Snow impacts billions across the world and inspires countless people into the field of meteorology. Here's a look behind-the-scenes of a snowfall forecast.
Category
🗞
NewsTranscript
00:00Welcome to AccuWeather's Ask the Experts. I'm your host, Jeff Cornish. We go beyond
00:16the forecast to give you the how and why on all the cool and interesting things that you've
00:20wondered about and wanted to ask in weather, space and science. And today we are talking
00:25about the science of snow prediction with Dan DePauly, AccuWeather's Senior Director
00:30of Forecast Operations. And Dan, this is a big part of our operation here in our business
00:35at AccuWeather, snowfall forecasting. And Dan, this is such an important thing to talk
00:39about as, well, it's impactful to so many people.
00:43It sure is good to be with you, Jeff. And I think snow is, for me, it's one of the most
00:46interesting things to forecast. It's really what inspired me to be a meteorologist back
00:52when I was a kid growing up in New Jersey. We'll talk a bit about that later on, I think.
00:56And snow is really impactful to people. It shuts down schools. It snarls supply chains
01:01with packages getting delivered later. And it can really be dangerous, too, obviously,
01:06from a traffic perspective. So it really impacts a lot of people throughout the winter season
01:09across the United States.
01:10And for the most part, with snow, the report card is very visible. People know how much
01:16snow fell in their backyard most of the time. Sometimes it's melting. But compared to rain,
01:20snow is something that people know when we're accurate or when the forecast doesn't work
01:26out as well.
01:27So we want to talk about what goes into the AccuWeather forecast with snowfall. How complicated
01:32is that process?
01:33Yeah, so it's certainly one of the more challenging things that we forecast as a team. And at
01:38AccuWeather, we've developed a philosophy over many decades about how we approach all
01:42types of forecasting, but especially snow. One of the things that's important to us is
01:46being consistent with our forecast and not flip-flopping around with different forecasts
01:50as a storm approaches. So trying to set ourselves up with a forecast initially that is sort
01:55of what we expect is the best forecast three, four days in advance that limits the potential
01:59for changes as we approach that storm from an accumulation perspective, as well as the
02:03timing of the snow. That's obviously very important. And how fast the snow falls is
02:07very important for road crews in terms of how they can keep up with the snow during
02:11the storm.
02:12We've seen a lot of change in weather data and modeling and so forth compared to 40 or
02:1750 years ago when there were maybe two options. So how many different models do we have at
02:22our disposal here at AccuWeather that we might consult when putting together a snowfall forecast?
02:26So we're looking at a variety of different models, both publicly available ones that
02:29the different governments across the world run, like the National Weather Service in
02:32the United States. We also have access to proprietary data sets. And we're looking at
02:36different cutting-edge tools like some of the new artificial intelligence models that
02:39are coming out sort of very rapidly from many different organizations. So we have about
02:44170 or now even more maybe different models that we look at and we combine them in different
02:48ways to see how the data looks. And we have different tools to look at those and our meteorologists
02:53use those to help refine the forecast even further.
02:56And we're really proud of our team's success and the AccuWeather forecast is created differently
03:02than a lot of others out there. When I worked in local TV, you know, I admit I was not aware
03:07of how much went into the AccuWeather forecast, especially when it comes to the human aspect
03:13of things. Because there are computer model driven apps out there where there's no human
03:18that ever really touches that forecast. That's very different at AccuWeather.
03:21Yeah, we're really proud of the team that we have of our meteorologists. We have close
03:25to 100 expert meteorologists across AccuWeather and many of them work on our snow forecasts
03:30and ice forecasts during the winter months. And it's really a consensus approach. That's
03:34an important part of the AccuWeather philosophy. That's not just my forecast or your forecast,
03:37Jeff. It's the entire team providing their input together and people's different experiences
03:42and knowledge bases to help come up with a forecast that's better than any one individual
03:45can do by themselves. And it's that human input into the forecast on top of the really
03:51cutting edge data that really sets AccuWeather's forecast apart that we provide not just a
03:56forecast of how much snow or when it will fall, but also the impacts of that snow or
04:02that ice, when it will happen, will it happen during the commute, will it fall overnight,
04:06will it melt on the road because it's been pretty warm ahead of the storm. So we try
04:09to really convey the impacts of the storm as well as the context of it. And that's the
04:12important piece that our meteorologists add on top of the data.
