Challenges Forecasting the Snow Level

The snow level is a critical variable for mountain weather forecasting, but it is both difficult to define and predict.  

In fact, there isn't even an agreed upon definition of snow level amongst meteorologists or the public.  One might say it is the level above which precipitation is snow and below which precipitation is rain, but its just not that simple.  Precipitation doesn't just change from snow to rain at a given altitude.  Snow takes time to melt as it falls.  Different types of snow crystals melt faster than others.  Sometimes it is snowing but not accumulating on the ground due to melt.  What do you do when there is a mixture of snow and rain or drizzle?  Where do you put the snow level? 

To begin, perhaps it is easier to begin with a description of the transition zone, the layer in which snow is warming, melting, and turning to rain as it falls.  

The schematic below provides a simple conceptual model of the transition zone.  The freezing level marks the top of the transition zone and is the highest level at which the temperature is 0°C.  Temperatures above this level are below 0°C.  

The transition zone (h/t Ron Stewart)

The freezing level and the snow level are not the same because snow does not melt instantly when it falls below the freezing level.  Instead, it begins a transition as it falls, gradually changing from dry snow to wet snow to slush (part ice part water) and finally rain.  The distance over which this occurs depends on many factors including the profiles of temperature and relative humidity below the freezing level, type and size of the snow crystals, and fall speed of the snow crystals. 

For example, if the relative humidity is low, a given snow crystal can penetrate farther below the freezing level since the crystal will warm more slowly because it is also cooled by sublimation.  A dense graupel particle can generally penetrate farther below the freezing level because they have a high fall speed and take longer to warm and melt fully compared to a small dendrite.  

In addition, because the melting of snow cools the atmosphere to the melting point (0C), there is often an isothermal 0C layer in the transition zone, as depicted above.  This layer can be quite deep, especially if precipitation rates are high.  Because of this effect, the snow level may lower some when the precipitation rate increases and rise some when the precipitation rate decreases. 

In the above schematic, I have denoted the level at which the precipitation is half snow and half slush as the snow level, but that may not conform to what you think is the snow level.  Perhaps it should be lower and at the level where precipitation is all rain or all rain or slush.  Alternatively, perhaps it should be higher and at the level where precipitation is all snow?  

Or, perhaps instead of focusing on the type of precipitation falling, we focus instead on the level at which snow is accumulating on the ground.  That's sometimes called the snow line.  But if we do that, what exactly is the ground?  Grassy surfaces?  Asphalt surfaces?  

There are many ambiguities.

As discussed in Van Cleave et al. 2011, about 15 years ago the National Weather Service Western Region actually established a definition of snow level for their applications:

"The snow level is the elevation above which snow will fall, and below which rain will fall. A mix of rain and snow may be observed at elevations within a few hundred feet of the snow level. Snow will not accumulate on the ground below the snow level and may not even accumulate at elevations above the snow level."

Even that definition is somewhat ambiguous, but they also defined techniques for calculating and forecasting the snow level and identified the wet bulb 0.5°C level as the best proxy for snow level in the western region. That said, there can be variations in precipitation type and accumulations on the ground around this level.  

One situation that can be particularly challenging involves stable layers in which wet-bulb temperatures are very near 0°C through a deep layer.  An example is provided by last night's GFS forecast for 1000 UTC 01 January 2026 (3 AM MST New Year's Day).  At that time, a precipitating cloud layer has spread over northern Utah in southwesterly flow aloft, but stable conditions remain at low levels as the cold air currently in place over the Salt Lake Valley is slow to mix out.  The freezing level in this case is located just above 750 mb, or near about 2500 m MSL (~8200 ft MSL).  Because the atmosphere is saturated at that level, the freezing level in this case also represents the wet-bulb zero level. However, below that level, the layer is very close to 0°C down to almost 800 mb.  This might enable snow to penetrate farther below the freezing level than if temperatures were increasing with decreasing altitude below the freezing level.  

Layers in which temperatures are relatively constant with height are called isothermal, which means having a constant temperature. Having a deep isothermal layer near 0°C is one way for the snow level to penetrate to low elevations.  One needs to keep a close eye on these layers as they can make a big difference for snow level forecasts.    

The Birth of "Snow Augmentation" at Sun Valley

While researching the 1976/77 drought year, I came across the article below in the Salt Lake Tribune describing the benefits of "Snow Augmentation" at Sun Valley (click to enlarge).  

