Denver, Colorado

Lake-effect Snow Explained

Posted by @docwx

This past week, we have been hit another meteorology buzz term of “Lake-effect Snow” (known in the meteorological community as LES).  Buffalo, NY was inundated by heavy snow amount with some places receiving > 80” in a 5-day period. It snowed so much that the NFL had to postpone and relocate the Buffalo Bills football game this past Sunday to Monday night in Detroit. Even the governor of New York, Andrew Cuomo, got into the LES talk by blaming the National Weather Service (NWS) about their terrible weather prediction, about which he was totally incorrect. The NWS did a fantastic job on letting people know of the severity of this LES event ahead of time.  Since, here in Colorado, we do not get to experience LES events, let’s talk briefly about what they are and what atmospheric processes cause them.

LES (lake effect snow) radar imageryA radar animation of the recent LES event over Buffalo (image courtesy of College of DuPage)

What is Lake-effect snow?
A snowstorm occurring near or downwind from the shore of a lake or large body of water resulting from the warming (destabilization) and moistening of cold air during passage over a relatively warm body of water. Lake-effect snow occurs when mean lake temperatures exceed mean land temperatures; therefore, the lake is not usually frozen for this phenomena to occur.

Lake effect snow average snowfallAverage Annual Snowfall Around the Great Lakes From 1971 – 2000.

Overview of the Lake-Effect Process
Lake-effect snow occurs to the lee (east) of the Great Lakes during the beginning of the cool season.  When the cold air mass travels across a relatively warmer lake, the air mass acquires heat and moisture, which destabilizes the air mass.  Cloud formation is further enhanced by thermal and frictional convergence and upslope along the lee shore.

LESConceptual Model of Lake-Effect Processes (Heat and moisture from lake + frictional convergence + upslope flow  = clouds and lake-effect precipitation) (image courtesy of COMET).

Lake-Effect Snow Characteristics
Lake-effect snow is considered a mesoscale convective snow phenomena.  Destabilization of the cold air mass is sometimes strong enough to produce thundersnow. Although convective in nature, the clouds in LES are very shallow compared to their Cumulonimbus counterparts (see Figure below), but can still pack quite a convective punch. These LES bands can come in a single- or multiband variety (See Figures below).

LESA vertical comparison between a Lake-Effect convective cloud and a summertime thunderstorm (image courtesy of COMET).

LES satellite imagerySatellite image of a Lake-Effect Single Snow Band event (image courtesy of COMET).

Satellite image of Lake-Effect Multiple Snow Bands event (image courtesy of COMET).Satellite image of Lake-Effect Multiple Snow Bands event (image courtesy of COMET).


Lake-effect snow occurs from late fall through winter, though lake-effect rain can occur from late summer through mid-fall.  Tremendous snowfall amounts and snowfall gradients are associated with LES events (see Figure below).  Lake-effect snowstorms rarely produce multiple fatalities directly, but are very disruptive to commerce and transportation, with the exception of the recent Buffalo event (November 2014), where at least 12 confirmed fatalities have been reported.

Image courtesy of Brad Panovich of WCNC via his facebook page.Image courtesy of Brad Panovich of WCNC via his facebook page.

While Buffalo is infamous for LES events, other parts of the Great Lakes region and North America and the world do experience this type of weather phenomena. The image below shows where these events can and do occur across North America.

Lake-Effect-type Phenomena across North America (image courtesy of COMET).Lake-Effect-type Phenomena across North America (image courtesy of COMET).

Typically, the rule of thumb is that once the Lakes are frozen over, the majority of the LES events will diminish or end the lake-effect season.  This is not always the case.  A frozen lake does not necessarily preclude a lake-effect event if ice cover is not continuous.  Sensible heat flux can occur through the ice but not as efficiently; therefore, if enough moisture is available, LES bands can still occur.