04:15And every storm is a learning experience. I know that you're a better forecaster now
04:19than you were five, six, seven years ago. I am more dialed in now just with weather
04:23than I was when I first graduated from college. And a lot of our forecasters have been doing
04:28this for decades. And they have a good knack for not just meteorology, but also the geography.
04:34So how important is geography and climatology in influencing that forecast, the subtle differences
04:40in wind direction for the Great Lakes and so forth?
04:42Yeah, I think both of those are key. To be a good meteorologist, understanding geography
04:47and what has happened in the past of climatology are both very important. You have to know,
04:51so a wind in a certain direction, does that cause cooling or warming? Is it coming off
04:55a warm body of water like is often the case early in the season in the Northeast where
04:58that Atlantic Ocean temperature is warmer than it is later in the season, that can cause
05:02more rain than snow sometimes in December in parts of the Northeast and Mid-Atlantic.
05:05And what has happened in the past is very important. So are you predicting something
05:09that has never occurred in recorded history? That's something to know. Or what's the frequency
05:13of this type of snow event in a certain area? And also knowing that a snowstorm of one inch
05:18in Atlanta is very different impact-wise than a one-inch snowstorm in Buffalo, New York.
05:22Absolutely. Is there one part of the country or maybe one type of snow regime, whether
05:26it might be an Arctic front with snow squalls or lake effect snow, that you especially enjoy
05:33forecasting? I think probably based on where I grew up in New Jersey, these deep coastal
05:38lows, nor'easters in the winter that bring a variety of hazards. Sometimes it can be
05:41all rain, but in the winter they tend to be a mix of rain and snow. You're trying to find
05:45where that rain-snow line will be and when the changeover will happen. And some places
05:49you can get one to two feet of snow in pretty high population zones. And I think those to
05:53me are the ones that are most interesting. I grew up in the northwestern suburbs of Philadelphia
05:58and it's not fun to watch that rain-snow line nudge across the city coming into the northwestern
06:03suburbs when you want to hold on to the snow, especially if you're a kid hoping to get out
06:08of school the next day. And forecasting snow is so different in other parts of the country.
06:12When I worked in the Midwest, it was a little bit unnerving to forecast snow when the storm
06:16didn't exist on radar yet. It would develop over the plains on the lee side of the Rockies.
06:21There's a lot that goes into this in different parts of the country. So is there a difference
06:25in predicting snowfall maybe in November or December as opposed to the middle of winter
06:30or into the early spring? Yes, so depending on the time of year, it's definitely important to
06:34look at a variety of factors. Early in the season, first of all, a lot of people may not
06:38have been used to snow yet because it hasn't happened in six, eight months. So just people
06:42aren't as prepared for winter weather and driving in that. So conveying the impacts there is very
06:47important. Also sometimes there's not as much treatment on the roads, salt and other types
06:49of treatment on the roads, and sometimes that means that the roads can be more slick early
06:53on in the season. Temperatures, especially later in the season, you have to watch for the sun angle.
06:57The higher the sun angle by late February and March, if snow falls during the day in March,
07:02for instance, it has a harder time accumulating on the roads because the sun angle is higher,
07:05so it helps to melt the snow as it falls if it's not falling at a very good rate. So there's a
07:10variety of different factors throughout the season that we look at in addition to just what the
07:14forecast of how much liquid is coming from that storm to determine how much snow will fall.