Published in late December 1976 when Utah resorts were not operating due to a lack of snow, it describes initial use of artificial snowmaking at Sun Valley.  Today, Sun Valley probably has one of the best snowmaking systems in the world.  At that time, during the great snow drought of 1976/77, it enabled three runs to be open for the holidays.  As described in the article,

"Ribbons of artificial snow from two to three feet deep have been shot by air compressor-activated snow guns up to two-thirds of the way up the mountain."  

It goes on describe the artificial snow as "cube like crystals" with "a bit more density than nature's snow."  I'm sure that sounds familiar to today's skiers.  The system sounded pretty cutting edge for the time, pulling 1400 gallons a minute and putting down two acre-feet of snow in eight hours.  

I suspect in 1976 there were no Utah ski resorts with snowmaking.  It would be interesting to look into the history of snowmaking investment and expansion at our resorts.  Online articles suggest Deer Valley had a snowmaking system when it opened in 1981.  Prior to the 2002 Olympics, Snowbasin expanded and installed what was probably at the time the state's most expansive and sophisticated snowmaking system.  My recollection is that Alta was not yet making snow when I moved here in 1995, but perhaps my memory is inaccurate.  AI tells a lot of good stories, but they can be authoritative BS so I'm reluctant to use it here.  

That said, as we have learned this season, most contemporary snowmaking systems today are still at the mercy of Mother Nature.   If wet-bulb temperatures are too high, snowmaking efforts are hopeless.  This has been a major issue in Utah so far this season (although guns will be roaring in the during and in the wake of this latest cold surge).  All-weather snowmaking systems are available, but are expensive to buy and operate and not used to cover large areas of terrain.  

Let's hope we see a colder, snowier pattern emerge for 2026.  

Is This The Worst Christmas Skiing Ever in Utah?

I never like to bad mouth Utah skiing but I think it is accurate to say that this is a truly abysmal start to the ski season right up there with many of the worst seasonal starts on record.  

But is it the worst?  

Currently on this Christmas Day there is little to no natural snow at low-to-mid elevations.  For example, the base of Park City is as bad as it gets right now from a natural snow perspective, although there is a white ribbon of death thanks to artificial snowmaking.   

Source: https://www.parkcitymountain.com/the-mountain/mountain-conditions/mountain-cams.aspx. Screenshot from 1:20 PM MST 25 Dec 2025.

Where we have data, there are now several SNOTELs in the Wasatch Range that are at their record low snowpack water equivalent for the date.  These SNOTELs are Ben Lomond Trail, Farmington Lower, Farmington, Hardscrabble, Parley's Summit, and Payson Ranger Station.  At Ben Lomond Peak (records back to the 1978/79 winter), the 4.5" of SWE is just ahead of the record low of 3.7".  At Snowbird (records back to the 1989/90 water year) the 6.0" of SWE is just ahead of the record low of 4.6".

So, at some sites we have the worst natural snowpack during the SNOTEL period of record for Christmas.  Others are close. 

But "during the SNOTEL period of record" is an important caveat.  The oldest SNOTEL stations in the Wasatch Range began operations during the 1978/79 winter.  Others have even shorter records.  They fail to provide a comparison with what is the worst Christmas and ski season start on record:

1976/77

How bad was the start of the 1976/77 ski season?  During November and December of 1976, Alta-Guard observed only 31" of snow and 2.04" of water equivalent precipitation.  How low is that?  Really low.  For comparison, Alta ski area has recorded 44.5" of snow and 6.75" of water so far this November and December (they will get more through this weekend) and the Atwater SNOTEL which is located where those Alta Guard snow and precipitation measurements would have been taken in 1976 has 6.9" of water equivalent in the snowpack.  Thus, precipitation in 1976/77 and far less than we have had this November and December.  

There was also no snowmaking in 1976/77.  Snowmaking has been limited this year due to warmth (more on this below), but it still has saved our bacon and allowed resorts to operate with limited terrain.  In 1976/77 the ski areas didn't have it and weren't even open for skiing during the Christmas Holidays.  Back in 1976 there were these things called "newspapers."  They printed the latest news on paper and were delivered to your door.  It was a predecessor to modern social media.  On December 24, 1976, the Salt Lake Tribune contained an article entitled "Resorts Featuring Snowless Events."  Park City was promoting golf and tennis, gondola rides to view the new Jupiter Bowl Ski Runs, and dryland training classes.  Snowbird had "summer rates," tram rides, and music.  Perhaps it was a predecessor to Octoberfest.