Key Processes from a Forecasting Point of View

  • Localized instability - lapse rate and boundary layer depth

  • Fetch – the distance in which the cold air moves across the warmer lake water (the longer the fetch, the higher the snowfall potential)  

  • Wind direction and shear

  • Cloud microphysics – snow crystal habits

  • Synoptic (large) scale forcing

  • Orography/topography -- lake shape and orientation and lee-shore topography

  • Upstream moisture

  • Snow/ice cover on the lake

Forecasting Challenges
Lake-effect snowstorms are difficult to observe and forecast due to several factors.  First, LES bands are shallow systems (depth often < 3 km), and the lowest elevation radar scans “overshoot” the tops.  Secondly, the onset, intensity, orientation, and exact location are very sensitive to wind shear/direction and inversions in the lower troposphere.  Third, LES bands are difficult to distinguish from orographic influences in some locations (e.g., Great Salt Lake). Lastly, some operational NWP models do not have sufficient resolution, microphysics, etc. to resolve the scales of lake-effect snow bands quite yet.  The HRRRX, the ESRL version of the HRRR, performed well compared to its NCEP counterpart keeping the heaviest snow just of south of Buffalo.

11-h forecast of simulated composite reflectivity from HRRRX (ESRL version) initiated on 1900 UTC at 11/17/2014.11-h forecast of simulated composite reflectivity from HRRRX (ESRL version) initiated on 1900 UTC at 11/17/2014.

Final Thoughts
Lake-effect snow is weather phenomena that can be difficult to forecast due to the sensitivity of many atmospheric variables, especially upstream fetch and wind direction. These events can last as long as the cold air moves across the “relatively” warm lake without much deviation from the overall wind field.  In most cases, LES episodes do not create major societal distribution and multiple fatalities.   The November 17-21, 2014 event near Buffalo, NY was an historic event and somewhat atypical as far as LES events go.  Snowfall rates for this storm reached 5” per hour over many hours revealing the highly convective nature of the snow band.  In the end, the NWS did a fantastic job in forecasting and anticipating societal impacts, but mitigation for an outlier event is nearly impossible.

New York lake-effect snowfall totals map
Total snowfall from Nov. 17-21, 2014 lake-effect snow event in western New York (Image courtesy of the Weather Channel).

The State of the Atmosphere: Sunday, November 23, 2014

Posted by @brendansweather

We’re heading into the final full week of November and Thanksgiving, 2014. As we discussed in our previous post, the weather pattern for eastern Colorado has been overwhelmingly ‘normal’ since early in the week, with ongoing mountain snows.

Snowfall across central and western Colorado will continue through Tuesday, with winter weather advisories and
winter storm warnings up across the board. Snowfall totals will range from 10 to 20”+ through Monday evening. There will be a chance for a few snow showers across lower elevations through Tuesday as well, but we’re not expecting much impact from these showers. The latest hi-res NAM has a dusting of snow for Denver over this period, with a spotty inch or two across the Palmer Divide. Mean SREF plumes produce a ½” of snow at DIA, but likely most folks will see less or none. Greatest snow shower activity will likely be south and southeast of Denver Sunday. Temperatures will also be cooler over the next several days, with daytime highs near 40.

Here’s a look at the
WPC’s snowfall probability map for 8” or more of snow through Wednesday -- probabilities of 80% for the north central mountains.  

WPC snowfall probabilitiesWPC 3 Day Snowfall Probabilities Colorado

Thanksgiving Week
The forecast for Thanksgiving across eastern Colorado has gone from what looked “interesting” regarding snow, to likely nothing. Late last week the EURO was indeed showing a pretty hefty cold front and chance for snow by Thursday for Denver, even then its own ensembles were a bit more finicky. Also, the Canadian (which had a few snowy runs) backed off quickly, and well, the GFS never really showed anything. Chalk one up for the GFS in the win column? Maybe.

It seems to me that we were
way too far out for anyone to be getting too excited about anything in the first place. And the ECM tends to have a western bias at times with its troughing -- while the GFS can have an eastern bias as we saw much of early fall. It was worth watching if only because our preferred medium range model was showing snow -- but why anyone feels the need to share 10 day snowfall maps in any serious setting, is beyond me. Fact is, this time of year some model at any given run almost always shows a snowstorm 10 days out. Always look to see what the ensembles are trying to do, any if two of three models have no snow, maybe consider there’s something there. Get ready for a long winter of long-range maybe snows…

Most models and their ensembles now generally agree on ridging taking hold by midweek, which will lead to a more mild, dry Thanksgiving day. What we’ll likely need to watch Thursday is for a backdoor cold front that could make its way far enough west to knock temperatures down even for the Denver area, and maybe quite a bit across the eastern Plains of Colorado.  Interestingly, it seems like a lot of the same folks calling for a Turkey-snow are now calling for warm, when we really need to watch the potential for cooler temps closely. The chance for snow in Denver, at this time, looks pretty low.

