Climate Change & Cyclical Snowfall Averages

[Thundersnow]

I was just watching a report on WTVF (Channel 5), where they were talking about what might be expected for the rest of the winter... generally above normal temperatures and above normal precipitation.

What I thought was interesting is that Kelly Cox said that Nashville's yearly snowfall average is 8.3 inches.

Several years ago, I remember that the average used to be 11 inches.  Then, I would occasionally hear what the average was... 10 inches... then, more recently 9 inches.  Now, we're down to 8 inches.

That proves that we're not all just imagining it.  We're not getting as much snow as we used to.  And, as time goes on, that affects the long-term average. 

So, as we see less over time, statistically, we expect less. 

Maybe that cycle or trend or whatever it is will reverse at some point... we can only hope. 

[Coach B]

I've noticed the same thing.  I think I've heard where they go by the 30 year average instead of the all time average.  Maybe by the time we bottom out on the average it'll turn around and head in the other direction. 

[Thundersnow]

I kinda like crunching statistics in order to look at trends.

Our friends at the NWS here in Nashville put out a handy historical list of all monthly and yearly snowfall records in Nashville since record-keeping began in the season of 1884-1885.

That data can be found here:  http://www.srh.noaa.gov/ohx/climate/monthsnw.htm.


Well, I did a little calculation in a spreadsheet with that data.  Climatological averages are typically based on 30 years worth of data at any given time.

So, what I did was set up a rolling 30-year average of the snowfall beginning in the year 1914-1915 (which marked 30 years after record-keeping began).  The span of time ends with last year (2006-2007).

Attached is a line graph generated from these averages.  In this, you can see some obvious trends.

Since the 1980s, we have been going down, down, down.  Our 30-year average is now about where it was in the 1930s and 1940s.

Along about the 1960s, we seemed to go into a "climate change" cycle where our snowfall average went up considerably.  30-year seasonal averages stayed up (around a foot of snow per year) during the 1970s and 1980s (mainly because of some very snowy years in the early 1960s and then again in the late 1970s... which spiked the average).

But, since the end of the 1980s, we have been losing ground rapidly (the snowy years of 1996 and 2003 have helped us out a little, as you can see with graph toward the end).

What may make you feel a little better is that we have been here before.  Our 30-year annual snowfall average is not yet quite as low as it was in the 1930s.  If the cycle should ever repeat, then perhaps we'll go into a snowier trend like the area saw, beginning in the 1960s.

That doesn't appear to be happening this year though (yet anyway).

In the graph below, the numbers up the left side (Y-axis) are the snowfall averages in inches.  The years are shown on the bottom (X-axis).

[ Guests cannot view attachments ]

[Ron_Jarrell]

Interesting. Very nice info TS.

[Woodvegas]

Good info. What concerns me is that maybe the 60's, 70's, and 80's were an anomaly and what we are experiencing now is our true normal or average.

[Thundersnow]

Possibly, Woodvegas.  But, if you look at the early part of that graph, it appears the 1930s-40s low point had come off a slightly higher cycle before that in the late 1800s/early 1900s.

It's too bad the info doesn't go back a couple of hundred years.  We might really notice some revealing trends.  In the window of ~120 years, though, it does appear that there are some up and down cycles.  Yet, I agree that the 60s and 70s appear to be an anomalous period of more snow in the grand scheme of things on the graph.

However, history tends to repeat itself.  Who knows if that ever happens again in any of our life times though.

[Snowman]

Very nice writeup Thundersnow. Very Interesting.

[Thundersnow]

Thanks to ctbpharmd at Talkweather, I have calculated a similar graph for Memphis from the data provided at http://www.srh.noaa.gov/meg/memsnow.php:



Similar trend as Nashville... except that the greater snowfall averages in the late 1800s/early 1900s is more pronounced. But, there was a decline in between the 1920s and the 1950s (just like at Nashville). The increased snowfall averages between the 1960s and the 1980s is then evident... followed by the decline we're now experiencing.

[Mathman]

Just for fun (?!) I did a sine regression for the cumulative snow data on the noaa website and came up with some interesting results.  Although the predictor function does not do a very good job forecasting snowfall for any given year (you would not expect it to since there is so much year-to-year variation), it does match the overall pattern very well.  For example, it picks up on the maximum which ocurred near 1920 , the minimum which occured near 1950, and the maximum which occurred about 1980.  For what its worth, it predicts an upturn in snowfall after 2005, with another peak (of about 12 inches) in 2040.

Here is the formula for any of math nerds out there 

y = 4.159477sin[.106344 x -.789182] + 8.020243

(Time is measured from 1890.)  It predicts 5 inches for this year.   

[Thundersnow]

Interesting, Mathman.  By that reckoning, this starts to look like a 30-year cycle in swinging from low point to high point and vice versa.

If 1920 was a high point, and 1950 was a low point, and then 1980 was a high point... then that would put 2010 at about the low point before we start climbing again to the projected high point in 2040. 

That means, folks, that we only have at most two years left in this snow drought, before averages start going up again.     

[Mathman]

Thanks... I thought it was pretty neat.  The pattern suggested by the data coincides farily well with North Atlantic sea surface temperatures, too.   Mean SST's were high in the 40's and 50's and are quite warm again. Patterns like these  can help us learn whether we humans are really affecting global temperatures the way some say we are.

[Clay]

y = 4.159477sin[.106344 x -.789182] + 8.020243

...wait, which letters do you substitute for. What do you plug in for "X".

[Mathman]

The letters "sin" stand for the sine function.  The only variable in the formula is "X" which stands for the number of years since 1890.  For example, if you wanted to know the prediction for 1990, you would let X=100.   Be sure your calculator is in radian  and not degree mode.

Just remember that the predictions made for any given year are not very important, since the data the function is created from are very "noisy".  It is more interesting to graph the function (try a TI-83 plus for example) and overlay the noaa data to look for trends.

I'l be happy to tell you more if you want to play with it.

[Thundersnow]

Then, I guess it would be fair to say, Mathman, that we have a fairly high standard deviation in the statistical sample we're talking about here.

[Curt]

Thanks to ctbpharmd at Talkweather, I have calculated a similar graph for Memphis from the data provided at http://www.srh.noaa.gov/meg/memsnow.php:



Similar trend as Nashville... except that the greater snowfall averages in the late 1800s/early 1900s is more pronounced. But, there was a decline in between the 1920s and the 1950s (just like at Nashville). The increased snowfall averages between the 1960s and the 1980s is then evident... followed by the decline we're now experiencing.

Hey I am now posting here as well! I figured there werent too many from Memphis-WTN posting so I thought I wold chime in (although I may be moving to Nashville soon to be closer to the fiancee!).
Didnt know that TN Wx Zone even existed until 2 weeks ago.

In reference to this subject, it is amazing how closely Memphis and Nashville mirror each other in terms of cyclical trends, exceptbeing further north, Nashville naturally gets a little more snow. It would be interesting to see if the same snow storms have affected both cities, and which havent. These cyclical changes seem to occur in 20 year increments, so one has to believe we are on the threshold of a new cycle thank God. I dont believe global warming has played a part in the decrease recently-I truly think this is a cycle. Hopefully we will see a a gradual increase in the next couple of years.