Tuesday 22 November 2016

Foothills Magic: A Brief Overview of Thunderstorm Activity in Alberta


The following is based solely on my current understanding of the meteorology and prediction of thunderstorms, and of how these specifically relate to the province of Alberta. It is in no way meant to be a formal discussion. Rather, the intended audience is other chasers and weather enthusiasts residing in Alberta that wish to gain a better understanding of how and why thunderstorms occur here and in the manner in which they do – though others may also find the material interesting. I will also include some suggestions for further reading within this blog post. I have created and annotated many of the images used in this post on Microsoft Paint, so please forgive me for their somewhat primitive appearance. And of course, expert input is always greatly appreciated.

The subheadings will be as follows:

- Introduction
- Patterns of Thunderstorm and Tornado Activity in Alberta
- “Foothills Magic”: How the Terrain Influences Patterns of Thunderstorms
- Using the Position of the Jet Stream to Anticipate Certain Thunderstorm Patterns
- Forecasting Tips
- Conclusion and Recommended Reading

Introduction


It may be helpful to begin this discussion with a review of some basic principles of convection and conceptual models. The basic life cycle of an ordinary thunderstorm, as well as the three basic ingredients needed for thunderstorms are as follows:



Ample moisture is needed for cloud matter to be able to condense in the first place. Moisture quantity can be expressed using a number of different variables (a topic which is beyond the scope of this post). However, for the lay person, the dewpoint temperature is a good indicator of the true amount of moisture in the air – especially compared to relative humidity (RH). RH is simply a measure of how close the air is to saturation; so while an RH of 90% may seem high at any temperature, the actual amount of moisture in the air between a temperature of say, 5C and 20C (at 90% RH and the same pressure) is much different. I generally take notice when the dewpoint temperature rises above 10C in Alberta, and the higher, the better – with an important caveat. RH does become important once the dewpoint (Td) is high. You may have a surface Td of 13C, but if the air temperature (T) is 32C, the relative humidity is actually quite low – which tends to lead to high-based storms with strong evaporative cooling potential (a bad thing for chasers). Thus, I generally get excited when both dewpoint and RH are relatively high, leading to smaller T-Td depressions (the difference between air temperature and dewpoint). Many chasers would agree that a temperature of say, 24C with a dewpoint of 17C (a T-Td of 7C with moisture to a good depth above the surface) would be money for an Alberta severe setup.

The atmosphere is said to have instability in a given layer when displaced air parcels (imaginary chunks of air) tend to accelerate in the direction that they were displaced. This occurs due to positive or negative buoyancy – especially when relatively cold, dry air lies atop warm, moist air. A buoyant near-surface parcel will, once lifted to a level where it is warmer (and less dense) than its surroundings, rise freely on its own until it once again cools to become the same temperature as its environment (the main process at work in thunderstorm updrafts. This occurs when saturated air parcels condense and release latent heat, enabling them to be warmer than their environment - especially in an environment with steep lapse rates. However, a more complete discussion about "conditional instability" will not be given for the sake of simplicity). Likewise, a cold (dense) air parcel forced downwards, will continue to accelerate downward to the ground through warmer and less dense low level air (as in thunderstorm downdrafts). In contrast, stable configurations tend to see displaced parcels return to their original positions. This condition typically occurs with temperature inversions (where temperatures increase with height through a layer), and thus can prevent buoyant low level parcels from rising through the layer to their level of free convection (LFC – above which they rise on their own due to positive buoyancy). This condition is dreaded by chasers when it works against them, and is known as a “cap”. There are a number of different indices to assess atmospheric stability (ie the LI, or Lifted Index), as well as other parameters that, more accurately, assess potential updraft strength (ie CAPE, or Convective Available Potential Energy). More information will be listed in the “Forecasting Tips” section below.

