| Login Help
Home header left
Give button


The Convective Boundary Layer and the PM

In addition to the great service they offer to glider pilots, BLIPMAPS have given widespread currency to two key concepts in thermal soaring forecasts: The convective boundary layer, and the Forecast PM sounding. The two are interconnected. In the following discussion I will exemplify the defining properties of the convective boundary layer (BL), and argue that the use of forecast PM soundings, although seeming to elevate theory above practice in its reliance upon numerical models rather than balloon soundings, is often to be preferred over the traditional approach. I will also compare the PM sounding thermal forecast with the traditional AM balloon sounding forecast.

This tutorial assumes a decent familiarity with thermodynamic diagrams ("SkewT's") and their use in predicting lift height and strength and cumulus cloudbase. In a separate tutorial I have presented more than enough material to allow anyone interested to develop the necessary familiarity.

The Convective Boundary Layer

The following sequence of soundings illustrates the evolution of the convective BL at the Allentown-Bethlehem-Easton airport on July 3, 2005. The sequence starts at 1000Z (6 AM local) and ends at 2100Z (5 PM local). The red lines are the forecast temperature, the blue the forecast dewpoint. The most of the figures reproduced with permission from the excellent and widely used FSL interactive site.

The 6:00 AM sounding shows a nocturnal inversion from the surface (~400 ft) to about 1,000 ft. Above this inversion, between about 2,000 ft and 4,000 ft lies a region with a dry adiabatic lapse rate (DALR) - this region is essentially unchanged from its state late in the previous day where it was created by that day's convection. The origin of the nocturnal inversion is night time radiative cooling of the earth's surface. A second inversion, of quite different origin is also apparent beginning at about 4,000 ft. This subsidence inversion will generally persist throughout the soaring day. Photo #9687 | Sounding _0600
By 8 AM solar heating has begun to erode the nocturnal inversion and has forced the first 500 ft or so of the atmosphere to have a dry adiabatic lapse rate. Photo #9689 | Sounding_0800
One hour later the nocturnal inversion has almost gone, and a superadiabatic layer exists at the surface. Soaring to about 2,000 ft would be possible, at least for soaring birds which can generally be seen a few hours after the sun begins to hit the ground slowly rising in the growing but still shallow convective BL. It is clear from the 8 AM sounding that the surface temperature will only need to rise to about 23°C at which point the nocturnal inversion will be completely gone and the lapse rate will be dry adiabatic in a convectively mixed BL about 4,000 feet thick. This particular temperature is what is, or should be, understood by "trigger temperature". I will have more to say about this when comparing the forecast PM sounding method with the traditional AM balloon sounding method. Photo #9691 | Sounding_0900
The 11 AM sounding confirms that with a surface temperature of 23°C the soaring day should begin with lift to just less than 5,000 ft, the point at which the 23°C surface adiabat intersects the temperature lapse rate line. Photo #9693 | Sounding_1100
The 2 PM and 5 PM soundings show that as the surface temperature continues to increase the convective BL deepens eventually reaching a value of about 6,000 ft at 5 PM. Careful inspection of the superadiabatic layer at 2 PM and 5 PM reveals that by 5 PM the cycle is about to begin again: The shrinking superadiabatic layer will first disappear as insolation weakens to be followed shortly thereafter by radiative cooling and the establishment of another nocturnal inversion. Photo #9695 | Sounding_1400
Photo #9697 | Sounding_1700
Multi-Sounding Plots
Superimposing the 6 AM, 8 AM, and 11 AM soundings in a single skewT diagram very clearly shows how, by 11 AM, (green line, 1500 UCT) the nocturnal inversion is eroded and trigger temperature is reached.

Superimposition of the 11 AM, 2 PM and 5 PM soundings reveals something important which may not have been obvious from the individual soundings: Following trigger temperature the temperature lapse rate in the convective BL generally moves right as the day progresses. This of course is exactly what would be expected since warm air has been convectively transported aloft all day. Absent any other changes, this guarantees that the next day will need to be hotter to get the same kind of instability because the air aloft is already hotter before convection begins. Fortunately, it is often the case that a persistent flow of cooler and drier air sweeps away the warm air aloft allowing for another soarable day without too much in the way of higher surface temperatures.
Photo #9699 | Sounding_06001100
Photo #9701 | Sounding_11001700
The multi-sounding plots also reveal another basic attribute of the mixed BL: Surface moisture is transported aloft as the day progresses and accumulates at the base of the inversion. Inspection of the forecast soundings for this particular day suggests that it would have been blue, at least until late afternoon, but nonetheless, "dry" (i.e. blue) thermals would have transported moisture aloft. It can be seen that as soon as trigger temperature is reached the dewpoint aloft (green line) jumps to the right. Further convection increases the dewpoint aloft until, in the vicinity of the top of the BL, the temperature and the dewpoint start to converge. This is to be expected: The inversion traps the moisture can now only be removed by advection.


