Lessons in Boundary Layer Mixing, Transport, and Dispersion

A three-alarm fire occurred last night in an office building in Sugarhouse.  Sixty firefighters fought the blaze, which stared around 2 AM and continued into the morning.  One firefighter was injured and we hope they have a speedy recovery.

I observed the smoke plume during a morning hike in the foothills.  At 8 AM, the behavior of the smoke plume was strongly affected by the stable cold pool that forms overnight in the Salt Lake Valley.  The smoke rose perhaps a couple hundred meters before settling back down and being transported by the local flows, as illustrated below.  

I've annotated several key features.  The heating of a fire results in warmer air than the surrounding airmass.  As a result, the air becomes positively buoyant and rises, resulting in an ascending smoke plume.  As it does so, it cools and, if the air is stable as is often the case in the morning near the Earth's surface, it will eventually reach a level where it has the same temperature of the surroundings.  At this level, which is known as the equilibrium level, the air is neutrally buoyant.  However, the air continues to rise because it has momentum, but begins to decelerate.  Eventually, the updraft dies, at the overshooting top.  Here, the air is colder than the environment, and settles back down to the equilibrium level.  

In the photo above, the local flow was westerly at the fire site, resulting in the transport of smoke eastward (the photo is taken looking south).  However, near the time of the photo, there was an eddy present, so the smoke turned northward and then westward.  If you look carefully, you can see some smoke further east, above where I've added the label "old transport and dispersion."  Earlier that morning, there was a pronounced smoke layer here, suggesting that the eddy had just recently developed.  

The behavior of the smoke changed rapidly, however, by 8:46 am.  Note how in the photo below the smoke rises vertically through a much deeper layer.  It is also being transported westward.  

The vertical rise could be related to the warming of the air near the earth's surface.  At the University of Utah, where we collect 1 minute observations, the surface temperature increased about 5˚F during this period.  Another possibility is that the heat related by the combustion increased.  Either of these would result in a a warmer smoke plume able to ascend to greater heights.  In a wildfire, these two effects can be closely related since changes in the ambient atmosphere affect the combustion process such that wildfire intensity often increases during the day.  

The shift in the orientation of the plume is more difficult to explain.  On short time scales of minutes to tens of minutes, turbulence in the atmosphere and other larger-scale eddies can cause smoke dispersion to be fickle.  The outflow from Parleys Canyon typically results in easterly flow (which would transport smoke to the west) in the morning in the Sugarhouse area.  However, this morning at one observing site there was a clear shift in the wind from easterly to southerly at about 8 AM (see the image below), when the flow was weak.  After that, the flow was S-SW.  

These observations don't correlate all that well with the plume evolution, which illustrates the fickle nature of smoke transport and the challenges of using sparse observations to anticipate smoke transport during light wind periods.  This is a major challenge for anticipating the transport of smoke, pollutants, and other hazardous materials produced by fires, accidents, or terrorist incidents, especially during periods of weak flow. 

Lessons in Boundary Layer Meteorology

The boundary layer is the lower portion of the atmosphere that is affected by friction and the transfer of heat from the earth's surface into the atmosphere.  It's depth varies depending on the time of day, the environmental conditions, and the local land surface and topographic characteristics.

Typically the evolution of the boundary layer is conceptualized as shown below.  During the morning, sunlight heats the Earth's surface, resulting in a transfer of heat into the atmosphere.  This erodes the shallow stable layer that typically forms overnight from radiational cooling.  The boundary layer then continues to grow, eventually reaching maximum depth later in the day (the time varies depending on conditions).  Within the boundary layer, turbulence driven by wind shear and surface heating results in considerable mixing.  Concentrations of gasses like water vapor and carbon dioxide are often nearly constant with height in the boundary layer.  Hence, the "mixed layer" label in the schematic.  Often there is a stable layer or inversion at the top of the mixed layer. 

Source: COMET/METED

Near and after sunset, the Earth's surface cools rapidly.  Heat is transferred from the atmosphere, which cools rapidly near the Earth's surface.  This forms a shallow nocturnal stable boundary layer, that may be tens or perhaps one or two hundred meters deep.  Above this layer, the old remnants of the boundary layer remain.  This layer is called the residual layer.

In quiescent weather conditions, this pattern repeats itself daily: 1. The sun rises;  2. The nocturnal stable boundary layer is "burned off";  3. The boundary layer grows rapidly into the residual layer and constituents (including pollutants) are mixed through its depth;  4. The sun sets and the nocturnal boundary layer forms and strengthens, leaving a residual layer aloft.

