Some Thoughts on Using Long-Range Forecasts

Snow is currently short of scant in the Wasatch with 6" being reported by the Collins automated sensor and spotty coverage on high-north aspects.

Alta Ski Area cam image from 8:40 AM MDT 31 Oct 2025. Source: Alta Ski Area.

While there's no skiing, it does look like a splendid Halloween and weekend for late fall, so enjoy.  

Given that we are playing the waiting game for snow, I thought I would share a few thoughts on using long-range forecast guidance.  

The National Weather Service Global Forecast System (GFS) produces forecasts at ~13 km grid spacing out to 16 days.  The European Center for Medium Range Weather Forecasting similarly produces a "control" forecast at ~9 km grid spacing out to 15 days with their Integrated Forecast System.  For simplicity I will call this the ECMWF control, although you may know it as the Euro, EC, ECMWF, or HRES.  Plots from these modeling systems are readily available on weather.utah.edu and other sites.  However, when you are looking many days out, there is little point in relying on forecasts from the GFS or ECMWF "control" when a large ensemble is available.  Here's why.

In the case of models run by the ECMWF, the ECMWF control is simply the so-called "unperturbed control member" of their 51 member ECMWF medium-range Ensemble Prediction System or ENS.  This means that the initial conditions for the ECMWF control represent the "best guess" as to what the atmosphere looks like when the model run is started.  The other 50 members are run with the exact same model, but with perturbed initial conditions.  Differences in the forecasts arise from these slightly different initial conditions, as well as the perturbations that are added to the model physics to account for uncertainties in how we simulate physical processes such the generation of clouds and precipitation.  The latter is referred to as "stochastic" (or random) physics.  We don't know for example exactly what the characteristics of cloud droplets are in a cloud, so we use stochastic physics to try and account for the range of possibilities.  Together, the differing initial conditions and stochastic physics lead to differing forecasts, with the spread growing with time.  

It turns out that the ECMWF control slightly outperforms the other members of the ENS at short lead times.  However, at longer lead times, the advantage becomes small and in a statistical sense it has very little advantage over the other ensemble members.  So, when you are looking 5+ days into the future, there's little point in relying on the ECMWF control.  Each member of the ENS is equally likely, just as when you roll a die, any of the six faces is equally likely to end up on top.  The ENS die just has a lot more faces.  

The National Weather Service Global Ensemble Modeling System or GEFS has 31 members.  The GFS is NOT a GEFS member.  The GFS and GEFS use the same modeling system, but the GFS is run at higher resolution (~13 km) than the GEFS members (~25 km), including the GEFS control run. Despite this resolution advantage, the gap between the GFS and the individual GEFS members closes with lead time. 

The Utah Snow Ensemble attempts to squeeze as much as we can from the 51 ENS members and 31 GEFS members through the use of downscaling and other techniques.  We are trying to take advantage of all of those runs to provide information as to the likelihood of precipitation and snowfall amounts at high resolution.  Rather than relying on the GFS or ECMWF control, I think of all members of the ensemble as equally likely.  Additionally, over many events, the GEFS or ENS mean is going to outperform the individual forecasts from the GFS or ECMWF control.  There's strength in numbers.  

Let's use last night's forecast as an example.  The GFS is bringing in a pretty healthy system for next Thursday.  Below is the forecast valid 1500 UTC (0800 MDT) Thursday 6 November.  

If we pull out the forecast for Alta-Collins, the GFS puts out about 1.3" of water through 1200 UTC (0500 MDT) Friday 7 November.  Bank on it?  
  

Well if we look at the Utah Snow Ensemble, a different picture emerges.  The GFS is actually an outlier compared to all of the forecasts produced by the GEFS and ENS (note that the Utah Snow Ensemble goes to 10 days instead of 7 like the GFS forecast above).  Through 1200 UTC 7 November, only 4 out of 31 GEFS forecasts produce > 1" of water equivalent.  The wettest ENS member is around 0.6".  

The GFS could verify, but it's a lower probability outcome than inferred from the full GEFS and ENS ensembles.  

