this model, the CFS v2, is not the most reliable in our forecasting toolkit. It does tend to have a warmer bias quite often, so if it’s showing cold it’s worth keeping in mind.

For a second opinion, the National Weather Service Climate Prediction Center tweeted the following late Tuesday:

“Despite the warm winter for most parts of the Lower-48, winter does not appear to be going down without a fight! Below-normal temperatures are favored across much of the country in today’s Week-2 outlook as we look to close out February and begin March.”