Lift, or lifting mechanisms, force air parcels upward. As discussed above, if the troposphere has a deep (throughout a great depth) unstable layer, a lifting mechanism can raise buoyant lower level air parcels to their LFCs, “triggering” deep, moist convection. Sometimes folks use the word “trigger” for what is assumed to mean “lift” – however, I find the latter to be preferable as it is more accurate and specific in describing the processes at play. There are numerous different lifting “mechanisms”, with some occurring in the low levels of the troposphere, and others occurring in the upper levels. Positive vorticity advection (PVA), warm air advection (WAA), and jet streak dynamics are all examples of processes that can lead to synoptic-scale rising air (recommended reading resources to follow). Air circulations resulting from the interaction of air masses of contrasting densities (ie along fronts, drylines, outflow boundaries, etc), as well as air being forced to rise over terrain features, represent low level lifting mechanisms. The vertical motions induced from these lifting mechanisms can be orders of magnitude greater than those resulting from upper air features.

The words “synoptic” and “mesoscale” will be used in this post, especially when describing the “synoptic background environment”. This is taken to mean the greater overall surface and upper air pattern surrounding the area of discussion. “Synoptic-scale” features are those that have spatial dimensions of over 1000km and time scales on the order of days (ie larger shortwave troughs and frontal systems/mid-latitude cyclones). “Mesoscale” features exist over areas of hundreds of kilometres or less, and last for several hours (ie sea breezes, mountain-plain circulations, etc). The following graphic gives an example of some major synoptic features:




The jet stream is a fast-flowing river of air that resides near the tropopause between contrasting air masses. Warm air masses occupy taller atmospheric columns of air, and thus have higher “heights”. The opposite is true for colder air masses. So when these two contrasting air masses come together, a jet stream develops near the top of the atmospheric columns, where the height imbalance is greatest. Greater imbalances yield stronger restorative forces, and thus a stronger jet stream. Once again, a topic worth investigating further, is how jet streams form, and related “forces” (such as PGF (pressure gradient force) and Coriolis force).

Depicted in the previous image are shortwave ridges and troughs, and associated upper convergence and divergence. What is shown is a very basic, oversimplified version of the processes at play (once again, worth reading up on this topic), but in general, upper divergence can be found immediately downstream (often east) of shortwave troughs, and upper convergence can be found immediately upstream of shortwave troughs (to the west). Divergence can be thought of as the spreading out of air parcels, and convergence as the piling up of air parcels. Therefore, when divergence occurs in the upper levels, the mass of air columns beneath it decrease – resulting in lower surface pressures (if not offset by other factors). Likewise, when air converges aloft, mass is added to air columns below, and surface pressures increase. The lower atmosphere responds to these changes of surface pressure in an attempt to establish equilibrium, with air flowing (or converging) into areas of low surface pressure, and away (diverging) from areas of higher pressure. When air converges (diverges) at the surface beneath divergence (convergence) aloft, rising (sinking) motion occurs through the depth of the troposphere, as shown in the idealized schematic. Areas where large-scale (synoptic) ascent is occurring are favourable areas of thunderstorm development, while areas of synoptic subsidence tend to suppress it. Once again, there can be numerous processes at play simultaneously, so this is somewhat of an oversimplification – however it is still helpful to envision how the air behaves near such synoptic features.

The general take home message here is that being downstream of shortwave troughs is where you want to be if you’re looking for thunderstorms, since the synoptic background environment is favourable to their development.  

Patterns of Thunderstorm Activity in Alberta


Maps of lightning flash density per square kilometre in Alberta (between 1999 and 2013), as well as reported tornadoes (between 1980 and 2009) are shown below. An annotated topographical map is also shown – which when compared to the other images, some interesting patterns can be seen.

