How Social Media Can Mislead You During Winter Storm Forecasting

Each winter, t​here always seems to be a race for who can put out a snowfall forecast the soonest, but an early forecast for snow can often be wildly inaccurate.

W​here the heaviest and most snow will fall can vary widely from one model run to the next. H​ere's what you need to know when looking at snowfall maps on social media.

(​MORE: Winter Storms 101)

Patience is needed with snowfall forecasts: The eagerness to know the impact a snow or ice storm will have is understandable, but sometimes, that question can't be answered until within one to three days of its arrival.

This thirst to know what will happen can often be fed by those computer model forecasts of snow amounts that go viral on social media. Sometimes, you'll see these maps with pretty blue, pink and purple contours a week or so before a storm's arrival.

There's nothing that can be done about these maps; they will never disappear from social media. But you can do your part by not buying into what they are showing and not spreading them like a nasty virus.

Meteorologists might see signs of a significant winter storm five to seven days in advance, but it's too early for specific snow forecasts: For instance, there might be forecast models suggesting on a Monday morning that a winter storm could affect several regions of the U.S. the following weekend.

Instead of showing specific details, forecasters usually just give a heads-up that we are watching for a particular storm in a general time frame. This is typically followed by the caveat that the forecast will change and we will provide more details as they are available.

Without robust data, numerical forecast models could have difficulty resolving important details of how a winter storm may come together from that incipient disturbance(s). It's a basic concept of forecast models that initial errors in analysis or short-term forecasts grow with time.

Throw in questions surrounding the location and depth of the cold air source, and there are myriad uncertain scenarios that forecasters face when dealing with a winter storm in a long-range forecast.

(15-min details: For even more granular weather data tracking in your area, view your 15-minute details forecast in our Premium Pro experience.)

T​hose uncertainties lead to the flip-flopping often seen in long-range snowfall forecasts from computer models: For example, you might see a map posted for your city on social media that shows 8 to 12 inches of snowfall (model run No. 1 below) from a storm that's still five or more days away. That same model could change its tune and forecast 3 to 5 inches in its next update 12 hours later (model run No. 2 below).

On top of that, those maps often don't account for how much snow can be produced from the forecast amount of liquid equivalent precipitation, also known as the snow ratio. Snow totals on the maps can also be inflated in areas where sleet is expected rather than snow.

Y​ou can see the difference in forecasts in the maps below. These forecasts were made less than a day apart and have huge discrepancies in where the highest snowfall totals will be. One forecast shows up to 18 inches of snow in coastal North Carolina, while the other model shows very little snow there; instead, it's shifted inland.

H​ere's what to do next time you hear that there is a possibility of a significant winter storm in your area: First of all, resist the temptation to run with the forecast on any snow map you might encounter on social media.

Instead, consider the source. If the person who posted the image isn't a familiar face that you've seen on TV, or a trained meteorologist from the National Weather Service or a private forecast company such as weather.com, a quick Google search of their name will almost always tell you if this is a reputable source.

Chris Dolce has been a senior meteorologist with weather.com for over 10 years after beginning his career with The Weather Channel in the early 2000s.