All you need is a cloud, precipitation, and an atmosphere below freezing for snow, right? Well, not really. There is a lot that goes into snow and it’s one of the reasons why forecasting snow is so challenging. While some just “rip and read” raw model output there’s a lot of value meteorologists can add to a forecast by closely inspecting the atmosphere.
Ice crystals cannot form in a cloud that is warmer than -4ºC. Most ice crystals grow through deposition which is where water changes from vapor to ice onto an ice nuclei (a subset of cloud condensation nuclei or pieces of salt, dirt, clay, etc.). In the Bergeron process the growth of ice crystals occurs at the expense of liquid droplets. This process is maximized between -12ºC and -18ºC.
This is important for several reasons. If you want snow you first need to make sure the cloud is colder than -4ºC. You need ice! Even if you’re subfreezing a cloud doesn’t necessarily have ice in it. For heterogeneous nucleation those ice nuclei need to be activated which won’t happen unless it’s colder than -4ºC. You really want the cloud colder than -6º or -8ºC unless there is seeder-feeder (another topic for another day) to introduce ice to the cloud from above. Once you know your cloud has ice you want to maximize the efficiency of the snow growth which occurs between -12ºC and -18ºC.
So how does this play into the Christmas Eve and Christmas Day event? First we need to determine how much precipitation this storm will produce. It isn’t much of a storm! The normally over-juiced NAM shows between 0.15″ and 0.30″ of QPF across Connecticut.
The NAM is ‘roid addicted model that is almost always far too robust with precipitation forecasts. The 15z SREF plume which shows us the accumulated QPF of each SREF members is always useful.
Notice there’s a fairly tight clustering between 0.1″ and 0.25″ of QPF. The ensemble mean QPF (solid black line) is 0.19″. The European model is a bit more paltry with just about half that amount – near 0.10″ of total QPF from the Christmas eve/day wave.
So is 1″-2″ a slam dunk forecast? Split the difference between the Euro and NAM, agree with the SREF mean, and slap on a 10:1 snow:liquid ratio? Maybe not. Let’s take a closer look at the 18z NAM and how it gets us to 0.20″ or 0.30″ of liquid. What does the atmosphere look like?
Here you can see, at least initially, the cloud extends up through about 550mb and intersects the -12ºC isotherm. There’s modest upward vertical motion through the bottom half of the troposphere and enough lift in a cold enough and moist enough column for a period of snow. Up to an inch, quite possibly. But look at what happens by 12z Christmas morning.
Yikes! The cloud is now much shallower and the coldest part of the cloud only reaches about -7ºC. That’s marginal for snow growth through deposition and there’s virtually no moisture above (you can forget seeder-feeder). If anything that looks like a freezing drizzle or snow grains kind of sounding. Certainly not getting much accumulation. As it turns out, more than half of the NAM’s precipitation falls into an atmosphere that looks like this. Yuck.
So while taking a consensus approach and blending model QPF forecasts is frequently a good way to go… in this case looking at how the atmosphere will evolve over time is just as important as the final numbers.
Here’s my snowfall forecast through Christmas morning. There is an area of better snow growth to the north (i.e. a bit colder through the cloud) so I have 1″-2″ along and north of the Pike.
As always it’s possible that by tomorrow the forecast soundings will change some. A heavier thump of precipitation into an initially colder atmosphere? A possibility. But at least for right now I’m not too excited about much accumulation by Christmas morning. That said, it won’t take much too get a coating or quick inch on the ground to put all of us in the Christmas spirit!