One of my blog readers Ralph Fato has a thing for stats and a passion for the weather. Over the past few months, he has crunched seasonal snowfall data for a number of cities, and cross-referenced them with the conditions of various meteorological indicies -- ENSO (El Nino / La Nina, which is measured by the SOI - Southern Oscillation Index), NAO, PNA, AMO, and PDO.

The chart above (click to enlarge) takes the ENSO status, and direction in which it is moving, and compares it to snowfall amounts for 13 cities (mostly in the Northeast U.S.). Possible conclusions from this chart, Ralph writes: "Pittsburgh, PA on average gets more snow during weak El Nino and 2nd most during weak La Nina. It also gets more snow on average while La Nina is strengthening."

I think it's fascinating that, in Fargo, ND, there is a clear decrease in snow as you go from La Nina to El Nino, but the opposite is true in Baltimore and Philadelphia. Unfortunately, the ENSO phase doesn't seem to matter in any statistical way for the other cities.The next chart is the same thing, but mixing in the NAO phase, which unfortunately makes it more complicated:

Here he points out "Baltimore gets almost double the amount of snow during negative NAO. Philly gets 56% more. NYC gets more snow during a Negative NAO in El Nino. And least amount with a Positive NAO and La Nina." Although clear progressions are not present in this chart, there are some takeaways (for example, Pos LaNina is generally bad for East Coast snowfall).

So why look at this data? Well, one of the three ways you can forecast the upcoming winter is to look for past winters with similar index values. The relationship of these to seasonal snowfall is not known, but he has come up with some interesting data. Ralph's latest graph is based on the current state of our indices this Fall:

This, of course, is subject to change but right now Ralph translates: "There have been 11 Weak La Nina Years since 1950. 4 of them with a Warm AMO. Only 2 had both a Warm AMO and a Cold PDO combo like we have this year. Most shockingly all years with a weak La Nina had a Negative NAO (using November-March average)." There are two things that are apparent to me from this graph, if you want snow, pray for a Weak La Nina / Warm AMO / Warm PDO, Negative NAO / Negative PNA. The opposite is true (except for Pittsburgh) for +AMO, -PDO, _NAO, -PNA.

Looking at his above research for State College, PA (home of AccuWeather HQ), we get these numbers:

Ralph says: "With a Moderate El Nino and a Cold PDO State college gets 2 feet of snow on average compared to the 4 foot average. With a weak El Nino and a Warm PDO you get 5.5 feet on average." With this year's conditions (weak La Nina and cold PDO) we can expect an average which is very close to our normal snowfall.

So what does it all mean? Research into these types of relationships is just beginning in the meteorological community, and I'll caution readers that "correlation isn't causation" -- there are other things involved that could trick you into thinking there is a valid relationship, including climate change, other indices, "persistence" and potentially even solar activity. There are also more indicies than these.

Additionally, some of the numbers above are too small to be statistically relevant (for example the State College listing above for an average of 5.5 feet was only four seasons and varied wildly from below (36") to way above average (89"). And finally, none of these relationships will become true unless the indices trend in the directions we expect.

Report a Typo