We provide tables for Alta-Collins precipitation and snowfall at https://weather.utah.edu/text/ensgefsdslccforecast.html.  It takes a bit of time to get used to these, but I'll focus solely on the total water equivalent precipitation table below (right click and open in another window to enlarge).  I've highlighted in red the total accumulated precipitation through 6 AM MDT Friday 7 November (we have a bug for dealing with daylight time so that's going to be 5 AM local time).  The rows include the minimum from the ensemble on the top and maximum on the bottom.  P10 indicates the so called tenth percentile.  10% of the forecasts are at or below this value.  P50 is the fiftieth percentile or median.  Half of the forecasts are below this value and half are above.  In this case P50 is 0.17", so half the ensemble members are at or below 0.17".  Much lower than the GFS.

The GFS's 1.3" sits somewhere between the P75 value and the P90 value.  Those are the seventyfifth and ninetieth percentiles.  Let's guesstimate it to be at about P85, which means 85% of the ensemble members are below the GFS amount and 15% are above.  

So, the GFS could verify, but a look at all of the forecasts suggests that's a lower probability outcome.  The full Utah Ensemble suggests the "over under" through Friday morning is not the GFS's 1.3" but more like 0.16" (the P50 value).

The discussion above assumes the Utah Snow Ensemble is unbiased.   In other words, on average, it produces about the same amount of precipitation at Colins as is observed and that it's probabilities, evaluated over many events, reflect real world probabilities (i.e, when the Utah Snow Ensemble says there's a 10% chance of 1" or more of precipitation it actually happens 10% of the time).  I don't actually know if this is the case.  The GEFS and ENS underpredict precipitation at most mountain sites.  Our downscaling helps with this, but undoubtably biases remain, vary by site, and in some cases could be high rather than low biases.  I haven't had time to look into this, so I'm going to leave this for your, the student, to investigate.  

Warm and Dry to Start November

Warm, dry, ridgy weather looks to predominate the end of October and start of November.  All members of the Utah Snow Ensemble have no precipitation at Alta-Collins through 0600 UTC (0000 MDT) Tuesday 4 November.  Then a few get excited, but others a meh, so we will have to see what the storm track decides. 

The waiting game is now in full swing.

Update on Weekend Changes

On Monday we discussed the possibility of big changes this weekend with the potential for a significant snowfall event (see Big Changes Next Weekend).  Changes are still coming, but right now it appears that the pattern will not be quite as favorable for mountain snow, although there will still be some.  

The latest GFS is still bringing a front through northern Utah this weekend.  Below is the forecast valid 0000 UTC 26 Oct (6 PM MDT Saturday) with the front pushing into northwest Utah.  It's still looking like we will have a frontal passage Saturday night with Sunday significantly cooler.  

That said, the moisture and precipitation with the front isn't quite as robust as advertised a few days ago.  There's still a decent chance of valley showers and mountain snow, but amounts are not as high. 

Then, instead of the trough digging through the western US, it lifts off to the northeast.  We get a bit more precipitation Sunday and Sunday night as the atmospheric river and jet stream behind the front move through, but we never get into the deep, unstable, cold air.  Sigh.  

The net result is that the water and snow numbers for Collins are on the low end, with all but two members under a 10" total for the storm cycle and half the members under 5.5 inches.  

We'll see how things shake out, but this is looking like a good time to keep expectations low and hope for the best.

Big Changes Next Weekend

Although this week will be mainly quiet with the exception of a weak monsoon trough that will move across southern Utah Wednesday night and Thursday, next weekend looks exciting.

The latest GFS is forecasting a strong cold front to be moving across northern Utah late Saturday.  Behind the front, a strong jet extends across the Pacific and into the western United States.  

The European is on board with this general pattern shift, but is faster with the front, which it forecasts to be in eastern Utah by late Saturday.  

Regardless of timing, the frontal passage should bring another round of precipitation with it to add to the 4.95" October record at the Salt Lake City International Airport.  Additionally, it also looks to bring mountain snow and even the possibility of valley flakes, although much with regards to that will depend on what happens in the wake of the front.  

