Retired senior lecturer in the Department of Meteorology at Penn State, where he was lead faculty for PSU's online certificate in forecasting.
By: 24hourprof, 1:23 PM GMT on August 29, 2013
When I arrived at Penn State in the early 1980s, Toby Carlson and I had several discussions about very warm air with very low relative humidity originating over the Sahara Desert and moving across the Atlantic between 5,000 and 15,000 feet during the summer and early autumn months (essentially, hurricane season). Toby's research on the origin of dusty air over the Atlantic intrigued me (check out Toby's paper published in 1972), especially his observation that Saharan dust contains distinctive minerals such as radon-222, which allows scientists to forensically trace the source of dusty air over the Atlantic Ocean back to Africa (here's the abstract of an early paper Toby published in Science circa 1970).
At any rate, when I read Dr. Masters' blog on the Saharan Air Layer (SAL) on Sunday, August 25, I jumped at the chance of contributing to his interesting discussion. I include the SAL analysis and the infrared satellite image from Meteosat at 12Z yesterday morning below, courtesy of the Cooperative Institute of Meteorological Satellite Studies. The yellows, oranges, and reds indicate various concentrations of Saharan dust, while cold cloud tops associated with showers and thunderstorms appear white on the infrared imagery.
The 12Z analysis of the Saharan Air Layer. Concentrations of dust are depicted by yellow, orange, and red. These data are superimposed on the 12Z infrared image from Meteosat. Courtesy of the Cooperative Institute for Meteorological Satellite Studies.
There are several consequences of Saharan Air moving across the Atlantic, including the negative impact that the SAL has on tropical cyclones (or developing easterly waves coming off the African Coast). I'll talk more about the SAL and easterly waves / tropical cyclones later in my blog, but, first, I believe it's important for readers to understand the prevailing weather pattern over Africa and the tropical Atlantic during late summer and early fall, when the frequency of easterly waves traveling westward from Africa typically increases.
The 12Z GFS temperature (red) and dew-point soundings over the Sahara Desert (23.84 degrees north, 15.42 degrees east) on August 28, 2013. Note the well-mixed layer extending from the desert floor to 600 mb (about 4 kilometers up). The wide separation of the temperature (red) and dew-point (green) indicates very low relative humidity ("low RH"). Courtesy of NOAA.
I'll start with the Saharan Air Layer, which, obviously, originates over the Sahara Desert from late spring to early fall. Considering its source region, it should come as no surprise that the SAL is very warm, very dusty, and has very low relative humidity. I randomly chose a point over the Sahara and generated the GFS model skew-T at 12Z yesterday morning (above). As you can see, the troposphere was well mixed from the earth's surface to about 600 mb (about 4000 meters). At times, the well-mixed layer can extend to 500 mb (roughly 5500 meters). Like clockwork, the SAL moves westward off the coast of Africa every few or several days during hurricane season. Of course, hot, desert air moving off the coast of Africa gets modified (low levels are cooled), so a reasonable general rule is that the SAL is largely found between 850 mb and 500 mb (1500 meters to roughly 6000 meters) as it moves westward over tropical Atlantic waters.
There is also vertical wind shear associated with the Saharan Air Layer. Let's explore this idea further by looking closely at conditions over Africa during summer. Given the major-league heating over the Sahara from late spring to early fall, it stands to reason that a summer monsoon (a seasonal reversal of wind direction) develops over central Africa (see this schematic, which displays the monsoonal regions of the world based on a monsoon index devised by S.P. Khromov in 1957.). Indeed, southerly winds over the Atlantic and Indian Oceans develop in response to the heating over the Sahara, carrying tropical moisture northward toward the southern edge of the great desert. To see what I mean, check out the GFS model analysis of streamlines representing 10-meter winds at 12Z on August 28 (below; larger image). I circled southerly winds associated with monsoonal flow...note that southerly winds in both areas have a westerly component (I'll refer to this westerly component in just a moment).
The 12Z GFS model analysis of 10-meter streamlines over northern and central Africa on August 28, 2013. Larger image. Courtesy of Penn State.
Meanwhile, scorching, hot Harmattan winds blowing from the northeast (revisit the GFS model analysis above) converge with the southerly monsoonal flow, creating a relatively strong temperature gradient stretching east-west across Africa. To see what I mean, check out (below; larger image) the 12Z GFS model analysis of 2-meter temperatures on August 28, 2013. Local meteorologists used to refer to this feature as the Intertropical Front (ITF).
The 12Z GFS model analysis of 2-meter temperatures on August 28, 2013. The convergence of hot desert winds and cooler, monsoonal winds creates a frontal zone sometimes called the Intertropical Front. Larger image. Courtesy of Penn State.