Western Canada lightning flash density per square km, between 1999-2013. The area between Edson and Rocky Mountain House often sees the most annual lightning activity out of anywhere in prairie provinces.
Image retrieved from here

Image retrieved from here

Foothills region in purple hatched area.
Image retrieved from here



First, and most obvious, is the relative maximum of lightning flashes that were detected along the central foothills of Alberta. Also notice the tight gradient of lightning flash frequency between the central Rockies and the central foothills – suggesting that something special must be going on in that area. Lightning activity tends to peak along the Alberta foothills in an area bounded by Calgary, Hinton, and Swan Hills, with gradually less activity trending across the plains to the east. Some relative minima exist in the far northeast, the dry south, and along the continental divide.

Also notice the distribution of tornado reports across Alberta. They tend to be more aligned with the patterns of population across the province (not shown) and are more evenly distributed, as opposed to patterns of lightning. One can infer at least two things from this: that more tornadoes are likely to have occurred than what have been reported here due to the disparity between active areas of lightning and those same areas where few tornadoes have reportedly occurred (ie unpopulated areas where people are less likely to see tornadoes); and that despite more thunderstorms occurring along the foothills (west of Highway 2), more significant (F2+) tornadoes have occurred to the east of Highway 2 over the Alberta plain.

Nonetheless, the importance of the Alberta foothills in influencing thunderstorm patterns in Alberta cannot be overstated. So what is this “Foothills Magic” all about?

"Foothills Magic": How the Terrain Influences Thunderstorm Activity in Alberta

The shape of the land in Alberta plays a direct role in our weather and climate. Some common weather patterns that are a direct result of our lee-side geography include chinooks, upslope fog and precipitation events, cold air damming, lee troughs and Alberta “clipper-genesis”, occasional drylines, and foothills convection.

In the summer, when the sun is strong and synoptic pressure gradient weak, a noticeable mesoscale wind circulation – called the “Mountain-Plain” circulation (and technically the “Plain-Mountain” circulation during the day) – can develop near the foothills and adjacent plains. In the early morning, the summer sun warms the NE and E slopes of the front range of the Rockies more readily than the nearby plains. This leads to columns of rising air along the front range, with relatively low surface pressures resulting there. In response, moisture-laden air residing over croplands and parklands to the east is advected (flows) toward the foothills, causing it to gradually pool against the terrain. Thus, cumulus clouds typically develop here first, where there is both a source of lift (due to orographic convergence) and moisture. If there is sufficient instability and weak capping, towering cumuli may develop by late morning or early afternoon. The upper branch of the circulation is advected east over the Alberta plain, where the compensating subsidence leads to air sinking back to the surface, completing the circulation (and strengthening “capping” there).




Another factor that can enhance this circulation during the day (when the sun is overhead) is the difference in surface albedo. Darker surfaces, such as the conifer-covered foothills, have a lower albedo – leading to less reflection of solar energy back to space, and more absorption, translating to warming. Thus, sensible heat fluxes (flow) to the troposphere are greater over the foothills than the adjacent plains, leading to locally lower pressure and the reinforcement of easterly flow.


The Rocky Mountains themselves can behave as so-called “high level heat sources”, since the thermals that develop due to warming of their slopes lead to air pressure being lower here than at the same height to the east over the plains. Thus, under certain synoptic regimes, thunderstorm activity may be confined solely to the Rocky Mountains, and not even make it out into the foothills. Prevailing westerly winds ascend the continental divide, leading to less capping here than to the east, where it is subsides over the lee slope.

In any case, when there is a strong synoptic pressure gradient (evidenced by more tightly-packed surface isobars), stronger surface winds result – leading to these mesoscale wind circulations becoming embedded in the flow, and perhaps being overwhelmed. Still, even in energetic environments with an approaching upper disturbance, the foothills are often the first place to “break the cap” due to the forced ascent of moist easterly flow up the terrain. This makes the foothills the obvious play for most chasers on many setups, since convergence generated along the front range almost always goes, and often goes first (due also to the fact that upper disturbances come from the west over the mountains). Most storms that have become significant tornadic supercells in Alberta have had their origins along the foothills, being supported by upper energy that moved east out over the plains.