The Utah Snow Ensemble forecast for Alta-Collins has a few ENS (European Ensemble) members producing a bit with of snow with the monsoon trough Thursday, although most keep us dry.  However, during and following the frontal passage, many members are keen on a significant snowfall event.  Most are generating something in the 8 to 22 inch range.  

I think it's a bit early to be worried about numbers as this is a pattern in which the details are hard to forecast with precision at such lead times.  It is, however, worthy paying attention to.  We'll see how things evolve over the next couple of days.  

A Dry Weekend

It's a rarity here in the Great Salt Lake rainforest, but it looks like we have a dry spell settling in that will last through the weekend.  

The trough that brought us rain over the past few days is exiting the area today.  On Saturday, a low-amplitude upper-level ridge looks to be moving in, with no monsoon moisture or Pacific storms to be found in our area.  

There is a brush-by system that flirts with northern Utah late Sunday but right now the precipitation from that looks not push into Utah through 0000 UTC 20 Oct (1800 MDT Sunday).

That system could bring a bit of precipitation to northern Utah, mainly in the mountains, on Sunday night. For Alta Collins, for example, a bit over half of the Utah Snow Ensemble members produce no precipitation Sunday night, but a fee give us some light amounts Sunday Night, with a couple doing a bit better than that. 

That's the main threat for precipitation over the next few days if you are hoping to add to aour already record-breaking October totals.  The next major system that could flirt with us comes in around October 26th. 

The Importance of Research for Teaching

Higher education is currently under siege and facing many challenges including declining public confidence and trust, funding cuts, and political interference.  Amongst the many questions being raised concerns whether university professors need to teach more, with greater prioritization of teaching over research.  When addressing this question, the public, politicians, and even professors and provosts often assume that research and teaching are mutually exclusive activities.  This stovepiped view of research and teaching has some conveniences, but the reality is that the best professors create a strong synergy between research and teaching to the benefit of students.  

During my 30 years at the University of Utah, I have served as the lead or co-principal investigator on more than 30 external research grants, but have also won college and university teaching awards.  I am classified by my department as a "research intensive" faculty member, but I also care deeply about teaching.  I know many other faculty who are similarly committed to both research and teaching, and recognize that there are many benefits of research for teaching (and vice versa).  

In many ways research is teaching because a good deal of external research funding supports graduate and undergraduate students.  Well over half of my grant support (maybe more than 75% in some cases) goes to graduate and undergraduate students and, as a result, much of the time I spend on research is actually spent mentoring graduate and undergraduate students.  This includes guiding them through their research and helping them become scientists who can innovate, create new knowledge and techniques, work in a team, and be effective if not outstanding communicators.  

In other words, teaching.  

Research also enables the latest advances to be brought into the classroom.  Readers of this blog are well aware of the products on weather.utah.edu.  Although we get no direct funding for the web site, we have developed it to provide access to the latest forecast techniques and enable their use in the classroom.  For example, students can use the Utah Snow Ensemble, which uses downscaling and machine learning techniques to produce a large snowfall ensemble.  This product is not only being used in upper-division atmospheric sciences classes, but even introductory ones such as Atmos 1000: Secrets of the Greatest Snow on Earth.

Research funding also provides experiences for students who may not be a formal member of my research group, including those in physical labs or in the field.  I nearly always build in a research experience opportunity for students into my National Science Foundation grants.  For example, we have brought the Doppler on Wheels (DOW) radar to Utah on two occasions to teach students about radar meteorology and mountain precipitation.  

Fun times in the Doppler on Wheels

For this winter we are planning a large field campaign focused on the northern Wasatch.  We currently have undergraduate and graduate students helping to configure and deploy observing systems for that campaign.  During field operations, students will launch weather balloons and operate DOW radars.  All of this because of research.  

Finally, teaching also benefits research.  Having to teach classes requires a much deeper understanding of fundamental subject matter, which makes you a better scientist, but it also exposes gaps in understanding.  Classroom discussions have helped me to identify or refine research questions and ideas on many occasions.  

This synergy between research and teaching, and the benefits of research for teaching, seems to be missing in the public dialog about the future of higher education.  In fact, I would say it is even inadequately considered on college campuses.  Most policies for evaluating faculty productivity or workload, for example, very clearly distinguish between research and teaching, which are often treated in near isolation for faculty reviews.  