07:20And I know that when I worked in local TV, I worked at one station in northwest Pennsylvania
07:26where a lot of the time in that market, people would typically run these model-driven forecasts
07:33and they would predict snowfall down to 8.3 inches or something like that. It was a model-driven
07:39thing and that would kind of broad brush things, but it was almost communicating too much certainty,
07:44too much precision. We tend to forecast in ranges and ultimately that's what I prefer doing when I
07:49work there as well. So why do we predict in ranges as opposed to precise amounts of snow?
07:56So one reason is the fact that between a couple of inches, there's usually not a big difference
08:02in impact. Three to six inches, pretty similar impact. It's a plowable snow, it's going to cause
08:06disruption to roads. Six to twelve inches, you're talking a more significant snowstorm. And one or
08:10two inches, pretty similar types of impacts and similar types of response are needed in those
08:15types of events. Also, there's a lot of just small changes in the amount of liquid that comes from a
08:20storm have larger impacts on how much snow falls. We'll talk a bit about snow ratio coming up and
08:26that's something that we can dive into more. So it's a bit more difficult than rain to
08:30actually predict the amount and that's one of the reasons that ranges are very helpful for people to
08:34see what the impact will be. And it's more honest too. We try to be honest. You know, we want to
08:39give you a confident forecast. If we just waffle with our uncertainty, that's not going to help
08:43anybody. But at the same time, there's a little more honesty in a forecast range and that AccuWeather
08:48WinterCast is a fantastic feature because that's a very transparent look at our confidence level in
08:53any of those ranges. Yeah, and with WinterCast is one of the really unique features of the AccuWeather
08:58app and on our website, you can see not just the forecast of how much snow we expect, for instance,
09:02three to six inches, but what's the probability of more snow than that six to ten or what's the
09:06probability of less snow than that one to three inches, for example, to give you a sense for what
09:10our confidence is in that forecast. Okay, well we have our first viewer question. This one comes to
09:14us from Paul in Pennsylvania and how is snowfall measured? I've measured snow in my backyard before
09:20and have nearly a foot of snow. Then I see on the news it was only officially about eight inches. So
09:25why are the amounts so different? So it really depends how you measure and there is a best
09:30practice to measuring that the National Weather Service has said and it's really the way you want
09:33to do it is have a snowboard. So a flat, it could just be a wooden board that you put out well before
09:38the snow starts to fall so it's at the actual air temperature and then you measure on that board with
09:43a ruler and you don't brush it off every hour either because that can inflate amounts too.
09:48But measuring on the grass can sometimes cause an inflated amount because if you push that ruler
09:51too far down it goes into the grass underneath the snow you can get more snow than has fallen.
09:57Also drifting is pretty common in open spaces so you can, if you are going to measure in an open
10:01field, take a number of measurements and average those measurements to get the best representation
10:05of what has fallen. Okay, all very good advice there and and even beyond that sometimes what
10:11happens at the airport is going to be different than six, seven, eight miles away and those of
10:15you in lake effect snow areas certainly know the variety there. So a lot of good stuff so far Dan,
10:19different ways to answer these questions and a lot to think about with snow. Again Dan, our Senior
10:24Director of Forecast Operations, he's got a big job here at IQ Weather and coming up we're going to be
10:28putting your knowledge to the test in our weather-wise segment, is this really a thing?
10:34Like can you really get sick if you go outside with wet hair? We'll look at that but next we're
10:39going to explain what snow ratios are and why they are so important when a storm is coming.
10:44We'll answer more of your viewer questions as well when Ask the Experts returns.
10:59Welcome back to AccuWeather's Ask the Experts, I'm your host Jeff Cornish and today we are talking
11:03about the science of snow prediction at AccuWeather's Senior Director of Forecast Operations
11:09and AccuWeather snow expert, he's here for a specific reason, Dan Dupodwin is back and Dan,
11:14we're talking all about snow and how AccuWeather forecasts snow and you know there's a lot to this.
11:19People sometimes may not realize how intentional snow is and how important it is to us.