The GEM’s 5 day temperature anomaly from Nov 25 through Nov 30 actually looks pretty good to me. The coldest anomalies will be east of Colorado, but eastern Colorado will range from near normal to several degrees below normal through the period.

Canadian model temperature anomaly forecastWeatherBell Analytics

Beyond Thanksgiving things become a bit more murky. Last week the ECM and Canadian models both showed a very active pattern through the first week of December, but latest runs are less convincing. Plenty of time to see how things will shake out here…

In short -- cooler temperatures Sunday and Monday with a few snow showers drifting off the mountains across the Denver area. Greatest chance for any accumulation east of the mountains will likely be along the Palmer Ridge, and even then any accumulation will be very light. Still a few questions especially with regard to temperatures for Thanksgiving day, but snow chances generally look low for the I-25 urban corridor on Turkey Day. We’ll keep a close eye on things and offer updates as needed.

All for now!

Normal November Pattern Settles in, Mountain Snow Through the Weekend

Posted by @MattMakens247wx

We’ve snapped back to a normal weather pattern this week in Denver after an extended period of bitter cold. Afternoon temperatures in the 40s and low 50s is right on par for the time of year, and this trend will continue right into the weekend.

The afternoon “warmth” is due in part to downsloping northwesterly winds that have also created these periods of standing waveform clouds. The
mountain waves are pronounced during times of a strong northwesterly fetch and humidity at the mid-levels.

To the visual forecaster, these cloud formations indicate more than dry/warm-ish Denver days. These also indicate that enough moisture is within the flow to be aware that mountain snow is also possible. Indeed that is the case. Here is the GFS, as a singular example, of the snowfall expected through 5pm Sunday. This assumes a 10:1 ratio, which is underestimate in this type of flow regime. 20:1 would produce double these amounts and would seem more appropriate.
I am working on a post discussing our major modeling shortfalls when it comes to snow prediction; you’ll see it in the coming weeks.

GFS snowfall through weekend
WeatherBell Analytics

The northern and central mountains with a north to westerly orientation fare well in these weather patterns. Resorts like Beaver Creek and Vail send lots of press releases on these days, too. Further, Steamboat Springs is known for “champagne powder” during the same periods of time that Denver is warm and under mountain wave clouds.

I chose to stop at 5pm Sunday night for a reason. That reason is that that is when this pattern will take a brief break. A trough, a weak one, will scoot over the state later Sunday through Monday and will bring a slight cooldown to Denver and a chance for our own snow. This is still the GFS, but through 5pm Wednesday night. As of this posting, we aren’t expecting a significant snowfall anywhere outside of the mountains. It’ll be another minor affair for the metro area.

GFS 7dayWeatherBell Analytics

As far as the temperature change. Denver will likely go from near 50 Saturday to the 30s for Sunday through Tuesday...also a relatively minor affair.

Bottom line, Thanksgiving skiers and travelers will come away with a pretty good week across Colorado. HOWEVER, I’m guessing many of you will be headed elsewhere for Thanksgiving...east perhaps?

Currently, Buffalo NY is getting walloped
with over 5 feet of snow. The Great Lakes region will certainly be a snow covered Thanksgiving from what’s fallen there the past few days.

NY Buffalo Lake-effect SnowCarolyn Thompson/AP Photo

Additional snowfall through the next 7 days doesn’t look nearly as major, which is better news for flights to the major hubs back east. Through Thanksgiving Day we will likely see snowfall in these areas:

GFS usa snowfall forecast WeatherBell Analytics 

That’s a lot of snow to fall in Northern Minnesota, but for the most part the upcoming snowfall is quite manageable. HOWEVER, if you are to fly to the northeast Sunday, Monday, or Tuesday there could be great delays as that’s when most of the snow will fall and the wind will be awful around the Great Lakes with that system.