It is for the reasons above that I am opposed to sledgehammer proposals to simply increase faculty teaching assignments.  In some disciplines or for some faculty members this might make sense, but in other situations, it could harm the mutually beneficial balance of research and teaching that benefits students.  It would certainly harm the quality of instruction in scientifically intensive disciplines that are evolving rapidly.  

How the October 2025 Deluge Compares to the Past

More thunder and rain last night.  Observations from the Salt Lake City International Airport indicate another 0.68" in the gauge, all if it falling yesterday evening and overnight.  If my math is right, that brings the monthly total up to 4.16", topping the previous record for the month of 3.91".  Really, it wasn't even a fair fight.  It's only October 14!

With the October record easily dispatched, I thought we might aim higher.  First let's have a look at the records for each calendar month:

January 1918: 3.89" 
February 1998: 4.89" 
March 1891: 4.66" 
April 1944: 4.90"
May 1908: 5.76"
June 1998: 3.84"
July 1982: 2.57"
August 1968: 3.66"
September 1982: 7.04" (!!!)
October 2025: 4.16" and counting
November 1875: 5.81" 
December 1983 and 1889: 4.37"

And then additional months at or above 4":
March 1876: 4.00"
April 1974: 4.57"
April 1986: 4.55"
April 1984: 4.43"
April 1886: 4.43"
April 2011: 4.06"
May 2011: 5.14"
May 1977: 4.76"
May 1876: 4.30"
May 1901: 4.27"
May 2015: 4.19"
May 1898: 4.19"
September 1973: 4.07"

So, calendar months that have observed 4" of precipitation or more are February, March, April, and May in the late winter and spring and September, October, November, and December in the fall and early winter.  There have been 22 calendar months with at least 4" of precipitation, most commonly in April (6 times) and May (7 times), which aligns well with the months that on average are wettest at the airport.  The November 1875 record (5.81") is an extreme outlier for that month.  I took a quick look at the daily records though and there was nothing that was obviously bad, so I'm including it here.

Of the months with ≥ 4", 10 are before 1928 when the observing site was in downtown Salt Lake City.  Downtown is a bit wetter on average than the airport, so it's possible this contributed, but there have also been changes in observing techniques and instruments that can also affect the frequency of extreme events. 

There are also five months in the 1980s, a period that was exceptionally wet in northern Utah.  The all time monthly record of 7.04" in September 1982 is really impressive and certainly helped moistened the soils prior to the epic 1982/83 winter that ultimately led to the famous flooding during the (late) spring runoff.  

Source: Marriott Digital Library

With those numbers in mind, a reasonable target to root for this month would be the coveted 5" barrier, which has only happened four times before.  Getting to 7" seems crazy, but not impossible.  That said, I'm going to set the over/under at 6.25" based on a quick look at the extended forecasts.  There's a lot of uncertainty in those forecasts though, so place your bets and see if you can beat the house.

There are a few other ways we might look at this exceptional run, which started on October 3 and currently goes through October 13, a stretch of 11 days.  You are probably already awaer that 4 October was the 2nd wettest day in Salt Lake (2.47").  For an 11 day stretch, this Oct 3 to 13 rates as the 30th wettest, although this count includes two exceptionally wet stretches, one in Fall 1982 and the other in Spring 1901.

Like our current stretch, each of those included one exceptionally wet day, 26 Sep 1982 (2.27") and 3 May 1901 (2.64", the wettest day on record).  This illustrates that one really wet event can set the table for a record breaking run if the pattern remains favorable for precipitation.  

Given that 1982 shows up conspicuously and the 1982/83 winter was an epic, should we get excited about this winter.  Of course you should get excited about this winter!  You are going to be skiing and a bad year in Utah is better than a good year most anywhere else (Japan excluded).  

Ah, but you probably want to know if this is a harbinger of a big winter, like fall 1982 was. The two are simply not comparable.  The 1982/83 winter was an enormous El Nino year.  We're currently looking at perhaps marginal La Nina conditions.  I look at what is happening now as not an indication of what the winter will bring.  It doesn't load the dice one way or another.  Your guess is still as good as mine.  

