The Physics Of Snow Accumulation

Well, what a week it's been! Rain, snow, hail, graupel, frost, thunder, blue lightning, freezing fog, freezing rain - the list goes on, writes iWeather Online's Fergal Tierney.  

Evening sun sets as snow falls in Wexford town. Pic Declan Kearney
For the vast majority of the country, meteorological winter has well and truly started with a flurry (excuse the pun!), with satellite pictures showing the vast majority of the country blanketed by snow and ice. But why did the snow accumulate so quickly this early in the season? Shouldn't it have melted, like it seems to do most other times? With land and sea surface temperatures still usually not at their coldest for another 6-8 weeks yet, you'd have been forgiven to think "ah yeah, but it won't stick" when you woke up last Saturday morning, but boy would you have been wrong!

Snow is an extremely complicated form of precipitation, and as such leads to a very tricky time for a forecaster. There are two "types" of snow - wet and dry, and the particular type we get is important in how it reacts when it hits the ground. Wet snow occurs when the temperature in the Planetary Boundary Layer (roughly the lowest 500-1000m of the atmosphere) is fairly constant and near zero. Partial melting means it contains some liquid water, and is therefore more dense than dry snow. It will compact quicker, so for the same amount of liquid-precipitation, it will form a thinner layer than dry snow, with a liquid equivalance of between 10:1 and 5:1. It will also fall in larger flakes, because two or more flakes will agglomerate as they fall, and it will not drift to well in wind. This type is the best for making snowballs.

Dry snow occurs when the PBL is a lot colder, inhibiting this partial melt, and therefore yielding a less dense and smaller flake. The lower liquid content makes this snow difficult to make snowballs with, and will easily drift as it blows in the wind. Its lower density means it will form deeper layers than wet snow, with a liquid equivalent of between 15:1 and 30:1. As we were affected by an Arctic airmass, with low moisture levels, most of the snow fell as dry snow.

An important factor in setting the scene for the week was the fact that the first snow fell onto frozen terrain. A hard frost had set in Friday night, with air temperatures down to -3°C in the east, and grass temperatures a few degrees lower. This meant that anything that fell remained intact, building up a layer that would then further help to keep the ground temperature low, but also insulate any further snowfall from the remaining warmth radiating up from the ground. 
The dry nature of the snow layer Meant it contained lots of air, which is a good insulator, adding to the effect. This is also why snow accumulates more on grass than on concrete pavements, roads, etc. The grass blades are more exposed to cold air than concrete, which will stay warm for longer as it's in full contact with the ground's internal warmth. The air pockets around grass insulate the snow from the this warmth, allowing it to build quicker. Of course, the more it builds, the greater its bulk, therefore the greater the energy input required from the ground to raise its temperature and cause melting.

Another factor is that the snow fell at nightime.  With nightime snow there is no uv radiation present to cause sublimation of the snow as it falls, therefore more of it reaches the ground than when it falls in daylight. Dayime uv radiation also heats the ground and other surfaces, reducing the chances of it sticking, but with a covering of snow present, this radiation doesn't get a chance to heat the ground, therefore air temperatures remain low.

Now of course the synoptics have changed, and much warmer air has moved in behind yesterday's front. Temperatures and dewpoints rose above 0°C, kicking off melting of the snow layer. Once some bare ground becomes exposed, daytime uv radiation can accelerate melting around the edges, exposing more and more ground, which means more and more melting, and so on. 
The depth of lying snow reported at Casement fell from 21 to 16cms in one hour Friday afternoon as temperatures rose slightly and precipitation turned to rain. It will take some time, however, for all the snow to thaw, with nightime freezing fog lingering throughout the day in places during the next few days. Further snow is expected next week, especially over high ground, so it will be interesting to observe whether this  falls during the day or at night. And as it's a polar airmass this time, it will more than likely be wet snow, so will probably not lead to the same depths as of late.

Maybe that's where the term "Depths of Winter" comes from?