The Folly of Betting on the GFS and "DModel/Dt"

Just a quick post today following up on some themes from posts the past few days.

The graphic below lops through GFS 228, 204, 180, 156, 132, 108, and 84 hour forecasts valid at 0000 UTC 4 December (5 PM MST Sunday).  Imagine trying to forecast for Sunday afternoon based on just this single forecast system.  Good luck.

The loop also illustrates the folly of forecasting based on model trends, or what forecasters call "DModel/Dt" (i.e., the rate of change of the model forecast with respect to time).  Clearly, there is no evidence of a consistent trend in those forecasts.  The pattern is too chaotic, leading to a lack of run-to-run consistency, even in the more recent forecasts.

It's fun to talk about DModel/Dt, but studies have found that it has little forecast value for medium-range forecasts.  Hamill (2003) examined the value of DModel/Dt and here's what they found:
"Extrapolation of forecast trends was shown to have little forecast value. Also, there was only a small amount of information on forecast accuracy from the amount of discrepancy between short-term lagged forecasts. The lack of validity of this rule of thumb suggest that others should also be carefully scrutinized before use.
Let's put this rule of thumb to bed.

GOES-16 is no longer transmitting ABI data

16-panel image of all GOES-16 ABI Bands, 1332 UTC on 30 November 2017 (Click to enlarge)

In preparation for its move from 89.5º W Longitude to the operational GOES-East position at 75.2º W Longitude, GOES-16 Instruments — the ABI, the GLM, and others — have been placed in ‘safe mode.’  In that mode, the instruments do not scan or transmit data.  This occurred shortly after the 1330 UTC Full Disk image, and the 1332 CONUS Image, shown above.  GOES-16 instrumentation will start scanning and transmitting again, sometime between 14 and 20 December.  In contrast to earlier GOES Satellites, GOES-R series satellites will not transmit data when they are shifting longitude.

Other examples of the final preliminary, non-operational GOES-16 ABI images are shown below: (1) Visible (0.64 µm) imagery centered over snow-covered Mount Washington, New Hampshire, (2) Full Disk Water Vapor (6.9 µm) imagery and (3) a closer view of Water Vapor (7.3 µm, 6.9 µm and 6.2 µm) images showing mountain waves over Wyoming and Colorado.

GOES-16 Visible (0.64 µm) images, centered on Mount Washington, New Hampshire (Click to animate)

GOES-16 Visible (0.64 µm) images, centered on Mount Washington, New Hampshire [click to animate]

GOES-16 Water Vapor (6.9 µm) images (Click to animate)

GOES-16 Water Vapor (6.9 µm) images [click to animate]

GOES-16 Lower-level (7.3 µm), Mid-level (6.9 µm) and Upper-level (6.2 µm) Water Vapor images [click to play animation]

GOES-16 Lower-level (7.3 µm), Mid-level (6.9 µm) and Upper-level (6.2 µm) Water Vapor images [click to animate]

Super Dark Days: Advice from a Meteorologist

Yesterday was an amazingly dark day.  In fact, according to UW meteorologist Mark Albright, it was the darkest day in Seattle (at the UW) since Dec. 8, 2015.  For the technically oriented among you, we had .61 MegaJoules per meter square yesterday, while on December  8, 2015 there was .57.   Essentially the same.  Typically this time of the year we have 2-3 times that amount.

The Seattle Space Needle PanoCan at noon yesterday says it all (see below).
Dark and dismal.

Why so dark?  We start with our northern latitude and the fact that we are within a month of the winter solstice.  Thus, we are near rock bottom for radiation reaching the top of the atmosphere.

Then we are in the cloudiest/stormiest time of the year.  End of November is our worst.

The visible satellite image and the regional radar at noon yesterday tells the story.  Thick clouds and rain greatly reduce the feeble solar radiation reaching the surface.

How bad was it?  Here are the radiation and weather measurements from the roof of my department in Seattle for the 72 hr ending 5 AM today (Wed.)  The bottom panel is solar radiation.

Almost nothing....I couldn't believe it when I first looked at the plot.    The third panel gives temperature (black line).   THE SUN WAS SO FEEBLE THAT THERE WAS NO DAYTIME WARMING.  In fact, temperature dropped during the day.

OK, the sun situation is really bad now.  What do you do?  

Listen to a meteorologist!

My first advice has nothing to do with weather.  Get out during the middle of the day.  Take a walk, go for a run, ride a bike.  Just get outside during the brightest time of the day...that really helps me.

Second, go to whether there are less clouds...and nearly every day there are places to get a lot more light.     For example, head for rain shadow areas in the lee of mountains..