A Wet Start to October

Total October precipitation at the Salt Lake City International Airport through October 11 is 3.48".  This is already good for the 7th wettest October on record (tied with 2004).  

Ten wettest Octobers on record at the Salt Lake City International Airport. Source: https://xmacis.rcc-acis.org/.

The record (3.91") is only 0.43" away.  Average precipitation the rest of the month would eclipse that, so we have a pretty good shot of having the wettest October on record.

Rocktober Is Coming

Wow, what a forecast.  Buckle up as a lot is going to be happening through Sunday.  Things get exciting around here when the mid-latitude storm track phases with monsoon moisture and we will have both in spades this weekend.  

I'll start with the GFS forecast valid 1200 UTC 11 October (0600 MDT Saturday).  A deep trough sits along the Pacific coast and monsoon moisture is streaming into the four corners area, producing the Predecessor Rain Event (PRE) discussed the the previous post in advance of the remnants of Hurricane Priscilla over northern Arizona, southeast Utah, and southwest Colorado.  In addition, a strong cold front sits over eastern Nevada.

By 0000 UTC 12 October (1800 MDT Saturday) that strong cold front is over northern Utah and lighting up parts of northwest Utah.  

I'm not sure how Saturday is going to play out in the Salt Lake Valley ahead of the front.  There will be southwest flow ahead of the front but it's hard to say if we might get showers and thunderstorms from the monsoon moisture or if we might be "dry slotted" in an area of dry air that has wrapped round the trough.  We shall see.

By 0300 UTC 12 October (2100 MDT Saturday), the front is pushing across the Salt Lake Valley bringing rain and possibly thunderstorms.  This is a strong front and a deep trough, expect a big change overnight.  

By 1200 UTC 12 October (0600 MDT Sunday) we are post-frontal, with unstable northwesterly flow.  700-mb (crest-level temps) at this time are -5.6°C, compared to 6.8°C just 12-hours earlier on Saturday afternoon.  Snow levels will be a dropping with the front.  

Expect some adjustments in timing and details as the event approaches and consult forecasts, not only for the Salt Lake area, but especially if you are adventuring in southern Utah over the next few days.  The NWS currently has a flood watch up for that area starting this afternoon.  

Source: NWS; Snapshot from 1:06 PM MDT 9 Oct 2025

More mountain snow in the central Wasatch?  You betcha'.  Snow levels will be high Saturday, but they will be crashing down Saturday night if the forecast frontal timing is accurate.  Accumulations will depend strongly on how much precipitation falls during and after the frontal passage.  Through 0000 UTC 13 October (1800 MDT Sunday), the median downscaled GEFS and European ensembles are generating about 1.5" of water, but the former is around only 4" of snow whereas the latter is closer to 7".  Much is going to depend on details that are tough to anticipate now.  Stay tuned.

I'll add that median wet-bulb zero levels drop to as low as 6600 feet on Sunday morning.  It's not out of the realm of possibility to see some flakes down to bench levels.  

Buckle up.

Potential Southwest Gullywasher

A lot of exciting things can happen this time of year when the mid-latitude storm track interacts with monsoon moisture.  The latest model runs really get my attention in this regard as they are putting out a good deal of precipitation in the southwest US.

The latest GFS shows the precipitation picking up in northern Arizona on Thursday.  By 1200 UTC 10 OCT (0600 MDT Friday), precipitation is organized in a band that runs from SW to NE from northern Arizona and to western Colorado. 

This precipitation region is to the north and northeast of Hurricane Priscilla, which is forecast to be off the coast of Baja California at that time.  Such a precipitation feature is sometimes referred to as a Predecessor Rain Event, or PRE and occurs as enhanced vapor transport north of a hurricane experiences large-scale lift.  Forecast soundings from Flagstaff at that time are saturated with precipitable water values of 25.4 mm (exactly 1 inch if you prefer US units), which is quite high for early October.  

The precise track of Priscilla is still unclear, but the latest GFS has her decaying off the coast of Baja at 1200 UTC 11 Oct (0600 MDT Saturday) with a plume of integrated vapor transport exceeding 500 kg/m/s extending into northern Arizona.  