What made 1976 so bad?  The short answer is a big ridge.  Average 500-mb heights for November 1 to December 23, 1976 show a high amplitude ridge along the Pacific coast that acted to reduce storminess, yielding few storms.   Basically, it was a very dry pattern.  

In contrast, the mean pattern for the same period this year is far less amplified and although we are below average for precipitation in November and December, we have had some storms.  

The complicating factor this year, however, has been the warmth.  The high fraction of precipitation falling as rain instead of snow and frequent snow-loss events have prevented the buildup of a low-to-mid elevation snowpack.  

How warm has it been?  To use a scientific term, it has been bat-sh*t-crazy warm.  Below is a time series of mean Nov 1 to Dec 24 temperatures in Salt Lake City since 1875.  Do you see that data point on the far right?  The one that sticks up way above everything else?  That's this year.  

The mean temperature for 1 Nov to 24 Dec at the airport was 47F.  This is 4.1F warmer than the next warmest (42.9F in 1995).  That's like running a 3 minute mile.  Simply unbelievable.  

Unfortunately, we don't have complete records for temperatures at mountain sites.  There's data at Alta back to the 1940s, but there are a lot of missing days.  For what it's worth, the average temperature at Alta for the same November 1 to December 24 period was 35F this year (with complete data coverage).  That's the highest on record, but in other years, there's missing data, so this comparison is limited.  For what it's worth, the average temperature for the same period in 1976, when records were also complete, was 30.1F.

So to summarize, the worst Christmas skiing was in 1976/77.  There was practically no natural snow, even at upper elevations, due to persistent ridging and a lack of storms.  Precipitation was far less than we have seen so far this season and there was no snowmaking infrastructure to save the day.  

But this season is also different animal.  We have seen unusually high temperatures that are really unprecedented and exceptional in the Salt Lake Valley.  More analysis of the mountain observations is needed, but even at upper elevations this appears to have been a remarkably warm period.  This has led to a high fraction of precipitation falling as rain instead of snow at low-to-mid elevations and frequent snow ablation events.  We have even seen a substantial rain event at elevations up to about 9600 feet.  As a result, there's either no or very little snowpack below 8000 feet and a thin snowpack at higher elevations.  Although we now have snowmaking infrastructure, it has frequently sat dormant due to high temperatures.   

Skiing during the 1976/77 holiday period was bad due to a dry snow drought, but it's bad this season due to a dry-warm compound snow drought.  In a dry-warm compound snow drought, warmth exacerbates the impacts of below average precipitation, making a bad situation worse.  I suspect there is no historical analog to the start of this season in northern Utah.  We simply haven't ever seen temperatures in November and December this warm (and this sustained).  A good research question concerns whether or not this reflects an ongoing transition into a future in which dry-warm compound snow droughts become more common, especially during the early season.  This season is not a new normal, but it might reflect a new extreme that became more likely due to warming of the climate system.  

That said, it will be cold this weekend.  You may need to recalibrate.  

Enjoy Your Lump of Coal

Source: Cowboy State Daily

Merry Christmas to all from the Wasatch Weather Weenies.  

Apparently most of you have been naughty not nice because Santa delivered quite the lump of coal for Christmas.  Warmth and high snow levels materialized as anticipated last night.  It was a real soaker even at 9600 feet where Alta-Collins picked up only an inch of snow from 1.01" of water indicating a good deal of rainfall during the period.  

My memory is hazy this morning but I can't recall a mid-winter rain event of that magnitude at that elevation this time of year.  I do recall storms that started with a bit of rain or produced some with mist and rime due to a lack of ice nucleating particles in clouds, but a full inch of water with little snow at 9600 feet in December or January I suspect is pretty unusual.  That said, I'm working off memory here and leave it to you to dig into the data to see what you find.  

Meanwhile, it's 8:00 AM and 55°F at the Salt Lake City International Airport.  Skip the sleigh ride today and take the mountain bike out instead.  

A Rainy Night for Santa

A warm storm is on tap for the Night Before Christmas, one that will not make Santa or his reindeer happy.   

