Before being able to make an accurate forecast, one must understand the history of a location. This history can provide bounds to a “guess” and more importantly, it gives insight on what is “normal” at a location. Because I live in the South Eastern Pennsylvania area, I will focus on Philadelphia and Allentown but part 3 of this series will include a more national view of my winter forecast.
I’m not sure if there is a word I hate in meteorology more than the word “normal”. I prefer the term average (my statistical friends please hold off from letting your brain explode). Some might say “But Dennis, isn’t the normal the same exact thing as the average?” no…no it’s not, at least not always. I will always try to distinguish between the terms normal and average when I show data on this site.
The “official” 1981-2011 snowfall normal from the NWS (Source):
The “official” 1971-2001 snowfall normal from the NWS (Source):
I can feel the pain in your head right now. But how can normal change? Normal in this case is redefined every 10 years to include the most recent 30 years. And it just so happened that the past 30 years have been snowier than the previous 30 year period. For the remainder of this series, I am going to use the 1981-2011 definition because that is what the local news outlets will most frequently cite.
When do you expect snowfall to fall? Some people hope for a white Christmas others love the March super-storms, I love a good presidents day storm myself. But when does snow actually fall in our area?
In both cities, February is the big month. Interestingly though, if you compare the new normals versus the previous normals, you would notice that in Allentown January is actually the bigger snow month. Also interesting is that the “tail months” (Oct, Nov,Mar,Apr) have all decreased in both cities.
Current Forecast: A normal year (well what did you expect after reading that?)
Although this does not pose much in a forecasting sense, it would give you some insight into what the “maximum” snowfall we can expect.
Because I have enough of a history, I can say that only one year in Philadelphia was 3 standard deviations above the average from 1884-2001. The large dataset (http://www.fi.edu/weather/data2/wthrsnow.txt) also tells me that to be a top 10 storm, only about 1.5 standard deviations needs to fall.
Last 5 Years
When trying to make a forecast using only history, the last 5 years can provide the most insight into what could happen this year. Looking at how much snow fell and comparing to average can be useful in determining a forecast for the current year. For example, if every year in the past 5 has been below normal, well then chances are this year will be normal or above and vice versa. This is rarely the case. Only a handful of times in the past 100+ years has there been 5+ years of above or below average snowfall, including two 8 year stretches of below normal snowfall. One between 1924 and 1932 (interesting time frame) and between 1968 and 1976.
So what can the past 5 years tell us?
So, no such luck in every year being above or below. In fact, Allentown has trended almost exactly normal over the past 5 years while Philadelphia has trended well above (both at 31.2 inches). So using this information, the following forecast can be inferred:
Allentown will be near normal
Philadelphia will be below normal
But, what if we give the previous year more weight?
Last year was incredibly anomalous It started strong with almost 7 inches falling in Allentown on Halloween! Wow! Snow lovers ere excited, parents were worried the kids would be in school until July, human sacrifice, dogs and cats living together, MASS HYSTERIA! (sorry, was channeling my inner Peter Venkman) . Then, the winter came and went with almost no snow falling. In fact, the Franklin Institute data shows that it was the 4th least snowfall since 1884! So this must mean that this year will be a shoe in for a snowpocolypse…
Well…not exactly. 4.1 inches of snow fell in 1930-31..and the following year a massive 8 inches of snow fell. In fact, out of the 7 winters in which less than 5 inches of snow fell, only 2 have followed with above average snowfall (20 inches in 1973-74 and 29 inches in 1919-20). An extreme two year snow drought was experienced when in 1949-50 2 inches of snow fell and was followed by the 1950-51 winter which had 4.6 inches of snow fall. This information tells me that you can not always use the previous year as a indication of what is to come.
Forecast based on Local History
Ouch…not much help here. But based on local history my forecast would be the following.
Philadelphia 16.3 inches (below average)
Allentown 32 inches (Average) (Meaning that I am ignoring last year)
Note: THIS IS NOT MY OFFICIAL FORECAST YET. This is simply what I’m thinking by looking ONLY at history and at nothing else happening in the world. Meaning I have not looked at ocean/atmospheric patterns yet (part 2 of the series) or anything else for that matter. My official forecast will be issued in part 4 of this series. Expect that sometime in the next few weeks.
I am happy to answer questions in the comment section of this post or by e-mail if you feel better with that method. I love discussion, it is the best way for anyone to learn. Just know that at this point there is no right or wrong answer.