Snowfall simulation for the weekend from the GFS model. ( Ryan Maue Twitter feed, @RyanMaue )

But please don’t spread rumors that we’re calling for snow this weekend, because we’re not.

The reasons why snow is highly unlikely?

1) Forecasts of precipitation from computer models 6-7 days into the future are usually not very good.

2) This particular model seems to be overlooking the fact that low-level temperatures are probably too warm

The European model suggests temperatures will be in the 70s Sunday. (

4) It’s abnormally early in the season for snow. In fact, it would set a new record for earliest accumulation, which currently stands as Oct 10, 1979 with 0.3”.

Some other interesting facts to consider...

* The earliest instance of at least a trace of snow in D.C. occurred Oct. 5, 1892. At Dulles, it was Oct. 10, 1979 (1.3” fell), and in Baltimore, October 9, 1903.

* The October daily snowfall record in D.C. is 2” from Oct. 30, 1925. 0.2” fell next day for a storm total of 2.2” (biggest monthly single snowfall in D.C.)

* Prior to last year’s trace during the October 29 Snowtober-fest, the last time D.C. had at least a trace of snow in October was 1979.

In other words, the odds for early October snow are extraordinarily long in the current climate.

Don’t hold your breath just yet!

(Jason Samenow contributed to this post)