To illustrate this, let's take a look at the forecast from the High Resolution Rapid Refresh (HRRR) for Little Cottonwood Canyon (click to enlarge).  The HRRR calls for periods of precipitation overnight (green box below) adding up to 1.5" of snow water equivalent at Alta-Collins through 7 AM Christmas morning.  

That's the good news.  The bad news is that the overnight period is remarkably warm, with the wet-bulb-zero level, the altitude where the wet-bulb temperature reaches 0ºC and roughly the top of the melting layer, reaching as high as 10,700 feet (the snow line is usually several hundred feet below the wet bulb-zero level).  Our algorithms suggest this period, highlighted with green shading above, may produce rain even at the 9600 foot altitude of Alta-Collins.  

Snow during other periods at Alta-Collins adds up to 8.4" by 7 AM, although we don't account for losses due to rain, so that might be optimistic.  That 8.4" will probably be wet and dense.  Even by 7 am tomorrow, the HRRR has the wet-bulb-zero level at 9300 feet, which suggests that the precipitation will probably be all rain below 8000 feet overnight.  

The HRRR is a relatively wet model.  Others, such as the RRFS ensemble below, are calling for less precipitation.  Through 7 AM tomorrow (14z 25 Feb), the RRFS members produce 0.45 to 1.2" of water.  Some of those members also produce rain to near the Alta-Collins level during the warmest part of the storm.  Snowfall is in the 3-8" range by 7 am, with amounts varying depending on water equivalent, snow level, and snow-to-liquid ratio.  

So, Santa will likely be delivering presents in the rain below 8000 feet.  He might also encounter rain or a rain-snow mix up to 9500 feet or even a bit higher depending on his delivery schedule.  Christmas skiers probably won't find The Greatest Snow on Earth under the tree tomorrow morning, but perhaps 4-8" of wet, dense snow at 9500 feet. Expect a strong gradient in snowfall amount with altitude between about 8500 and 10000 feet and the most snow above 10000 feet.  

This is one forecast that I hope is totally wrong or out to lunch.  There is a wide range of forecast water equivalent being produced by the models, so maybe we can get lucky and we can have the storm come in colder than anticipated and produce a lot of water, with more substantial accumulations over 8000 ft.  It is Christmas after all.  

There will be a cooling trend over the next few days with some snowshowers on the 25th and 26th and a snowstorm Friday night and Saturday.  

Welcome to Mordor

Wind and warmth continue to dominate the weather here in Mordor, with the fires of Mt. Doom casting their glow over the Wasatch this morning.  

The average temperature since November 1 at the Salt Lake City International Airport of 46.3°F is a remarkable 2.4°F higher than any prior November 1 – December 21 period (note: the graph below includes data from the downtown weather bureau office prior to 1928).  

For the first 3 weeks of December (i.e., December 1–21) the mean temperature (43.3°F) is also the highest on record, although the gap to #2 (42.6°F in 1889) is not as large.  

Frodo and Samwise Gamgee continue to their quest to destroy the One Ring and break Sauron's evil grip on the weather of Utah, but remain well short of their goal.  

Hopes for the cover of precipitation increase around 0000 UTC 25 December (5 PM MST Christmas Eve) as southerly flow and an atmospheric river penetrate inland into Utah in advance of the very persistent upper-level trough that has been hanging off the west coast the past few days.  

The 82 members of the Utah Snow Ensemble have various ideas on the timing and intensity of the various bits and pieces of the storm, including the structure of the trough has it moves through Utah, but for the 3-day period from 0000 UTC 25 December to 0000 UTC 28 December (red-box below), 77 of the 82 members are producing 1–2.5" of water equivalent at Alta-Collins and 74 of the 82 members are producing 10–22" of snowfall.  

Most members are also calling for a cooling trend during that 3-day period, although it should be noted that at the start (i.e., 0000 UTC 25 December), median wet-bulb 0.5°C levels, which roughly denote the top of the melting layer, are at 10,000 feet.  This is a storm cycle that is going to start out very warm, perhaps with rain at the base of Park City on the evening of Christmas Eve.  By Christmas morning perhaps snow levels will be down to the base, but that's a touch-and-go forecast right now.  Monitor forecasts for this 3-day period if you are interested in details.  

Whether or not this modest storm cycle will enable Frodo and Samwise to reach Mt. Doom, destroy the ring, and loosen Sauron's grip on Utah weather remains unclear.