I'll add here that hot desert air ascends slantwise over the cooler (and more dense) tropical air hugging the earth's surface, paving the way for showers and thunderstorms. Check out the idealized cross section below, which nicely summarizes the discussion about monsoonal flow and the scorching, hot Harmattan winds).
An idealized cross section (larger image) showing the convergence of moist monsoonal winds and hot Harmattan winds, forming the Intertropical Front. (ITF). Hot Harmattan winds overrun cooler, denser moist air, setting the stage for recurrent showers and thunderstorms. Courtesy of, and copyright by, the Penn State online certificate program in weather forecasting.
The horizontal temperature gradient associated with the ITF decreases with increasing altitude above the earth's surface. To understand my point, check out the flash below. On the left, there are two soundings on the north and south side of the Intertropical Front...one in the hot, desert air mass and the other in the cooler tropical air mass. Note that temperature decreases faster in the hot desert air mass (essentially, dry adiabatic) than the cooler tropical air mass (here, the release of latent heat of condensation slows the rate of temperature decrease with increasing altitude. The bottom line is that the horizontal temperature gradient decreases steadily up to about 650 mb.
(Left) The temperature soundings in the hot, desert air mass over the Sahara (orange) and the cooler, more moist air associated with monsoonal flow (green). Because temperature decreases faster with increasing altitude in the hot air mass, the relatively large temperature gradient at the earth's surface decreases with increasing height, eventually going to zero at about 650 mb. (Right) The vertical variation of the zonal (east-west) component of the wind with altitude. Low-Level winds lose their westerly component, given the "reversed" north-south temperature gradient. Eventually, winds develop an easterly component, which increases with increasing altitude up to 650 mb, where the Mid-Level African Easterly Jet (MLAEJ) resides. Larger image. Courtesy of, and copyright by, the Penn State online certificate program in weather forecasting.
At about 650 mb (3500 meters), the horizontal temperature gradient goes to zero, marking the pressure altitude where the fastest easterly winds typically blow. Hold that thought for a moment. The plot on the right shows how the zonal (east-west) component of the wind changes with altitude. Remember when I said that the southerly monsoonal winds at ten meters have a westerly component? Well, you can see their footprint on the lower part of the plot. But the "reversed" temperature gradient causes the westerly wind component to decrease with height above the ground. Indeed, winds develop an easterly component in the lower troposphere, which then increases with increasing altitude until 650 mb (3.5 kilometers), where the temperature gradient goes to zero. At this altitude, easterly winds reach a maximum speed.
The 12Z GFS analysis of 650-mb isotachs (in knots) and 650-mb streamlines on August 28, 2013. The elongated maximum in easterly winds is the Mid-Level African Easterly Jet. Larger image. Courtesy of Penn State.
This maximum of easterly winds is the Mid-Level African Easterly Jet, which you can readily observe on the 12Z GFS model analysis of 650-mb isotachs (in knots) and 650-mb streamlines on August 28, 2013 (above; larger image). I note that this mid-level jet lies in the vicinity of the Intertropical Front.
In a similar fashion, Saharan air layers moving off the coast of Africa increase the easterly vertical wind shear over tropical seas, creating an unfavorable environment for the further development of easterly waves.
If wind shear doesn't do the trick, entrainment of air with low relative humidity will. Indeed, when an embryonic tropical wave entrains SAL air in the middle troposphere, chances are that it will not become a tropical cyclone. That's because mid-level air with low relative humidity entrained into the central-core thunderstorms causes stronger downdrafts to develop (entrainment promotes evaporational cooling that promotes sinking parcels of air to become more negatively buoyant...see the idealized schematic describing the impact of entrainment of air with low relative humidity below; larger image).
The impact of the entrainment of air with low relative humidity on thunderstorm downdrafts (in this context, "dry" means low relative humidity). Larger image. Courtesy of, and copyright by, the Penn State online certificate program in weather forecasting.
In turn, stronger downdrafts are able to penetrate into the boundary layer, injecting lower theta-e air (air that's more stable), which, acts to snuff out new convection in the vicinity of the central core. As a result, new convection is forced to develop farther away from the central core. Air pressure then decreases away from the core in response to the convective warming associated with these new thunderstorms. As a result, the pressure gradient decreases across the cloud cluster, thus weakening the already fragile system. The die is now cast and the system fizzles (or at least doesn't develop any further).
Here endeth my two cents.