That's impressively high IVT given the elevation of that region.  The GFS also continues to produce heavy precipitation in the same banded region.  

Total accumulated precipitation produced by the GFS exceeds 2 inches in a band that runs from just west of Flagstaff through SE UT and into SW CO.  Precipitation maxima produced by the GFS along this band exceed 3 inches, and this is a model that cannot resolve fine-scale precipitation features like thunderstorms.

Source: Tropical Tidbits

The ECMWF has a somewhat different idea about Priscilla's track and brings her through southeast Arizona.  That smears the precipitation out more across the four-corners area and gives heavy precipitaiton to southeast Arizona and southwest New Mexico.

Source: Tropical Tidbits

This has all the hallmarks of a late-monsoon, high-impact event for the southwest with heavy precipitation and flooding.  

Little Cottonwood Forecast Guidance

Last week we described a new HRRR-derived forecast product for Snowbasin.  So my friends in the central Wasatch don't feel neglected, we've also dropped a similar product for Little Cottonwood Canyon on weather.utah.edu.  It's available from a link along the top bar or directly at https://home.chpc.utah.edu/~steenburgh/ml/hrrrlccforecast.html.  It should update regularly, but we're not watching things all the time, so always check the dates before using.  

Like the new Snowbasin forecast guidance, the Little Cottonwood forecast guidance is based on the HRRR and uses machine learning to improve the prediction of several variables.  There's a graphical product (the examples below are the forecast from 0000 UTC 4 October and not todays):

and a tabular summary:

We provide wind forecasts for three locations: Mt. Baldy (AMB; 11,066 ft), Top of the Collins chair (ALT; 10,443 ft), and Cardiff Peak (IFF; 10,059 ft).  AMB is a wind-exposed location above Alta Ski Area and IFF the only upper-elevation wind-observing site on the north side of the canyon.  Each of these is important for avalanche forecasting and mitigation in the upper canyon.  We provide temperature forecasts for AMB, ALT, the Collins Snow Study Plot (CLN; 9662 ft), Base of Alta (ATH20; 8752 ft), and Elberts in the mid canyon (ELBUT; 7600 ft). All of the wind and temperture forecasts are based on machine learning algorithms applied to the HRRR.  

There is also the height of the 0°C wet-bulb temperature level, which is a rough estimate of the top of the melting layer (snow level is typically several hundred feet below this).  The precipitation fields, including snow-to-liquid ratio/water content, hourly and totally precipitation, and snowfall amounts, are for CLN.  Water equivalent precipitation is directly from the HRRR. Snow-to-liquid ratio/water content are based on a machine learning algorithm developed by Michael Pletcher, a graduate student in my group.  We apply these to the water equivalent precipitation to obtain the snowfall forecasts.  

The graphics are organized so that the precipitation related plots are plotted on the left and the temperature, RH, and wind plots are on the right.  From top to bottom, the precipitation-related plots include:

1: Hourly water equivalent (bars) and total water equivalent (black line) at CLN.
2: 700-mb temperature (purple line) and relative humidity (green line) over CLN.  These are useful for estimating cloud conditions at crest level.
3: Wet-bulb zero level (black line) and 1000 ft below the wet-bulb zero level (green line) with color-fill between.  This is based on a high-resolution profile over Salt Lake City and provides an approximation for the melting layer.  We use the upstream profile so we can have data down to the Salt Lake Valley and because the National Weather Service provides a very high resolution profile from the HRRR over the Salt Lake City airport.  Color fill is used to indicate the elevation of key locations in Little Cottonwood Canyon.
4. Hourly snow-to-liquid ratio (bars) at CLN with light blue used for non-precipitating period and dark blue for precipitating periods. Inclusion of non-precipitating periods allows one to have an estimated snow-to-liquid ratio during periods when the HRRR doesn't produce precip (but Mother Nature might). 
5. Hourly snow (bars) and total snowfall (black line) at CLN.

