I'm not finished yet, but my preliminary results using one statistical method involving detrended temperature data showed a 16% reduction in snowfall overall.
Out of all snow "events" (anything over 0.5" I count as an event and omit the rest):
- 32% of them would have decreased snowfall in today's climate if they happened now than some X date in the past (e.g. an 8" snowfall in 1970 would be only 4" today)
- 15% of them would be "total losses" (complete rainstorms)
- 34% of them would see no change (similar accumulations)
- 19% of them would actually see increased snowfall with bigger accumulations
Of those 19%, I can highlight a few examples of very cold storms where a few degrees increase wouldn't flip it to rain, but would actually juice up the storm:
1) Feb 1979 (PD1) would have dropped 22" if it happened today rather than 16" at IAD
2) Feb 1983 would have dropped 30" today instead of just 23" (!!)
3) Jan 1996 would also have dumped 30" instead of 24"
Then as we get into more recent storms like 2010 and 2016, the effect would be minimal because the temperature difference would be minimal. Interestingly, in the first 2010 storm, day 1 yielded 4" less, but day 2 yielded 2" more, so the result was 30". The actual was 32.4", so the loss was minimal, but more on the front end of that storm when the BL temps were still warm-ish and lots of snowfall was lost to melting.
The 2016 storm had almost no change (obviously).
Now, on the flip side... how many major storms did we lose?
1) We lost a footer in Feb 1987... assuming my method is correct (more or less), that storm would be the perfect track rainstorm. Total shutout.
2) There were four 8" storms throughout the 1960s... and we lost them all! Four 8-inchers in the 60s got zeroed out. (though there were another 3-4 storms in the 60s that made up for those losses by adding more snowfall... one 10" storm would produce 15" today, for example).
So this was a qualitative look at how snowstorms from the past would perform in today's climate. It was exciting to share, even though I'm not finished with it yet. I haven't looked at ENSO, PDO, and other methods that may possibly affect snowfall like adjusting SWE, and so on. Still hoping to get full results out in its own thread here.