During the winter storm season when southwesterly flow is prevalent, head towards the Olympic Rain Shadow, northeast of the Olympic Mountains.  Northern Whidbey Island, Sequim, Port Townsend, Victoria or the southern San Juan Island will do.

Here is an see the clearing NE of the Olympics?

After fronts go through and the winds turn more westerly, head east of the Cascade crest over the eastern slopes.  Places like Cle Elum. (see satellite image of such a situation).

But during the winter (November to February) don't do too far down into the Columbia Basin.  Why?  Because it often fills up with low clouds.  Very bad.
As shown below.

Sometimes in mid-winter under high pressure, the lowlands are full of low clouds, sometimes from the Willamette Valley to Bellingham.   But the mountains are in the clear! To get out of it, just go up in up to the passes or climb one of our hills.

Now, 10-20 times a year, we get a Puget Sound convergence zone and there are clouds (and often rain) right over Seattle (see below).  Want to escape the murk?  No problemo.  Just head north or south by 10 miles and you will be in sun!

Finally, head up into the snow in our mountains.   Snow reflects solar radiation and makes it much brighter, WHEN there is no active weather going on.

Pretty bright at Snoqualmie Summit!

There are more "secrets" I can tell you, but the message is clear.  Don't be passive in accepting a dark fate. You can do something about it.  90% of the time you can get to a LOT more light with a short drive.

Do what I do on one of the dark days....take a look at the visible satellite imagery (here), find the nearest bright spots, and make your plans!

He May Like the Dark Side.  You don't have to.

Disastrous Heartbreak Ridge to Develop Next Week

The medium range models have been calling for the ridge from hell, which I'll officially name "Heartbreak Ridge", to develop over the western U.S. next week.

Just check out these GEFS forecasts from this morning.

Penn State E-wall
Blocking patterns don't come much uglier than that, and the ECMWF ensemble forecast (mean and standard deviation below) is no prettier.

I've been in denial about these forecasts for a few days, hoping that we might get something Sunday/Monday for storm chasing and to help the skiing some, but it hit me today that we could be hosed big time.

Let's put the situation into historical perspective. Records for the Snowbird SNOTEL cover 29 years.  Time series of snowpack water equivalent from 1 October to 1 February are provided below.  We currently sit at 3.1 inches, which I've circled below.  There are only 2 water years with less snow, 2000 and 2010.     

Source: NWS
That wouldn't be cause for panic, but that block scares the bejesus out of me. I stuck an arrow for the future on the graph, assuming we get perhaps 0.5" of water out of the Sunday/Monday storm (it could be more or less, but I lean toward less).  We would go through the first week of December near the lowest snowpack in that 29 years.  If instead we hold at the current snowpack, 3.1 inches would match the low for Dec 10 set in 2010.  

Although there is some uncertainty in the SNOTEL data, we're near the bottom of the barrel.  Note that the SNOTEL records don't go back to 1976/77, better known as the "drought year", when only 13.5" fell in November and 17" in December at Alta.  That year may have been worse.  

Bottom line: Burn skis, sacrifice a goat, or whatever else you can think of to conjure snow up Sunday/Monday.  We could be facing a really ugly start to December.  

About the only thing that keeps me going in times like these are thoughts of the 100 inch storm in November 2001.  We'll need it if this continues.  

The Deception of a Single GFS Forecast

Having cut my teeth in an era before ensembles, I confess some tendency to pull up the medium range forecasts produced by the Global Forecast System (GFS) when I arrive at the office during times of drought to scan for the next hope for flakes.

This is a colossal mistake, especially in the current pattern.  

The predictability at 4-7 days seems to have been remarkably low the past few weeks, leading to individual GFS forecasts that might be described as "all over the place," although one might argue that in general they have trended to drier as the forecast lead-time decreases.

Here's an example of the types of changes that one sees.  The 162 hour forecast from the GFS initialized at 0600 UTC 27 January November shows a major storm for all of the mountains of Utah on Sunday afternoon.  Great hope right?

Two days later, the 114 hour forecast from the run initialized at 0600 UTC last night says NO SNOW FOR YOU!

Let's consult an ensemble forecast.  We had a problem with our 0000 UTC NAEFS processing, so I'll use the plume forecast for Alta-Collins from yesterday.  Look at that spread.  There are some members generating over 3" of water and 30" of snow and others absolutely nothing.

I personally like plume diagrams, but the members producing the heaviest precipitation are the ones that attract the eye's attention and cause the heart to race.  Close inspection of that diagram shows that most members (about 40%) are producing less than 0.5" of water.  Thus, if one never bothered looking at the GFS, but instead the ensembles, the possibility that we're not going to see much is there.  On the other hand, the possibility of a larger storm is small, but not zero.