From top-to-bottom, the temperature, RH, and wind plots are:

1. Hourly temperature at AMB, ALT, CLN, ATH20, and ELBUT following the embedded legend.
2. Hourly RH at AMB, ALT, ATH20, and ELBUT  following the embedded legend (CLN has not had a RH sensor to enable training for this variable).
3. Hourly wind speed (red line), wind gusts (blue line), and wind direction (black dots) at AMB. Wind speed and gusts based on the left-hand y-axis.  Wind direction the right-hand y-axis.
4. Same as 3 except for ALT.
5. Same as 3 except for IFF.  Winds at this site will be weaker and more erratic, and that is evident in this plot.

The primary strengths of this product are its intentional design for Little Cottonwood Canyon and the use of machine learning to improve the HRRR forecasts, which are too low-resolution to adequately capture local effects.  The weaknesses are that its based on one model (the HRRR), rather than an ensemble, and we are not currently using any machine learning to improve the water equivalent forecasts (that could be done but I only have so much time in the day).  One, however, can also consult our experimental plumes from the RRFS Ensemble (https://weather.utah.edu/index.php?runcode=2025100606&t=rrfsqsf&d=PL&r=CLN) and Utah Snow Ensemble (https://weather.utah.edu/index.php?runcode=2025100600&t=ensgefsds&d=PL&r=CLN).  In doing this, one should recognize that the plumes are for the nearest grid point to CLN and that the RRFS and Utah Snow Ensemble have different terrain representations than the HRRR.  

Comments appreciated.  Criticism ignored.  Actually, that's not true, but like all of our online products, we do what we can with limited time.  Yes, we know the web page is not mobile-device friendly.

An Epic Storm

Pretty remarkable water numbers from this latest storm.  

At 7 PM MDT, the NWS reported a record calendar-day rainfall of 2.39".

The 2.39" appears to be the 2nd largest calendar-day rainfall on record, topped only by the 2.64" on 3 May 1901.  

Source: https://xmacis.rcc-acis.org/

In other news, the Collins stake sits at around 8 inches.

Source: Alta Ski Area

Snowbasin Forecast Guidance

For a long time, I've neglected our friends in the northern Wasatch, but that is changing this year as we will be conducting a major field campaign focused on that region this winter.  More on that in a future post.  In preparation for that campaign, we just dropped a new Snowbasin forecast guidance product on weather.utah.edu.  It's available from a link along the top bar or directly at https://home.chpc.utah.edu/~steenburgh/ml/hrrrsnowbasinforecast.html.  It should update regularly, but we're not watching things all the time, so always check the dates before using.  

This new Snowbasin forecast guidance is based on the HRRR and uses machine learning to improve the prediction of several variables.  There's a graphical product (the example below is from last season):

and a tabular summary:

We provide wind forecasts for three locations: Ogden Peak (OGP), Strawberry Gondola Top (SB2), and the Boardwalk snow-study plot (SBBWK).  The first two are on the Snowbasin Ridge line and very wind exposed.  The latter is in a wind sheltered mid-mountain area.  We provide temperature forecasts for the aforementioned locations plus the Middle Bowl observing site (SNI), which is near the bottom of the Middle Bowl chair, and the main base area (SBE).  All of these forecasts are based on machine learning algorithms applied to the HRRR.  

There is also the height of the 0°C wet-bulb temperature level, which is a rough estimate of the top of the melting layer (snow level is typically several hundred feet below this).  This is based on a profile from the HRRR just upstream of Snowbasin and referred to as the wet-bulb-zero level or WBZ.  The precipitation fields, including snow-to-liquid ratio/water content, hourly and totally precipitation, and snowfall amounts, are for the Boardwalk site.  Water equivalent precipitation is directly from the HRRR. Snow-to-liquid ratio/water content are based on a new machine learning algorithm developed by Michael Pletcher, a graduate student in my group.  We apply these to the water equivalent precipitation to obtain the snowfall forecasts.  

