Upcoming Early Autumn Chill: A Hint for Winter?

[donsutherland1]

In his "Autumn Moment," Carl Sandburg wrote, "I cried over beautiful things knowing no beautiful thing lasts. The field of cornflower yellow is a scarf at the neck of the copper sunburned woman, the mother of the year, the taker of seeds. The northwest wind comes and the yellow is torn full of holes..."

This week the northwest wind is coming to the Northeastern United States with a vengeance. As a result, the region will experience its sharpest chill since last spring. Readings in all the major cities will dip below 50 degrees for the first time. A number of outlying areas, especially those in the normally colder valleys and also across central and upstate New York and central and northern New England will witness their first frost.

In those areas, what "yellow" remains will be "torn full of holes."

While the first frost is a regular rite of the autumnal season, the depth of the coming chill after a benign late September will seem all the more rude precisely because of the pleasant temperatures that preceded it.

If there is to be comfort found, cold comfort at that for some, it is that the weather has been even more wintry in the past at this time of year.

For instance on September 29, 1821, Worcester County, Massachusetts received some snow.

On October 2, 1899, Boston saw snow flurries fly through the air.

On October 3, 1769, Philadelphia experienced a period of snow.

The oncoming chill will not be comparable to those historic events. Neveretheless, it will be unseasonable. And, as the pattern becomes more progressive by the coming weekend, the chill will abate.

Even as it fades from memory, the early chill may offer some subtle hints for the upcoming winter. Culling the hints from the range of data, of course, can be challenging.

A number of winters that resulted following similar occurrences could be characterized as follows:

- Cold and snowy
- Warm and wet
- Seasonably cold but dry

In most cases, winters (December through February) that wound up warm and wet were preceded by unseasonably cold autumns (September through November). On the contrary, in most cases, winters that wound up cold and snowy were preceded by autumns that saw normal to above normal readings. Winters that wound up seasonably cold but dry typically saw near normal autumn readings along with above normal precipitation.

Out of the sample of years in which unseasonable cold shots occurred in the first days of October, four have been chosen only for purposes of further examination:

1945-46: Mild fall led to cold and snowy winter: 31.4" in NYC
1947-48: Warm fall led to cold and snowy winter: 63.2" in NYC
1974-75: Very chilly fall led to warm and wet winter: 13.1" in NYC
1979-80: Mild fall led to near normal and very dry winter: 12.7" in NYC
1992-93: Very chilly fall led to near normal winter: 24.5" in NYC

They were chosen mainly for their timing and not other factors such as ENSO.

Following the early chill in 1945, record heat occurred a few weeks later. The same thing happened in 1979. In contrast, in 1974, an additional shot of record cold followed a few weeks later.

Right now, given fluctuations in the SOI, it appears unlikely that record heat will occur through at least the middle of the month. Nevertheless, a decent warmup appears likely due to the deamplification that will take place beginning this weekend.

In terms of how their summers compared with Summer 2003 in the Northeast, the following data is available:

Summer 2003: Near normal temperatures; normal to above normal rain
Summer 1945: Near normal temperatures; above normal rainfall
Summer 1947: Above normal temperatures; above normal rainfall
Summer 1974: Near normal temperatures; normal to below normal rain
Summer 1979: Normal to below normal readings; below normal rain
Summer 1992: Much below normal readings; above normal rain

In terms of ENSO, 1945-46 and 1947-48 were Neutral winters. 1974-75 was a La Nina (C-) winter. 1979-80 saw the onset of a weak El Nino (W-). 1992-93 was an El Nino winter (W-).

All said, such winters (including additional ones not highlighted) that remained neutral with regard to ENSO typically saw above normal snowfall. Those that saw the onset of a weak El Nino after a neutral fall saw below normal snowfall.

Right now, analogs and statistical forecasting suggest that Winter 2003-04 will remain neutral with some possibility of the onset of a weak El Nino. It is not likely that a La Nina will develop.

For now, the possibilities for the winter remain plentiful and leave much to consider. In about two weeks, the hurricane analogs should offer an additional hint.

[Stephanie]

Thanks for the analog information. I'd say from all of the information that we've been given so far, I'm going to go with an "average" winter temps and precipitation wise for the Northeast.

[TexasStooge]

I've already felt a slight chill this morning, they said it was 63*F at 6:00AM, it felt colder than that.

[JCT777]

I can't wait for the first temps below 40. We had a day last week with low temps in the mid-40's. It felt great!

[Colin]

We should feel that sometime later this week John...especially far N&W of Philadelphia...we shall see!

[donsutherland1]

JCT777,

This week should see the mercury slip below 40 degrees for the first time this season in Spring Mount. You may see some frost, too.

[donsutherland1]

Stephanie,

Whatever happens should be interesting.

At this time, I'm honing in more closely at certain analogs (some of which are favorable; others which are quite disappointing). The hurricane analogs will offer a piece of the overall puzzle.

Certainly, from this vantage point, a number of winter forecasts/preliminary forecasts are already available. They range from cold and snowy east of the Mississippi River to mild with below normal snowfall in the Mid-Atlantic region.

The years listed at the beginning of this thread show that one should not necessarily reach immediate conclusions based on the early cold spell. However, one should begin to watch how things evolve and certain kinds of evolutions have led to differing outcomes. The neutral to weak El Nino vs. neutral all winter situation is also revealing.

[JCT777]

JCT777,

This week should see the mercury slip below 40 degrees for the first time this season in Spring Mount. You may see some frost, too.

It got down to 40 degrees this morning, not too far from having frost. Friday morning will likely dip below 40, and I would not be surprised to see frost then.

[WxDen]

With the now projected possible impending weak El Nino by Mid-winter I'm in hopes of another snowy winter here. Good research on the analogs Don. However, there is one that still deserves a mention at least. 1977-78. That year also featured a colder than normal October. If you base it on Octobers alone around here. '76-77 and'77-78 were very much below normal and both winters went on to be some of the coldest and snowiest on record for this area. Of course alot of other influencing factors were different. But, if that El nino were to come on;somewhat more similarities. The winters '72-73 and '74-75 were horrible here too. Each had a cold October and then mild winter. 72-73 was a strong el Nino though. '74-75 followed a La nina year, which HM explained and detailed in a post over on wright. So, actually things are looking more promising I think.

[donsutherland1]

WxDen,

The analogs I referred to here--and are not necessarily going to be relevant through the winter--only pertained to other years in which the first few days of October had a sharp cold snap, mainly to show that one cannot reach immediate conclusions based on such weather.

I did offer a few things to look for in terms of mileposts as to how things might be evolving.

I agree that 1977-78 is in play for the overall winter at this stage (as is 1960-61), though Autumn 1977 had a weak El Nino. Autumn 2003 likely will average out with a neutral ENSO signal.

I fully agree with respect to HM's post. It was an outstanding post and an excellent reference.

At both WWBB and here, I'll post my idea for the winter, probably around mid-October, as one of the factors I consider pertains to hurricane analogs (bullish at this point) but this doesn't yet factor in Larry or any other storm that might develop in the future. QBO does look East through December and maybe January.