Humans like single model runs.  They are easy to look at and interpret, and they produce plausible forecasts.

However, for medium-range forecasts, they can be deceptive.  They do not provide estimates of the range of possibilities, and in a pattern like this, that's a problem.

I go to bed at night in a pattern like this expecting the worst (dust on dirt) and hoping for the best (major dumpage.

Prescribed burn in Wisconsin

GOES-16 Visible (0.64 µm, left) and Shortwave Infrared (3.9 µm, right) images, with plots of hourly surface reports [click to play MP4 animation]

GOES-16 Visible (0.64 µm, left) and Shortwave Infrared (3.9 µm, right) images, with plots of hourly surface reports [click to play MP4 animation]

* GOES-16 data posted on this page are preliminary, non-operational and are undergoing testing *

GOES-16 “Red” Visible (0.64 µm) and Shortwave Infrared (3.9 µm) images (above; also available as an animated GIF) showed signatures associated with a prescribed burn in western Wisconsin on 28 November 2017. The Shortwave Infrared images revealed a warm thermal anomaly or “hot spot” (dark black to yellow to red pixels) — and on the visible images, a thin smoke plume could be seen drifting southeastward from the fire source.

Early in the animation sequence, however, a band of cirrus cloud was moving over the fire — yet a faint thermal signature (darker gray to black pixels) could occasionally be seen on the Shortwave Infrared imagery. The cirrus cloud layer was thin enough to allow some of the heat energy emitted by the fire to pass through and reach the satellite detectors. Once the cirrus moved to the south, the fire’s hot spot became much more apparent.

A toggle between Terra MODIS Shortwave Infrared (3.7µm) and Infrared Window (11.0 µm) images at 1812 UTC (below) also showed a faint warm fire signature through the cirrus clouds — the cloud-top Infrared Window brightness temperature directly over the fire in northern Monroe County was -33ºC, while the warmest Shortwave Infrared brightness temperature of the subtle fire signature was +1ºC.

Terra MODIS Shortwave Infrared (3.7 µm) and Infrared Window (11.0 µm) images [click to enlarge]

Terra MODIS Shortwave Infrared (3.7 µm) and Infrared Window (11.0 µm) images [click to enlarge]

As was seen on the GOES-16 imagery, after the band of cirrus moved south of the fire an Aqua MODIS Shortwave Infrared (3.7 µm) image at 1912 UTC (below) displayed a pronounced fire hot spot signature.

Aqua MODIS Shortwave Infrared (3.7 µm) image [click to enlarge]

Aqua MODIS Shortwave Infrared (3.7 µm) image [click to enlarge]

(Thanks to Dave Schmidt, NWS La Crosse, for bringing this case to our attention!)

Lee-side cold frontal gravity wave

GOES-16 Lower-level (7.3 µm, left), Mid-level (6.9 µm, center) and Upper-level (6.2 µm, right) Water Vapor images, with hourly surface wind barbs plotted in yellow [click to play MP4 animation]

GOES-16 Lower-level (7.3 µm, left), Mid-level (6.9 µm, center) and Upper-level (6.2 µm, right) Water Vapor images, with hourly surface wind barbs plotted in yellow [click to play MP4 animation]

* GOES-16 data posted on this page are preliminary, non-operational and are undergoing testing *

As a strong cold front (surface analyses) moved southward from Colorado and Nebraska across New Mexico, Texas and Oklahoma on 28 November 2017, the subtle curved arc signature of a lee-side cold frontal gravity wave could be seen on GOES-16 Lower-level (7.3 µm), Mid-level (6.9 µm) and Upper-level (6.2 µm) Water Vapor images (above).

Closer views of imagery from each of the 3 water vapor bands are shown below.

GOES-16 Upper-level (6.2 µm) images, with hourly surface wind barbs plotted in yellow [click to play MP4 animation]

GOES-16 Upper-level (6.2 µm) images, with hourly surface wind barbs plotted in yellow [click to play MP4 animation]

GOES-16 Mid-level (6.9 µm) images, with hourly surface wind barbs plotted in yellow [click to play MP4 animation]

GOES-16 Mid-level (6.9 µm) images, with hourly surface wind barbs plotted in yellow [click to play MP4 animation]

GOES-16 Lower-level (7.3 µm) images, with hourly surface wind barbs plotted in yellow [click to play MP4 animation]

GOES-16 Lower-level (7.3 µm) images, with hourly surface wind barbs plotted in yellow [click to play MP4 animation]

Yesterday’s Storm Chasing Bounty

Yesterday's frontal passage was a bummer for skiers, providing little in the way of the white stuff in the Cottonwood Canyons, but provided us with plenty of excitement thanks to the Doppler on Wheels.