Evidence of these processes was very apparent on my ride above Ensign Peak this morning.  At the time, the top of the pollution layer was perhaps 1 km above the valley floor.  There was a very clear discontinuity in the pollution at that level.  It was early enough that I suspect that discontinuity did not mark the top of today's boundary layer, but instead the top of the residual layer, which was loaded with pollutants from emissions from yesterday.

I've added by eye yesterday's sounding (red=temperature, green=dewpoint) from the airport.  In this case, the temperature or dewpoint decrease if the line slopes to the left and increase if it slopes to the right.  The top of the residual layer was very near the elevation of an inversion in yesterday's sounding, as one might expect.  Below that inversion, the atmosphere yesterday afternoon was relatively well mixed.  For example, temperature decreased rapidly with height at a rate of about 10ºC per kilometer, which is consistent the density being constant with height.  Dewpoint decreased with height at a rate close to that expected if the concentration of water vapor is constant with height.  However, the decrease with height near the surface was more rapid than one might expect if the atmosphere is well mixed.  This isn't unusual as the turbulence typically can't mix the atmosphere fast enough right near the ground to make the water vapor concentration constant with height if there is evaporation or transpiration occurring.

Expect views like this most days this week due to the presence of high pressure.  Note that the sharp top of the residual layer is most apparent in the morning and evening if you are at an elevation somewhat above the valley floor and you face somewhat toward the sun (but perhaps not right at it).  This is an optical effect related to how pollution scatters sunlight.

Announcement

I'm pleased to announce that we will be exhibiting the Doppler on Wheels mobile radar (pictured below) at the Natural History Museum of Utah this coming Saturday, November 4, from 10–5 PM.  Bring yourself, the family, and friends.  The exhibit is also one of their Behind the Scenes days when they open the place up for a public viewing of their collections.  Come and geek out!

Yesterday's Big Blow

Some impressive winds were reported around northern Utah yesterday and I think it's safe to say that they exceeded forecast expectations.  Peak gusts at valley locations reported to MesoWest include the following:

Sherwood Hills: 82 mph (0210 UTC/2010 MDT)
SR201 at I-80: 80 mph (0130 UTC/1930 MDT)
Syracuse: 76 mph (0120 UTC/1930 MDT)
Great Salt Lake Marina: 73 mph (0136 UTC/1936 MDT)

Many mountain sites, especially in the northern Wasatch, also reported strong gusts.

The setup for the event was very similar to many other strong southerly wind events in the spring.  With a developing surface trough extending from the southern Sierra Nevada across central Nevada and northern Utah.

RAP sea level pressure analysis, surface observations, and GOES IR satellite imagery at 0100 UTC 16 June 2016

In such situations, strong surface heating typically has two effects.  One is to contribute to the intensification of the surface trough and the associated low-level pressure gradient.  The other is to grow a deep surface-based mixed layer or what meteorologists sometimes call a convective boundary layer or CBL.  Thermals driven by intense heating penetrate upwards through the CBL, and these updrafts are utilized by gliders for lift.  The updrafts are just one side of the story, however, as the CBL also features areas of descending motion.  These updrafts and downdrafts transport and mix momentum through the CBL and lead to gusty surface winds.

During the spring, meteorologists commonly use the 700-mb flow to get some idea of the potential for strong winds under such conditions, but late yesterday, the 700-mb flow wasn't really all that impressive with modest southerly flow (30-35 knots) east of the trough over Utah.  In the spring, I've seen stronger flow than this produce less impressive winds than we saw yesterday.

RAP 700-mb geopotential height and winds and GOES IR satellite imagery at 0100 UTC 16 June 2016

Yesterday's CBL, however, was remarkably deep and so we were mixing momentum through a much deeper layer, tapping into stronger momentum air above 700 mb.  As shown in the afternoon (0000 UTC/1800 MDT) sounding from the Salt Lake City Airport, the CBL extended to almost 450 mb, or about 20,000 feet above sea level.  Winds between 700 and 500 mb peaked at about 50 knots, suggesting that the 700-mb analysis doesn't capture the strongest winds in the CBL and thus the potential for high gusts at the surface.

Source: SPC

On the other hand, even looking at that sounding in hindsight, I'm not sure if I would have gone for 80 mph valley gusts.  Nevertheless, I'll be making a mental note to pay closer attention to the CBL depth and the strength of flow throughout the CBL, especially for potential events in mid June when the surface heating is so strong.

