Box AFD (Nash) with quite the self-deprecating, open voice... and a message to hobbyists to boot... some excerpts:
Models: still quite a bit of variation and the models are
struggling with exactly how a northern stream disturbance
currently diving south along the Ontario/Manitoba border and
southern stream energy across the ArkLaTex region will merge
later tomorrow...
A 50 mile difference in track, or 6 hour timing difference in
development means all the difference. What I`m trying to say
here is make sure you have big error bars around the forecast.
This is a great situation where probabilistic forecasts are much
better than a single deterministic one. But since I`ve got to
make a deterministic forecast, we`ll just have to work with it.
Scenario:... NAM is believed to be a little too far east. More or
less went with a 12z Canadian/ECMWF concept...
Snowfall: ...I went with about
7-8:1 ratios during the day Saturday. As we get into the
evening, rising to 9-11:1. For those playing at home and viewing
snowfall predictions from online weather models, please be aware
that most use a strict 10:1 ratio which will get you into
trouble more often than not. Although I`m going with the lower
snow ratios, there is an unknown if we end up getting a
mesoscale band or well developed comma head forming. If that
were to happen, that means we`ve got strong vertical motion in
the snow growth zone resulting in large dendrites and a much
higher SLR. After putting everything in a blender, I came out
with amounts over 6" across higher elevations of MA with 2-4"
widespread at lower elevations. Perhaps an 1" as you get just
inland of the coast.
Chances of my forecast being perfect?
Pretty low. Snowfall amount probabilities suggest the potential
for 10"+ across the higher terrain. Later forecasts will
hopefully fine tune things.