As the snow intensity drops, temperatures will be remain quite cold. If you are setting out that wardrobe for Grandma’s house, pack extra layers for those areas to the east. The quick skiff of snow and “cold” that Denver will feel early in the week will become jarring cold for the eastern half of the country by Thanksgiving.

This is showing above or below average temperatures for Thanksgiving morning. Yes this is showing nearly 20-degrees colder than average for the southeast near Atlanta, and 10-degrees colder than average for New York City. Ugh, we will see Al Roker bundled up again at the Macy’s Thanksgiving Day Parade. Meanwhile, the southwestern quarter of the country will have mild weather, including Denver. For those needing to warm-up, Phoenix, LA, or San Diego will do just that for you.

GFS temperaturesWeatherBell Analytics


Denver’s Thanksgiving Day weather is still a bit more in question. Several models have been trying to develop our next system by next Thursday and Friday. The EURO/Canadian have been most consistent with this, but the the GFS is seeing it too (at times) to an extent. A long ways out, but something we’re watching, and if it turns out to be true -- those anomalous cold temperatures pictured above will likely be a bit further west as well.


I hope this gives you an early heads up on the coming week. We, of course, will get far more in depth the closer we get to the holiday week. I, for one, am tuning the skis right now and will be hitting Colorado’s fresh powder. I like what I see coming to our mountains this week.

Comments on Last Week

Posted by @brendansweather

85 degrees.

That’s the temperature swing we saw between last Sunday’s high of 71
° and Thursday morning’s low of -14° at Denver International Airport (DIA).

Last week’s cold wallop was nothing to scoff at, especially so early in the season. We broke two all time record lows on November 12 and 13, of -13
° and -14° F respectively, and two record low maximums. The record low for November 13 wasn’t just broken, it was smashed. The previous record low was -3° F set in 1916, making the new record 11 degrees colder.

The cold air arrived in a hurry. Last Monday morning we recorded a high of 69
° in downtown Denver at 10:04am. By 10:20am we had dropped into the 40s, and by afternoon we had fallen into the teens with light snow.

Here’s a look at the giant temperature swing last week.

Denver temperature over last weekWeather5280


Denver also saw it’s first measurable snowfall of the year at DIA, recording 0.1” on Tuesday November 11, and 2.2” on Wednesday the 12th. DIA recorded an additional 1.7” of snow on Saturday, bringing the totals for the week to 4” even. Not a huge snow event for Denver as I think we made clear all along.

Worth noting is that a few locations that did
much better with snowfall, however. The Boulder area won big from the Tuesday/Wednesday snowfall with upwards of 8” being reported. The Palmer Divide did best with Saturday’s snowfall as a persistent snow band setup Saturday evening delivering from 2 - 5” of snow -- it was one of the locations we were watching for the potential to bust ‘high’ under banding.

Competitive edge
I was told last Saturday from one of our readers that we were running much colder for the coming week than any other forecast they had seen. Of course, this made me look. Leaving names out of it, this was one of the forecasts I saw on Saturday November 1 (left) for last week’s cold shot. The coldest maximum over the four days was 30 degrees, and forecast for Wednesday. This turned out to be a 24 degree temperature bust. The forecast low of 16 was still 10 degrees warmer than the actual high of 6 for Wednesday.

Denver media forecast bustDenver outlet forecast bust

By Saturday we had already dropped forecast highs into the low teens for Wednesday, and
released this chart of blended forecast model daily maximums for the coming week -- nowhere near freezing for Tuesday, Wednesday, and Thursday.

There were other reports of forecasts being off by as much as 40 degrees within 24 hours of the forecast day last last week, which is really inexcusable.

On the right is a forecast from Thursday calling for 3 - 6” of snow in Denver Saturday night into Sunday. We went out with 0.5 to 3” for Denver and stuck with it. Officially DIA recorded 1.7” of snow, with 1.25” recorded at our station downtown Denver. The wildcard was banded snowfall, but this wasn’t a 3 - 6” snowstorm for Denver, as we saw.

The point here is not about pointing fingers -- because, hell, we all blow forecasts.

The point is, at Weather5280, we’re sort of obsessed with getting it right. And it’s not just about being competitive in our forecasts, it’s about making sure our readers are always well-informed.