We deployed that morning to a rattlesnake speedway in the Utah desert where a dark cloud rose from the desert floor and we headed straight into the storm.  If you have no idea what that means, watch the video below. 

More accurately, we set up along the side of the causeway to Stansbury Island, just north of the Tooele Valley.  When I drove out to meet the team, the surface front had already pushed into the northern Tooele Valley, with low level "fractus clouds" seen in the photo below at levels just abouve the ground, near its leading nose.  At this time, the front was quite shallow.  

The rattlesnake speedway in the Utah desert was actually the Stansbury Island Causeway, ideal for surveying the frontal structure over the Tooele Valley and precipitation processes over the Oquirrh Mountains.  We could also can over the Stansbury Mountains (background below). 

The shallow nature of the front was very apparent in the radar data we collected.  A Doppler radar is capable of measuring how fast scatterers in the atmosphere, in this case snow and rain, are moving toward or away from the radar.  This allows us to use radar scans, known as PPIs, which are oriented at a slight angle to the horizon, to infer changes in the wind across the area and in the vertical.  In the plot below, cool colors represent flow toward the radar, warm away, with the radar in the middle of the image.  There is a clear indication of flow from the northwest near the radar and south-southwest at ranges more removed from the radar site.  Since the radar scan is tilted at a slight angle to the horizon, this is an indication of strong vertical wind shear in the frontal zone not far above the Earth's surface.  

We can also configure the DOW to scan in vertical slices, known as RHIs.  The RHI below is oriented to the east and scans over the southern Great Salt Lake, eventually hitting the lower slope of the Oquirrh Mountains near Point of the Mountain where they rise above the south lake shore.  Doppler velocities in this image are primarily away from the radar, consistent with northwesterly flow, except near the ground just to the west of Point of the Mountain.  This reflects the splitting of flow around the north end of the Oquirrh Mountains, with the flow there having perhaps a slight NNE component, which results in a weak flow component toward the radar (green).  

The DOW is also a polarimetric radar, which means it sends out and receives radar energy in two planes, one horizontal and one vertical.  The shape of the raindrops or snowflakes can be inferred using this information.  One product we use to do this is known as "differential reflectivity," with reflectivity the amount of radar energy is scattered by back to the radar.  If the horizontal and vertical radar energy is similar, the differential reflectivity is zero, and the precipitation is likely circular.  If on the other hand, the horizontal radar energy is larger, then the precipitation is wider than it is high, and the differential reflectivity is positive.  Dendritic snow, those wonderful flakes with six arms that produce blower powder, often produces high differential reflectivity, because the flakes tend to fall "flat." 

The RHI below shows two layers of high differential reflectivity (indicated by yellows).  One is near the ground and reflects the melting layer in which snowflakes are sticking together and falling relatively flat.  The other is farther aloft and likely reflects a layer in which dendrites are growing and falling.  The temperatures in this layer were likely between -12ºC and -18ºC, which favors the formation of dendrites.  

An interesting aspect of the storm was a near-complete lack of any enhancement of precipitation over the mountains.  It was a frontally forced event, rather than a mountain forced event.  In fact, it clearly precipitated more over the Tooele Valley than over the Oquirrh Mountains.  Only in the late stages of the event did the front finally decide to move over the Oquirrhs.  An example is the RHI of horizontal radar reflectivity below, taken looking east-south-east across the northern Oquirrhs.  The yellows are ground clutter from the radar energy bouncing off the Earth's surface, with the sloping area representing the western slope of the Oquirrh Mountains.  Above the ground clutter, the transition from light to dark blue reflects increasing precipitation toward the mountains, but this doesn't reflect an orographic effect, but is the frontal band as it moved into the Oquirrh Mountains.  

It's such a pity that the storm didn't produce much in the Wasatch.  However, thanks to the mobile capabilities of the DOW, we were able to go to the storm and get a wonderful dataset.  

Fronts Are Stronger Aloft Than At the Surface Here in the Northwest

Another strong Pacific front will come through tomorrow across western Washington State.  The temperature change will be wimpy at the surface, with a decline of only a few degrees.

But go aloft, to say 5000-10,000 ft, and major temperatures declines will occur (5-15F).

So why are fronts so wimpy at low levels around here?  The answer is below.