11:26People sometimes may not realize how intentional we are in trying to bring the greatest snowfall
11:31forecast that we can and the winter cast product that you talked about, AccuWeather.com and the
11:36AccuWeather app, you can look for that snowflake icon whenever you're dealing with snow in your
11:40forecast. But Dan, one of the big things it does either inflate or deflate a storm's potential for
11:46snowfall is snow ratio. So what is a snow ratio and how do we calculate this and look at this?
11:53So snow ratio, if we think about it, when you're forecasting just rain, all these models that we
11:58look at as one of the tools we use to forecast, they all predict how much liquid equivalent will
12:02come from a certain weather situation and snow that we need to convert that liquid equivalent
12:06to snow and the way we do that is snow ratio. So basically what it is, is it's how many inches of
12:11snow do we expect to come from one inch of liquid and that can vary significantly for many different
12:16storms and across the storm, during a storm you can start with a certain ratio and that can change
12:21if you stay in the same location as the storm evolves and that's really important understanding
12:25how much snow will fall. So for instance, we generally sort of a ballpark range that's a good
12:30average is 10 to 1. So from one every one inch of liquid you can expect 10 inches of snow. Okay and
12:36a lot of variables here because we need to nail down how much precipitation is coming and then
12:41what the snowfall ratio may be and obviously that's critically important in ultimately
12:47landing in our snowfall ranges. What are some influences that may make that snowfall ratio 30 to
12:531 versus 7 or 8 or 6 to 1? So basically what we look at is the temperature throughout the different
13:00parts of the atmosphere and that's important not just for the ratio of snow to liquid but also the
13:04precipitation type. So these are going to be mixing with sleet or rain at different times during the
13:09storm. So we look at the temperature profile not just at the surface but how does that temperature
13:13profile look throughout the entire column of the atmosphere above a certain location and that
13:18helps us determine whether it's going to be a wet heavy snow which can be like a 6 or 8 to 1 ratio
13:22which is pretty typical in parts of the northeast. It can be typical in the early and later part of
13:27the season as well when it's a bit warmer in the atmosphere. During the heart of the winter season
13:32especially out in the plains into the Rockies you can get 20, 30, some places even 40 to 1 ratios
13:37where you're getting 40 inches of snow just from one inch of liquid which is pretty impressive.
13:41That's a dry powdery fluffy snow that blows around really well to second cause even more
13:45issues with visibility and blizzard conditions. So there's really a wide range we look at in terms
13:52of snow ratio and how that ends up impacting the amount of snow from a storm. And when you're trying
13:56to remove that snow from your driveway you experience the difference there. There are times
14:00and this is maybe a little more common in the northern plains where you could take a leaf blower
14:05and that could clear your whole driveway easily or easily other times you're shoveling that wet
14:10snow and it really really takes a toll on on your back when you're lifting some of that wet stuff.
14:15Well Dan we want to talk to more or at least hear from other viewers with viewer questions.
14:19So we want to hear from Tom now in New York City. So Tom what do you want to ask the expert?
14:26How difficult is it to predict how much snow is going to fall during a storm?
14:30And is it any more difficult than predicting other weather events like hurricanes, tornadoes,
14:35or just rainfall? So I wouldn't say it's necessarily difficult. We've sort of gone through
14:40some of the snow ratio and other influences to a storm. We always look for the precipitation type
14:45and whether there'll be a changeover whether it will start as rain and go to snow or start as snow
14:49and go to rain. Whether there'll be ice in the mix as well can cut down on accumulations. So it's
14:53actually more difficult than other types of weather forecasts but it's certainly different.