Recipe for a Warm Night

Last night, temperatures fell little from yesterday afternoon's peak.  At the University of Utah, for example, we peaked at 72ºF yesterday afternoon.  Climatologically, the overnight minimum is roughly 20ºF lower than the afternoon maximum, but last night temperatures remained mild all night, dropping only briefly to a minimum of 61ºF.

Source: MesoWest

The recipe for a warm night is relatively well known.  Rather than clear and calm, you want clouds and wind, and we had them both last night.  But what makes the clouds and winds so effective at keeping temperatures elevated?

Let's start with the clouds.  It is often said that clouds act like a blanket, but this is a terrible analogy if ever there was one.  A blanket keeps you warm because it prevents the mixing of air near your body with cooler environmental air.  This slows the net loss of heat from your body to the atmosphere. In contrast, clouds contribute to warmer nights by providing an additional source of energy to the Earth's surface—namely infrared (a.k.a., long wave) radiation emitted by the clouds themselves.  Often it is said that clouds "trap", "reradiate", or "reemit" radiation coming from the Earth's surface, but this is also an oversimplification that is somewhat misleading (we will save that discussion for another day).  With this extra source of energy, the Earth's surface cools at a rate slower than it would on a clear night.  All else being equal, low clouds usually result in warmer nights than high clouds because low clouds are usually warmer and thus emit more infrared radiation.

Now on to wind.  On a calm night, there is typically very little turbulence to mix the air near the Earth's surface.  As a result, the cooling is concentrated in a very shallow layer and temperatures fall dramatically.  Sinkholes and basins often observe the lowest overnight minimum temperatures as they become very calm at night.  Wind generates turbulence, however, and instead of forming a shallow layer near the surface with very cold temperatures, you are constantly mixing the air, leading to warmer conditions near the surface.

So, last night with extensive cloud cover and strong winds, we simply didn't see temperatures drop as they do on a clear, calm night.  Thank the radiation from the clouds and the mixing from the wind.

Intricacies of Local Temperature Trends, Part I

As discussed in the previous post, there's a pretty good chance that the average temperature for this June and July will be the highest observed since 1948 (when I began my analysis) and very possible since the beginning of records in the Salt Lake Valley in 1874 (at least that's as far back as the National Weather Service goes).

The average temperature is typically calculated by averaging the maximum and minimum temperature rather than calculating an average based on hourly observations (although not a necessity currently at the Salt Lake Airport.  Such an average, however, obscures the dramatic difference in the trends of maximum and minimum summertime temperatures over the past few decades.

The graph below presents the average summertime maximum temperature at the Salt Lake City Airport each year since 1948.    There are large ups and downs from year to year, but a very gradual upward trend during the period.  The average maximum summertime temperature during 1948–1957, the first decade of this record, was 87.4ºF.  Over the past decade (2003–2012), however, it was 89.6ºF.  An increase of just over 2ºF.

Source: http://xmacis.rcc-acis.org

In contrast, the graph below presents the average summertime minimum temperature at the Salt Lake City Airport each year since 1948.  One also sees large ups and down from year to year (note that these roughly correlate with the ups and downs in maximum temperature), but the trend is larger and much more apparent.  The average minimum summertime temperature during 1948–1957 was 57.9ºF.  Over the past decade (2003–2012), however, it was 63.0ºF, an increase of about 5ºF, roughly 2.5 times the rate of the maximum temperature increase.

This is not an uncommon finding.  Over the past few decades, many stations have seen a larger increase in minimum temperature than maximum temperature, which reduces what meteorologists call the diurnal (daily) temperature range or DTR.

In a future post, we will talk about possible causes for these trends at the Salt Lake City Airport, but for today I merely want to discuss why the maximum and minimum temperature trends might differ.

The layer of the atmosphere that is in contact with the ground and through which air is mixed is called the boundary layer.   In the afternoon during the summer, the boundary layer over Salt Lake City is typically quite deep, often extending more than 2000 meters (6000 feet) above the valley floor.  Eddies and turbulence distribute energy and pollution through the boundary layer and keep it well mixed (the updrafts favored by paragliders are a result of larger eddies in the boundary layer).  Think of a pot of water in full boil.

In contrast, at night and during the early morning when minimum temperatures are typically observed, the boundary layer over Salt Lake City is typically quite shallow, perhaps 100 m (300 ft) or less.  There's usually an inversion present near the surface (unlike our wintertime inversions, summertime nighttime inversions burn off quickly when the sun comes up).  Eddies and turbulence are quite weak and only mix air through a shallow layer.