The differing synoptic regimes that determine the nature of thunderstorm activity will be explored in the next section, however it can be seen that the foothills “magic” isn’t so magical after all…it’s science! The take home message: the foothills readily provide a source of lift as well as moisture that can be pooled there with easterly winds.

Using the Position of the Jet Stream to Anticipate Certain Thunderstorm Patterns


Of all the varying upper air patterns that can occur, three main ones have been chosen for the purposes of this post: the upper ridge (being equatorward of the upper jet), the upper trough (being poleward of the upper jet), and being beneath the upper jet itself. (Note: it doesn’t necessarily follow that being south of the jet means you’re under an upper ridge, or being north of the jet meaning you’re under an upper trough).

Idealized Skew-T at right. While CAPE may be present, there is often very strong capping


On summer days in Alberta under an upper ridge, it is typically hot and sunny. Air is subsiding on the synoptic scale, so the development of deep, moist convection is typically stifled by a “steel cap”. Only areas with the strongest low level convergence (ie along the Rocky Mountains or foothills under a weakening upper ridge) have a chance of breaking the cap. Any storms are typically diurnally driven (as a result of daytime heating), occurring by late afternoon or early evening, and fizzling out soon after. Thus, these times are often frustrating for chasers, seeing TCU or CBs blowing up in unchaseable terrain, only to dissipate before reaching the prairies – or seeing no action at all. CAPE may build to tantalizingly high values, but it is inaccessible due to excessive CIN. Many lay folks think that hot weather automatically means thunderstorms - however, they are unaware of the larger factors at play that are preventing their development. Indeed, it is prime time for the beach!

However, the time spent under an upper ridge is crucial for Alberta, since evapotranspiration (ET) rates peak under hot, sunny conditions. Being located in a place that rarely taps into mT (maritime tropical, ie Gulf of Mexico moisture) air, we often heavily rely on our own home-grown moisture for storms (though sometimes we borrow good moisture from the eastern prairies). So long as a dry westerly doesn’t come and scour it all out, several days of stewing and brewing in Alberta – especially after a wet spring when plants and crops aren’t stressed – can lead to the deepening of low level moisture over the province owing to ET. This is “storm juice”, and can fuel an outbreak of violent thunderstorms upon arrival of the next upper trough. So be patient under the ridge, tune up that chase car, sip some brews, and watch the dewpoint climb!


Lack of a "cap" means overturning likely once convective temperature is reached



Upper troughs are a different story. They are typically characterized by cooler and more unsettled conditions, and under them is often a touch drier than before the previous trough came through. Cold air aloft under the upper low/trough leads to steep lapse rates (temperatures dropping off sharply with height). Thus, there is often still instability – the strength of which depending on residual low level moisture – however, there is usually a distinct lack of capping. Under this regime, convective overturning is more vigorous and widespread, and is underway no later than when the convective temperature is reached, which is typically by late morning or noon. By late afternoon, the troposphere is largely overturned (restoring stability, through warming of the mid and upper levels by updrafts, and cooling of the low levels by downdrafts).

Similar to being under an upper ridge, there is a lack of vertical wind shear (change of wind speed and direction with height), so most thunderstorms tend to be ordinary or “pulsy” (relatively short-lived) in nature. Any severe weather is related to storms being slow-moving (due to weak shear) and dropping large amounts of rain and hail on affected locations. However, this is often a fascinating time to observe mesoscale phenomena influence thunderstorm behaviour. The lack of a cap means that storms will tend to fire along any prominent convergence boundary that develops as a result of differential heating. The foothills are an obvious location for CI, but more subtle differences in albedo over the prairies and lakes to the east can result in deep convection there too. Weak deep shear and synoptic pressure gradients permit upright convection to develop atop axes of convergent winds, which can also generate columns of low level vertical vorticity. This, combined with characteristic steep low level lapse rates can lead to the development of “cold core” type funnel clouds, or weak landspouts, when TCU build atop columns of vertical vorticity. Thus, it is not uncommon to be looking at funnel cloud advisories for nonmesocyclonic funnels the day after a tornado watch for mesocyclonic funnels, as a result of the passage of the shortwave trough.