14:57We're looking at different parts of the atmosphere. We're looking at different forecast parameters
15:00during a storm and it's important to also convey the confidence and what that confidence is in a
15:05given storm. Often many times in advance of a storm five, six, seven days we can be pretty
15:09confident that there's going to be a winter storm that's going to bring snow to a region. We can
15:13sort of rule out the fact it'll be rain. Other times though it takes to two or three days ahead
15:17of time to be more confident of what's going to happen. So conveying what we're confident in is
15:21always an important part of what we do at AccuWeather. Telling you as a as a viewer as a
15:25consumer of our app what we know is an important part of our philosophy. And some regions are more
15:30difficult than others. The Rockies and out west even the west coast can be more difficult and
15:33some of that reason is because of the fact that we have less understanding of the real-time
15:38atmosphere over the Pacific Ocean because we have less observations. So some of the tools we use
15:44have a bit harder of a time predicting the snow in the west or also like Alaska is another
15:49part of the country that is a bit more difficult to forecast snow. And there's often a an increase
15:54in our confidence when the storm or at least the dip in the jet stream that's going to produce a
15:58storm for the east coast when that dip in the jet stream finally reaches the west coast it's
16:03sampled much better. So suddenly we're getting high resolution information to feed into the
16:08models and therefore the forecast quality is going to improve at that time. So our next question
16:12comes from Laura in Maryland and Laura writes is there a snow storm from your childhood that made
16:18you want to become a meteorologist? There sure is and I think snow is probably the thing that
16:23really cemented my interest and passion for meteorology and swayed me to become a meteorologist.
16:30The storm that really pushed me in that direction was the blizzard of 1996. I grew up in central New
16:34Jersey and we had 30 inches of snow. It's really the snow storm on record at least in probably
16:38modern times in the middle of New Jersey for snow and I remember sledding. I basically made a big
16:45hill in my backyard of snow off our deck. We had so much snow we were able to make our own sledding
16:48hill in our backyard from that event. So the blizzard of 96 is certainly at the top of my
16:53list but also a few more from my childhood. Right before I but don't really remember when I was
16:57younger the superstorm of 93 was sort of right when I was four or five years old and then also
17:04there were a couple of storms including the President's Day storm in the early 2000s in
17:07February that dumped about almost two feet of snow. So I was fortunate enough to grow up in an era of
17:12plenty of important and impactful snow storms in the northeast. All right and there will be more
17:17to come which is always very exciting. Well Dan DePotman, AccuWeather senior director of forecast
17:22operations and snow expert. Dan thanks again for your insight, your time and all you do here
17:27to drive the bus and lead our team to success. Don't forget when you have a question about
17:32weather space or science you can write us or send us a video question at asktheexperts
17:36at accuweather.com. You can also call us at 888-566-6606. And coming up next we're going
17:44to test your knowledge on some of those popular cold weather myths that you hear all the time in
17:50a segment we call is this really a thing? Ask the experts will be right back so stay with us here
17:55on the AccuWeather Network.
18:05It's time now for weather wise and today we look at some cold weather myths in our segment
18:09is this really a thing? So let's start with that saying your mother or grandmother may have told
18:14you if you go outside with wet hair you'll catch a cold. Is this really a thing? The answer is no.
18:20Colds are caused by viruses so you will not get a cold just by being cold. However colder air is a
18:27better environment conducive to viruses and there is some research that suggests the lack of sun
18:33and vitamin d during the winter may play a role in a weakened immune system. Next up if you fall
18:40into icy water you'll quickly die. Is this really the case? Is this really a thing? Hypothermia is
18:46prolonged exposure to the cold. Drowning is actually the bigger risk when you initially fall
18:51into cold water. People panic. Cold water shock leads them to swallow water and that can lead them
18:57to drown. Knowing you won't quickly die from hypothermia can help you remain calm for a few
19:03minutes until help arrives. And finally put on a hat because most of your body heat escapes through
19:09your head. Is this really a thing? Again not really true. The amount of body heat that you lose is
19:14related to the surface area exposed to the cold. So your head is only about 10 percent of the
19:19body's surface area. So you do lose some of your body heat when you don't cover your head in the
19:24cold but not really the majority of it. Thanks again for joining us here on AccuWeather's Ask
19:29the Experts. I'm your host Jeff Cornish. Don't forget whenever you have a question about weather,
19:33space, or science you can write us or send us a video question at asktheexperts at accuweather.com.
19:39You can also call us at 888-566-6606. Have a great one.