5 March 2013

Why It’s So Difficult To Forecast Snowfall

Posted by Dan Satterfield

Numerical weather models are all indicating a significant coastal storm on Wednesday and Thursday along the Mid-Atlantic and Northeastern U.S. Coast.

Big Storm Brewing 

March looks like it will come in like a lion this week with a major storm likely to cause significant disruptions from Northern Virginia to Maine. The heaviest snow will probably fall just west fo the I-95 corridor but it will be a heavy wet snow, and it may be heavy enough to cause major power outages. Some guidance is indicating over 2 feet of snow and that might even cause structural damage to some buildings. Two feet of heavy, wet snow on a flat roof is a LOT of weight.

Along the coast there may be snow, but the wind and coastal flooding will be the main story. Winds could be gusting to over 60 mph from the Virginia Eastern Shore to Long Island, Wednesday into Thursday. The high tide at Ocean City, MD is just after 3 PM on Wednesday and at that time an onshore wind of 35-45 mph seems likely. This will probably be the strongest coastal storm for the Delmarva area (where I am) since Sandy and it will rival the February Boston blizzard over New England.

Why Predicting Snow Is So Difficult

Forecasting snowfall amounts is by far the most difficult thing a forecaster has to contend with and I thought I might list a few of the many factors that have to be considered.

1. Precip. Type and Thickness. Even if you are confident in the track of the storm and the amount of precipitation, you still have to decide if it will be rain or snow. The surface temperature is less of a factor here than the average temperature in the bottom 10,000 feet of the atmosphere. Meteorologists tend to look at the distance between two pressure levels (called the thickness) to decide on precip. type. Usually the best indicator is what we call the thickness. You can think of it as the average temperature of the layer between two pressure levels. For coastal areas the preferred thickness to look at is between 1000 millibars (pressure near the surface) and the 850 millibar level (around 1500 meters above the surface).

If the thickness is below 1300 meters then snow is possible, but if you are very near the ocean, I’d want it lower. In mountain areas, the thickness between the 1000 millibars (mb) and 500 millibars is more often used, and the magic number here is around 5400 meters. Below that you will see snow. In higher mountains 5460 meters  or lower will give you snow. You should also look at the thickness between the 1000 mb level and 700 mb level (around 3km above the surface) and if both are below the snow threshold, it is a good bet that you will see white flakes and not raindrops. If the temperature from the ground up is below freezing then it’s easy- it will be all snow.

2. Surface Temperature.  Okay, so it is going to snow, what happens when it falls. Here is where the surface temperature is important. If the temp. is below freezing then it will stick more than likely. If it is 1-2 °C then it will mainly melt, but if it snows very hard, then it can accumulate and in some cases get quite deep. I’ve seen a foot of snow fall with the temp. above freezing. If the surface temp. is above 4°C the snow falling may melt to rain anyhow!


Those blocks of snow may look heavy but they were very light because Antarctic snow is very dry. The rain to snow ratio is a significant factor that forecasters must consider when forecasting snowfall.

3. Ground temperature. This is yet another factor. Has it been warm recently and if so the ground will melt most of what falls as it reaches the surface. If it has been cold then even if the surface temp is above freezing the ground may be frozen and the snow will likely stick. The worst ice storm I’ve ever witnessed was in Winnipeg,Manitoba. It started raining at 3°C, but the ground was frozen solid, and the roads quickly became a solid glaze of ice that was so slick you could not stand on it!

4. Dendritic Growth. For really good snow events you want the snow flakes to grow in an area where there is the best temperature and where rising motion in the atmosphere is greatest. This can give you a clue to how heavy the snow will be. This gets a bit complex so I will leave it at that. The best lift in the atmosphere will be determined by the formation of fronts in the troposphere (not just at the surface). The term meteorologists use for this is frontogenetical forcing (I told you it gets complex). If you’re really a glutton for punishment go to Google Scholar type in Conditional Symmetric Instability!

(Addition) 5. Snow to Rain Ratio. One last thing here, and that is the old rule that ten inches of snow equals an inch of rain. In reality it varies great;y from 3 to 1 in warmer situations to 25 to 1 in colder and perhaps much higher! So even knowing how much liquid equivalent rainfall will occur still leaves a forecaster with determining what the likely snow ratio will be! In Antarctica survival school I could easily pick up a huge block of snow to build an igloo because the liquid equivalent in the dry snow was so low, it was like holding polystyrene (Styrofoam).

This is just one forecast from the NWP models of snow totals. Hopefully, after reading this post you will understand why this is just a small piece of the puzzle and should not be taken literally! Rely on your local NWS office or a trusted on air metr. in your area.

In many events, a difference in one or two degrees can change a snowfall forecast by a dramatic amount. This is why I have a rule to never forecast snowfall totals more than 24 hours from an event. We actually (IMHO) do a pretty amazing job of getting the snowfall basically correct most of the time. Look at the Boston Blizzard for example.

There are a lot of folks on the internet who know how to look at the basic weather models, and tend to repeat whatever the raw guidance shows.  Even a few folks on TV who do that, so be careful when someone tells you that your going to get a big snow. Find someone who has given you decent forecasts before, and rely on that. Oh, and if they get it wrong, hopefully you have a good idea of just how hard it is to forecast snow.

If your on Delmarva, I hope you will watch us at WBOC, but here are some meteorologists I know who are very good at what they do, and know their stuff! These folks all work with the excellent NOAA forecasters and use their many years of experience to give you the best forecast available in their area.

Richmond- John Bernier

Lynchburg area- Sean Sublette

Boston- Harvey Leonard

Washington DC- Bob Ryan