Thermal Soaring Forecasting Using AM and PM Soundings

AM Sounding Method

I don't use this method, but many do. It is taught in all introductory texts and is the source of the Thermal Index (TI) concept. Understanding how AM soundings have been used in making thermal soaring forecasts is essential to understanding how and why the PM sounding method came to be used, and why that approach is now generally accepted to be superior. It should be noted that the distinction I am drawing is between AM and PM soundings - not RAOB (balloon) and model soundings. The AM sounding methodology works just as well with model soundings. It works better when, as is generally the case, the nearest RAOB site is hundreds of miles away.

The closest RAOB site to the center of my local soaring area (KABE) is KOKX (Brookhaven) about 100 miles East. Next closest is KIAD (Dulles), about 150 miles SW followed by KPIT ( Pittsburgh) 250 miles West. The distances and locations of these sites exemplify the first problem in relying upon RAOB data: It's generally not available from a site close enough to be useful on most days. This problem at least is rather easily fixed: Just use a 12Z forecast sounding from one of many numerical models. Because these models have so many grid points it is always possible to generate a sounding close to the home airport (or anywhere else for that matter, such as at task turnpoints).

Photo #9703 | Uvalde_1200An interesting example of the use of a forecast AM sounding is Day 5 (Aug 8) of the 2005 15M Nationals held at Uvalde. This was a day when the winner flew over 350 miles at about 86 mph - strong conditions clearly. The starting point for producing a thermal soaring forecast using the AM sounding method is of course the AM sounding. The next step is to estimate the surface temperature for the forecast time and to construct the surface adiabat for that temperature. The (almost) straight red line to the right of the temperature lapse rate line is the 35°C surface adiabat. The sounding, and the surface adiabat can now be used to make the forecast: At some point, about 8,500 feet in this case, the surface adiabat intersects the temperature lapse rate and (assuming no cloud flying) this intersection represents the maximum height for the day. The next step is to determine whether or not cumulus cloud will form, and if so at what height. That too is easily done by constructing the mixing ratio line passing through the surface dewpoint. If that line intersects the 35°C surface adiabat then cumulus cloud will form at the altitude corresponding to the intersection. If there is no intersection there will be no cu. On this day cu are indicated at about 7,000 ft marked "CCL".

The morning sounding also predicts a "trigger temperature" (blue line). This temperature cannot be defined without answering two questions - "for what height, and for what thermal index?" In the US at least the height is 3,000 ft agl and the TI is 3°F degrees. Underlying the choice of these values are the assumptions that for cross country soaring it's nice to be at 3,000 ft, and that to have sufficient instability to climb at all the air in a thermal needs to be 3°F degrees warmer than its surroundings. The (3,000 ft) trigger temperature is found by constructing an adiabat which at 3,000 ft agl is 3°F to the right of the environmental temperature lapse rate. That temperature is about 29.5°C.

Photo #9705 | Sounding_0600_1100The term "trigger" implies a tipping point. Nothing about the AM sounding suggests such a point. On the other hand, inspection of the forecast sounding sequence for ABE on a typical soarable day does suggest an obvious tipping point: The completion of the erosion of the nocturnal inversion. When that happens there is a sharp discontinuity in the depth of the mixed layer.

It remains only to determine the strength of the lift for an assumed surface temperature and this is where one of the limitations of the method becomes clear: Very large TI's are indicated for the lower levels, up to about 4,500 ft. A fundamental problem with analyzing an AM sounding in this manner is that we know full well that the temperature lapse rate which will prevail at 2 PM is quite different. This means that the TI values, especially those in the first few thousand feet are meaningless. Lacking a PM lapse rate it simply is not possible to assign an average TI, or for that matter to pick a representative altitude. An additional problem, well illustrated in this example, is the sensitivity of the predicted lift height to changes in surface temperature. A one degree shift would change to BL depth by hundreds of feet. Anyone who has tried to even a few thermal soaring forecasts will appreciate that errors of 5 degrees (C!), let alone 1 degree, are all too common.

Photo #9707 | Uvalde_1800Let's take a look at the 18z forecast sounding for the day. Now the TI values from the surface to the CCL have the meaning we want them to have: The difference, at a chosen height, between the environmental lapse rate and the surface adiabat. That difference averages to about 3°C (~5.5°F) from 1,000 ft to 7,000 ft. I seldom see those kind of numbers here in SE Pa, but then we seldom see Uvalde conditions.

The forecast sounding based prediction of BL depth is much less sensitive to changes in the surface temperature: Were that temperature to be higher, the temperature profile of the mixed BL would shift right, lower, it would shift left. In neither case would there be a large change in the thickness of the mixed layer.


The convective boundary layer and the PM forecast sounding methodology are commutative. The PM sounding defines and quantifies the convectively mixed layer to which thermal soaring is necessarily confined. The convective mixing creates the lapse rates which define the sounding.

The forecast sounding presents the fullest and most accurate picture of the potential thermal soaring conditions. It forecasts the conditions expected to prevail at the time soaring will take place and it does so in a manner which makes its predictions less sensitive to errors in the forecast surface temperature.

Posted: 1/18/2007 By: Richard Kellerman

Member Blogs 

Search Posts

Recent Posts

Legal Notice

The SSA policy on member posting is located here