The graphics are organized so that the precipitation related plots are plotted on the left and the temperature, RH, and wind plots are on the right.  From top to bottom, the precipitation-related plots include:

1: Hourly water equivalent (bars) and total water equivalent (black line) at Boardwalk.
2: 700-mb temperature (purple line) and relative humidity (green line) over Boardwalk.  These are useful for estimating cloud conditions at crest level, which is typically just below 700-mb at Snowbasin.
3: Wet-bulb zero level (black line) and 1000 ft below the wet-bulb zero level (green line) with color-fil between.  This is based on a high-resolution profile just upstream of Snowbasin and provides an approximation for the melting layer.  We use the profile upstream of Snowbasin so we can have data down to roughly the elevation of Ogden and because the National Weather Service provides a very high resolution profile from the HRRR over the Ogden airport.  Color fill is used to indicate the elevation of key locations at Snowbasin.
4. Hourly snow-to-liquid ratio (bars) at Boardwalk with light blue used for non-precipitating period and dark blue for precipitating periods. Inclusion of non-precipitating periods allows one to have an estimated snow-to-liquid ratio during periods when the HRRR doesn't produce precip (but Mother Nature might). 
5. Hourly snow (bars) and total snowfall (black line) at Boardwalk.

From top-to-bottom, the temperature, RH, and wind plots are:

1. Hourly temperature at OGP, SB2, SBBWK, SNI, and SBE following the embedded legend.
2. Hourly RH at OGP, SB2, SBBWK, SNI, and SBE following the embedded legend.
3. Hourly wind speed (red line), wind gusts (blue line), and wind direction (black dots) at OGP. Wind speed and gusts based on the left-hand y-axis.  Wind direction the right-hand y-axis.
4. Same as 3 except for SB2.
5. Same as 3 except for SBBWK.  Winds at this site will be weaker and more erratic, and that is evident in this plot.

The primary strengths of this product are its intentional design for Snowbasin and the use of machine learning to improve the HRRR forecasts, which are too low-resolution to adequately capture local effects.  The weaknesses are that its based on one model (the HRRR), rather than an ensemble, and we are not currently using any machine learning to improve the water equivalent forecasts (that could be done but I only have so much time in the day).  One, however, can also consult our experimental plumes from the RRFS Ensemble (https://weather.utah.edu/index.php?runcode=2025100306&t=rrfsqsf&d=PL&r=SNI) and Utah Snow Ensemble (https://weather.utah.edu/index.php?runcode=2025100300&t=ensgefsds&d=PL&r=SNI).  In doing this, one should recognize that the plumes are for the nearest grid point to Middle Bowl, not Boardwalk as in the HRRR-derived product above, and that the RRFS, and Utah Snow Ensemble have different terrain representations.  As a result, the product above is for about 8000 feet, whereas the RRFS plumes are for 6864 feet and the Utah Snow Ensemble for 8282 feet.  This can be important to consider during events with marginal snow levels.

Comments appreciated.  Criticism ignored.  Actually, that's not true, but like all of our online products, we do what we can with limited time.  Yes, we know the web page is not mobile-device friendly.

First Widespread Snow in the Central Wasatch?

Some high-altitude locations such as the top of Snowbird have already seen their first snow, but on Saturday we may see a bit more wide spread snowfall.

The latest GFS shows a bonafide cold trough moving across northern Utah at 1200 UTC 4 October (0600 MDT Saturday) with an accompanying cold front and precipitation system.  

Our GFS-derived machine-learened forecast product for upper Little Cottonwood shows temperatures on Mt. Baldy (11,000 ft) and at Collins (9,700 ft) plummeting Friday night and Saturday along with wet-bulb zero levels dropping to near 8000 feet by noon Saturday.  If that verifies, that would probably put the snow level Saturday afternoon around 7500 feet.  

For Collins' 9700 foot elevation, the GFS is producing a total of about 0.8" of water.  The storm begins with rain at that elevation Friday afternoon, but eventually changes to snow Friday night.  Total snowfall of just over 4".

The Utah Snow Ensemble is a bit more jacked. 75% of the 82 members produce 5.6" of snow by 12Z 5 Oct (6 AM Sunday).

Our snowfall estimates do not consider melt on the ground and the vertical resolution of the Utah snow ensemble is such that it's snow-level forecasts are somewhat uncertain.  That said, I expect to see the white stuff Saturday at elevations above about 8000 feet (possibly lower) with the first wide-spread accumulations at upper elevations.  My best guess right now is 3-6" for Alta-Collins.