Updated: 2:35 PM GMT on September 04, 2013
By: 24hourprof, 7:51 PM GMT on August 26, 2013
The photograph below, courtesy of the National Weather Service in Reno, Nevada, shows smoke from the American fire in Placer County, California, arriving in the Reno / Carson City area during the evening of Sunday, August 18, 2013 (you're looking southwest toward downtown Reno from the NWS Reno office; larger image). Note the shape of the leading edge of the smoke, which is consistent with the conceptual model of an outflow boundary (gust front) advancing eastward from a group of showers and thunderstorms, which had developed just to the west of the Reno-Carson City area (here's the 0037Z image of base reflectivity from the radar near Reno, NV, KRGX, on August 19 (5:37 P.M. PDT on August 18).
A wall of smoke invaded the Reno-Carson City area late Sunday afternoon as winds accompanying an outflow boundary spread eastward from showers and thunderstorms (0037Z image of base reflectivity) west of Reno and Carson City. Larger image. Courtesy of the National Weather Service.
The smoke from the American fire was the first round for the Reno area during the past week. Indeed, more smoke from the Rim fire near Yosemite National Park in California arrived over the Reno Carson City during the week (MODIS satellite image on Friday, August 23). As you probably already know, the Rim fire is still wreaking havoc in and around Yosemite National Park (and threatening majestic Sequoia trees).
Check out part of the Area Forecast Discussion (AFD) that the National Weather Service issued on Monday morning, August 19, documenting the role that outflow boundaries played in transporting smoke from the American fire into the Reno-Carson City area. Below is the annotated image of the base reflectivity at 0042Z on August 19 (5:42 P.M. PDT on August 18). I annotated the culprit outflow boundary. The corresponding image of Doppler velocities marks the footprint of the winds accompanying the eastward advance of rain-cooled air behind the outflow boundary (the radar at Reno, KRGX, lies in the upper-right corner of the two images). I added two white arrows to roughly indicate the winds behind the gust front (keep in mind that, for any given location, Doppler radar only measures the component of the wind blowing directly toward and/or directly away from the radar).
Four minutes later, the 0046Z images of base reflectivity and base velocities indicate that the outflow boundary was poised to move into the Reno area. I note that rain from these rather widely separated showers and thunderstorms were generally light (rainfall from 12Z on August 18 to 12Z on August 19), so please don't be fooled into thinking that there has to be heavy rain for the associated outflow boundaries to make a large impact on local weather.
The 0042Z image of base reflectivity on August 19, 2013 (5:42 P.M. PDT on August 18). The radar site near Reno, NV, KRGX, is located in the upper corner of the image. Larger image. Courtesy of the National Weather Service.
Just so everybody is up to speed on the conceptual model of an outflow boundary, I'll end my blog by delving a bit deeper into the scientific details. In general, the rain-cooled outflow from a thunderstorm qualifies as a density current, which is the technical term given to any relatively dense fluid that flows when injected into a less dense fluid. To see what I mean, check out this photograph of a density current in a laboratory tank, courtesy of John E. Simpson, University of Cambridge) and this computer simulation of a density current, courtesy of David Wojtowicz, National Center for Supercomputing Applications. Note that the greatest depth of the density current typically occurs along and near their leading edges.
An idealized schematic of an outflow boundary. Streamlines on the top panel indicate the downdraft of a thunderstorm and the winds behind and ahead of the outflow boundary (the vertical dotted line marks the position of the outflow boundary). Pressure, winds and temperature associated with the gust front appear in the second, third, and bottom panels, respectively. Courtesy of Penn State's online certificate program in weather forecasting.
A pressure-gradient force drives any density current. In the case of thunderstorm outflows, higher pressure forms in the rain-cooled air. For the most part, pressure increases in response to the rain-cooled air's higher density and greater column weights(I'm assuming hydrostatic equilibrium). To see what I mean, check out the idealized schematic above. For reference, the vertical, dotted line marks the position of the gust front, so plots and observations to the right of this vertical line indicate conditions ahead of the gust front, while plots to the left represent conditions behind the gust front. Okay, note the spike in pressure on the barograph trace (second panel from the top). Along the gust front, the increase in surface pressure is primarily non-hydrostatic, owing largely to the low-level convergence of the warm and rain-cooled air streams (the mass convergence leads to relatively large upward accelerations, and thus puts the associated pressure increase mostly in the non-hydrostatic category). This non-hydrostatic effect helps to explain the increase in surface pressure that's observed just ahead of the approaching gust front. But the hydrostatic spike in pressure behind the outflow boundary drives the advance of the rain-cooled air (sometimes referred to as the cold pool).