As a result, the nighttime boundary layer is very sensitive to any change in the surface energy balance.  If you provide an input of energy during the night, it is distributed through a shallow layer.  In contrast, the same energy input would be distributed through a very deep layer in the afternoon.  As a result, the same energy input in the morning typically results in a larger temperature increase than during the afternoon.

Chances are you've observed this first hand.  What happens when the sun comes up?  The temperature rises very rapidly.  In contrast, in the afternoon, the temperature rise is typically very slow, despite the fact that the incoming solar energy is much larger.  This is because during the morning, the small amount of energy provided by the low-angle sun is distributed through a shallow layer, so it gives you a big temperature increase, whereas in the afternoon, the large amount of energy provided by the high-angle sun is distributed through a deep layer, so you get less bang for the buck (there are some other factors at play, but we won't worry about it for the sake of this demonstration).

The bottom line is that the minimum temperature is more sensitive to a change in the surface energy balance.  In the near future we will explore some of the factors that might be causing the warming trend at the Salt Lake City airport, including the decreasing diurnal temperature range.  

Changes on the Playa

I've spent the last two weekends collecting field observations over the playa of the Great Salt Lake Desert.  The playa is an incredibly fascinating land surface.  One might even argue that it lives and breaths as it undergoes incredible changes due to weather, seasons, climate variability, climate change, and human disturbance.

For example, the storm on Friday and Friday night caused some noticeable changes to the brightness, texture, and wetness of the playa.  Compare the two photos below, which were taken last Sunday and yesterday.  They were taken in slightly different locations, but generally reflect the characteristics of the playa in the area in which we were operating.

Sunday 7 October

Sunday 14 October

Playas are known to have a strong influence on local circulations, not just in Utah, but other dryland regions of the world.  It will be interesting to see how the changes above affected the surface energy balance (e.g., the absorption of solar radiation, fluxes of heat and moisture to the atmosphere, etc.) and if those changes have an influence on airflows in and around the Great Salt Lake Desert.

It wasn't all hard work.  We did find time early yesterday afternoon to bag Volcano Peak just north of Wendover in the Silver Island Mountains.

Lessons in Land Surface Contrasts

I spent the day today at Dugway Proving Grounds where we will be running a major field program this coming year to improve weather prediction in complex terrain.  It was a long trip, but fascinating meteorologically because it offered up a nice perspective on how land-surface contrasts affect minimum temperatures.

The plotted temperatures below are the 24-h minimum temperature reported to MesoWest (the winds are from 2300 UTC so ignore them).  Notice that the minimum temperatures in the Salt Lake Valley vary widely (ignore a few as they are clearly bad), with a 57F at KSLC, a few 60s where the stations are a bit higher, and some 50s scattered about elsewhere.

Now, take a look to the west.  There is a more coherent pattern.  Over the Salt Flats, the minimum temperatures range from 56-62F, but if one moves south and east of the Salt Flats, where there is a higher density of stations near Dugway, the minimum temperatures are much lower, ranging from 43–50F.  Stations above 50F are at somewhat higher elevations and either in or above the morning inversion.

The contrast between the Salt Flats and surrounding desert illustrates nicely how the land-surface affects the surface energy balance.  The Salt Flats have a larger thermal inertia than the surrounding desert land surface.  This means it takes more energy compared to the surrounding desert to cause the same temperature change.  As a result, they cool off more slowly at night and typically see a higher minimum temperature than the surrounding desert.  They also heat up more slowly during the day and typically see a lower maximum temperature.  The site in the Salt Flats along I-80 was even warmer than KSLC this morning.

Topography strongly affects suface temperatures in northern Utah, but land-surface contrasts are quite important as well.  

Minimum Temperature Goes Nowhere

Despite my best wishcast, the temperature went absolutely nowhere last night at KSLC, remaining at about 8-10F for the entire night.  This occurred despite light winds (calm at times), clear skies, and low dewpoints, especially toward morning.

The minimum thusfar is 8F, and there's no way it will drop into the negative numbers.  What a waste of a perfectly good arctic airmass!

It is really quite remarkable that the temperature at KSLC did not drop at all.  Data collected for the PCAPS field program may shed some light on this, but I can't help but hypothesize that radiational cooling of the Earth's surface was countered overnight by the ground heat flux (i.e., heat flux from depth toward the surface).  If we had snow on the ground, the result would have been much different (and colder).  The magnitude of the ground heat flux is somewhat large at present because this event is so anomalously cold.  Thus, it may have put the breaks on the nocturnal cooling.

Other ideas?