Southwesterly jet axis over Alberta: Game on!



On the morning of potentially severe days, you may have read in some PASPC discussions the expression: “Jet+Alberta= bad”. Indeed, a strong southwesterly jet stream crossing the mountains, along with the upper disturbances riding along in the flow, lead to a synoptically energetic background environment that is conducive to the development of severe convection. The main ingredient present in this setup, that the other two regimes lacked, is vertical wind shear. Deep layer, as well as low level vertical wind shear are the main factors determining the organization of deep, moist convection, into convective “mode”. Capping is also still usually present, though not to the degree it was in previous days under the retreating ridge. Therefore, your classic “loaded gun” sounding can result, awaiting the erosion of the cap both from below as a result of daytime heating and moistening, as well as above with cooling aloft ahead of upper disturbances. If these disturbances are timed perfectly, instability can build to maximum values during the day, only to be released in the late afternoon or early evening as explosive deep convection. The associated wind “speed maxes” of upper disturbances can enhance wind shear, making severe, and perhaps supercellular convection likely. (A reminder that a supercell is a thunderstorm with a rotating updraft). However, especially in environments with high CAPE, you don't want upper forcing associated with the disturbance to be too strong - lest linear convection (as opposed to discrete) result, regardless of the orientation of shear vectors to initiation boundaries.




It may also be worth mentioning here that upper energy associated with the jet stream can arrive at any time of day. Therefore, if it arrives too early, overturning may kill chances of severe convection later in the day. If it arrives too late, a blue sky bust may result, with “nocturnals” (nighttime lightning storms) occurring instead. Convection that occurs at night and in the morning is typically “elevated” – that is, it ingests inflow air from some layer above the surface. This occurs because the cooling of near-surface air as a result of OLR (outgoing longwave radiation) tends to cause it to be too stable to be lifted by a thunderstorm updraft. In contrast, daytime convection is often “surface-based”, when capping has been overcome and updraft air is comprised of buoyant, low level air. This is important for tornadogenesis, which depends in large part on the processes and properties of low level air. This is why nighttime tornadoes are exceedingly rare in Alberta when storms are elevated, since low level air is too cold (stable) to be lifted to permit the stretching of near-ground vorticity.


Forecasting Tips:

- If a big system has come through and scoured out all of your moisture, and you’ve kissed your double-digit dewpoints goodbye for a few days, it is going to take a while for it to build to sufficiently deep levels again. This is not to say that severe thunderstorms can’t happen with lower dewpoints and shallow low level moisture, but storm severity will likely be related to the possibility of marginally severe hail and the strong wind gusts that result below high-based storms with cold downdrafts. These lead to progressive cold pools that tend to undercut updrafts, especially when shear is weak. You can expect to see a lot of outflow dominant shelves during this time. This is only mentioned because a lot of folks become excited when they see, say, 13C dewpoints in Alberta. But if these have just "arrived", they are likely only existing in the "skin layer" and will be mixed out with daytime heating.

- Dispel the notion that “hot” weather is good for tornadoes. As was discussed previously, high boundary layer (the turbulent layer of the lower troposphere that is affected by surface-based processes) relative humidity, as a result of high dewpoint temperatures and comparatively low air temperatures lead to smaller T-Td spreads, which lower your LCLs (Lifting Condensation Level – essentially cloud base height).
This ensures minimal evaporative cooling potential beneath low bases – both of which are usually necessary for tornadogenesis.



- Discrete convection is favoured when deep shear vectors are at increasingly perpendicular angles to initiating boundaries (ie along Alberta drylines), while linear convection is increasingly favoured when deep shear vectors are parallel to initiating boundaries (ie along "crashing cold fronts" that are oriented SW-NE, with a SW-NE deep shear vector as well.