 Here is the forecast map for 10 AM tomorrow at 700 hPa (about 10,000 ft).  There is a trough of lower pressure along the coast (solid line), with a slug of cold air with it.  If you look closely, you will see a wind shift with the front aloft and a substantial change in temperature behind (roughly a drop from -8 to -14C, a decline of 6C or about 11F)

A map of sea level pressure and 925 hPA  (about 2500 ft) temperature at 7 AM, shows the front making landfall and a temperature drop of only a few degrees C (3-4F).

A wimpy front at low levels and a much stronger front aloft, at least in terms of temperature.  

The larger temperature changes with the front aloft are even more clear cut in a time-height cross section, which shows the temperatures (red lines), winds (barbs), and humidity (color shading) for the next few days (time increases to the left, height in pressure, with 850 being 5000 ft).  Look at the red lines (temperature)....much greater undulations aloft tomorrow (1128/12-1129/00) .
Why are the  surface temperatures variations associated with our vaunted Pacific fronts so weak around here?

Because of the influence of the vast Pacific Ocean.  At the latitude of our coast, the sea surface temperature is roughly 11 C (52F), extending for thousands of miles!
A blow up section off of our coast show 11-12C water temperatures due west of the WA Coast (below).  Cold air moves off of Asia and Alaska at all levels, but is rapidly warmed over the lower few thousand feet by the relatively warm ocean.  Over time the temperature contrasts with fronts are weakened at low levels, but remain fairly strong aloft.  Those are the fronts that reached our shores.

Want to see the process from space?  Here is a NASA MODIS satellite image from Saturday over the north Pacific.  You can see cold air streaming off of Asia to Alaska, produce lines of clouds resulting from the instability of cold air over warm water.  This process results in warming of the lower atmosphere.

So Much Weather, So Little Time

Looking west from the University of Utah Campus at about 7:15 AM this morning
It was a Mordor-like sunrise over the Salt Lake Valley this morning (e.g., above), simultaneously spectacular and apocalyptic.  With the overnight southerly flow, there is a decent plume of dust over the western valley (evident below the fire-lit clouds above), but clear skies to the east. 

The weather over the past 24 hours has been quite remarkable if you know where to look.  While it was quite mild across much of northern Utah during the holiday weekend, a thin lens of cold air remained over the relatively cold waters of the Great Salt Lake.  Our Hat Island observing site, for example, hasn't eclipsed 55ºF over the past five days. 

Source: MesoWest
Yesterday, that airmass made its presence known by penetrating into the northern Salt Lake Valley.  At 4 PM (2300 UTC), a sharp lake-breeze front extended across the central part of the valley with southerly flow to the south and northwesterly flow to the north.  Temperatures dropped from 70ºF in the Sandy "banana belt" to near 50 (or lower) near the Great Salt Lake.  Almost climatology! 

Source: MesoWest
The sounding from the Salt Lake City airport at about that time showed the shallow nature of the cold lens, with a surface temperature of just under 10ºC (49ºF), but temperatures of almost 17ºC at 827 mb (about 5600 feet). 

Source: SPC
Forecasters predicting a record high for the Salt Lake Airport were likely flummoxed! 

Fortunately, the south winds saved the "day", blowing the lake breeze north and causing temperatures to rise rapidly around 8 PM. 

Source: MesoWest
The transition from cold, damp lake-modified air to warmer, drier air was remarkably abrupt, and led to a high temperature for the calendar day of 69ºF.  That is the 2nd latest 69ºF temperature recorded at the airport on record (December 1st is the latest), and it happened in the dark.  In addition, a temperature of 68ºF was recorded after midnight, which sets a record high for today.   This is not your grandparents (or even parents) climate that we are living in.

Moving on to the weather this morning, we are still in the warm southerlies with some wind-borne dust over the western Salt Lake Valley.  However, cooler air remains over the Great Salt Lake with northwesterly flow developing over northwest Utah. 

The HRRR brings the surface front into the northern Salt Lake Valley at 1900 UTC (1 PM MST) this afternoon.  It will be a dry frontal passage, with the precipitation trailing the surface front. 

We will be "storm chasing" with the DOW, which we plan to deploy just south of Stansbury Island to examine the front penetrating into the Tooele Valley and across the Great Salt Lake.  Although a dry frontal passage, we're hoping there is enough dust in the air to be able to examine some fine-scale aspects of the frontal passage.  We'll then work on whatever precipitation comes through, focusing on the frontal-band interaction with the Stansbury and Oquirrh Mountains.

For skiers, this looks like a pretty pathetic event.  The models have the band fall apart as it moves in.  Snowshowers are possible, but it won't add up to much, as we discussed over the weekend.   Sad!