- The reader is further encouraged to read up elsewhere on how different shear profiles lead to different modes of convection. In fact, this is one of the more helpful points to learn as a storm spotter, since it helps you anticipate storm motion and behaviour in advance (ie whether storms will be multicellular, whether boundary interactions favour the development of subsequent convection, whether storms may grow upscale, approximate speed and direction of deviant motion, whether LP or HP supercells are favoured, etc). All of this has practical implications for chasers looking to formulate strategic interception tactics while maintaining situational awareness – both of which increase safety. Moreover, it can save you gas if you’re driving across the province looking only for supercells or tornadoes, if you know that while thunderstorms are possible, the spinny stuff is highly unlikely!

- In the same vein as the above, the reader is encouraged to read up on general principles of meteorology, so as to make the most of model output and realtime data in helping them choose targets and anticipate storm behaviour. Moreover, that red blob on radar can tell you a lot about your storm and how it may soon evolve! Learn about how radar works, and about how to identify different shapes, reflectivity and velocity “signatures”, reflectivity gradients, etc to make the most of its products.

- Don’t misuse severe weather parameters, and know what atmospheric processes cause certain values to increase or decrease. Learn how to interpret Skew-T diagrams and hodographs, to assess the thermodynamic and kinematic properties of a given environment, and how this may affect your target area. Here is a personal list of values that get me excited – especially when they occur all together:



These are merely rough guidelines for more potent severe setups (when occurring in tandem), and not magic numbers. The atmosphere often behaves unexpectedly, and anomalous events can occur outside of your typical "severe" parameter space.

Remember that you can have all the CAPE in the world, but that does not guarantee a storm. It only speaks to the potential for updraft strength, so it is a sin to overlook CIN! And even if you have a ton of CAPE and seemingly weak or no CIN, you still need a lifting mechanism to realize that CAPE.

The same can be said for 0-6km bulk shear and your helicity values. They mean nothing if you can’t get a storm to form in that environment.

Also, huge CAPE in the absence of any shear may mean you get a strong pulse storm, but the chance of a mesocyclone developing is almost nil (ie no supercell funnels or tornadoes). Likewise, extreme shear with comparatively small CAPE will tend to rip your updraft to shreds. A good balance of CAPE to shear is important.

A final note on CAPE. Notice I've used the prefix "ML" ahead of CAPE and LCL above. This is "mixed-layer" CAPE, and is, in my experience, superior to using "SB" (surface-based) CAPE as it is more realistic when considering mixing of air in the boundary layer as a result of daytime heating. Yes, SBCAPE values are usually higher (making them more tempting to use to "hype" a setup)! I'll typically consult "MUCAPE" (most unstable CAPE - the layer of the highest CAPE within the lowest 300mb of the troposphere) when considering elevated convection - which I'm probably not chasing anyway! Similarly with MLLCL, I find it more realistic than SBLCL for the same reasons as listed above. 

Conclusion and Recommended Reading:


All I can say is learn, learn, learn! Understand the weather processes in your chase zone, and know how different large-scale weather regimes will affect the behaviour of storms there. This will help you to chase safer and perhaps have greater success in finding the storm you want (though there is still a large element of luck in being in the right place at the right time!). Enjoy that “Foothills Magic”!

Recommended reading:

www.theweatherprediction.com

This is a great site that gives understandable treatments to complex meteorological topics – some of which have been alluded to above. Learn about how to read weather charts, upper air patterns and sources of ascent, how to use Skew-Ts and hodographs, etc.

METED/COMET also has a plethora of good resources. Those of you who use radar regularly (especially the app “Radarscope”) should check out the course “Weather Radar Fundamentals”:

https://www.meted.ucar.edu/training_module.php?id=960#.WDT8FlyKLVI

Or a course for SKYWARN spotter training:

https://www.meted.ucar.edu/training_course.php?id=23

More advanced readers will also appreciate the 3 part “Principles of Convection” module on Meted, as well as Rich Thompson’s “Tornado Forecasting” on YouTube.