To complete the conceptual model of an outflow boundary, note that the top panel (revisit the schematic above) shows streamlines that trace the downdraft of a thunderstorm and the corresponding outflow of rain-cooled air. The bottom two rows show the horizontal fields of wind, and temperature behind and ahead of a gust front (note the abrupt change in temperature across the gust front).
As a general rule, the depth of the rain-cooled outflow from thunderstorms averages between 500 meters and 1.5 kilometers, although I've seen observations confirming outflows as deep as two kilometers. By and large, relatively deep outflows provide greater uplift, all other factors being equal. Like synoptic-scale cold fronts, the lift produced by outflow boundaries can initiate new storms (see the idealized schematic below). Of course, the initiation of new storms along the gust front depends on the strength of the uplift, stability, and other local conditions that determine whether air parcels can reach the level of free convection (short primer).
An outflow boundary can initiate new convection, assuming that the the strength of the uplift, stability, and other local conditions, are favorable. Courtesy of Penn State's online certificate program in weather forecasting.
Uplift along an outflow boundary goes hand in hand with low-level convergence, whose strength depends, in part, on the difference in temperatures between the rain-cooled outflow and the ambient environment. For the record, temperature differences tend to be rather small (two to five degrees Celsius) in environments where relative humidity in the lower troposphere is already high. That's because there's limited evaporation (and evaporational cooling) that might otherwise make cold pools colder. Indeed, cold pools tend to be rather weak when the relative humidity in the lower troposphere is already high. In contrast, temperature differences between cold pools and the ambient environment can be as large as 10 to 20 degrees Celsius when low-level relative humidity is on the low side. In such environments, evaporational cooling proceeds like gangbusters and cold pools tend to be pretty strong.
In addition to the temperature contrast between the rain-cooled outflow and the ambient environment, the strength of the low-level convergence along an outflow boundary also depends on the movement of the warm air relative to the rain-cooled outflow. When the storm-relative wind has a component that blows from the warm air toward the cold pool, low-level convergence along the outflow boundary is enhanced. In other words, a storm-relative headwind promotes more low-level convergence (and more uplift). Note the storm-relative headwind on the idealized schematic above.
In the case of the American fire, however, outflow from showers and thunderstorms did not initiate any new storms. Unfortunately, the gust front that passed the Reno-Carson City area transported smoke from the American fire into the region, degrading air quality and posing serious health hazards.
P.S. The air quality in the Reno-Carson City area improved late yesterday.
Updated: 10:02 AM GMT on August 28, 2013
By: 24hourprof, 2:53 PM GMT on August 17, 2013
Typhoon Utor made landfall early Monday, August 12, 2013, on the northern Philippine Island of Luzon as a powerful Category-4 storm (maximum sustained winds of 120 knots). The storm could have been more powerful at landfall than it was because, on Sunday (August 11), Utor's maximum sustained winds were 130 knots (150 mph), qualifying the storm as a supertyphoon. At the time, Utor also had a dramatically small "pinhole" eye (see 06Z water-vapor satellite image below...talk about cold cloud tops contaminating water-vapor images!). As it turned out, Utor weakened slightly before making landfall, causing the storm's maximum sustained winds to decrease from 150 to 140 mph.
The 06Z water-vapor image on August 11, 2013, showing the pinhole eye of Supertyphoon Utor. Courtesy of the Dundee Satellite Receiving Station.
The slight decrease in Utor's intensity from Sunday to landfall on Monday (ten knots) is interesting from a meteorological standpoint, but, from a human perspective, I am keenly aware that the tragic loss of life, the widespread human suffering and misery, and the devastating loss of property, render any marginal decrease in Utor's intensity virtually meaningless. So I present the meteorology of Utor's weakening with the sense of an overarching sadness for the victims and the people of Luzon.
At 06Z on Sunday, August 11, Utor's pinhole eye could also be observed on standard visible and IR imagery (the data were collected by the geostationary satellite, MT-SAT2; both images are courtesy of the Dundee Satellite Receiving Station). There are other types of satellite imagery that tropical forecasters utilize to assess the structure of tropical cyclones. Indeed, satellite images based on the detection of microwave radiation at wavelengths between 85 and 91 gigahertz (GHz) are very useful (convert gigahertz to micrometers), especially when the eye of a hurricane, typhoon, severe tropical cyclone, or a very severe cyclonic storm is shrouded by high clouds (these descriptions are given to tropical cyclones in various ocean basins once sustained winds reach 64 knots). Of course, Utor's pinhole high was not obscured by high clouds on August 11, but microwave imagery will help us to understand why Utor weakened slightly from Sunday to the time of landfall (after landfall, the lack of access to warm water, and, to a lesser extent, the interaction with the terrain of Luzon, sealed the deal on the storm's weakening). I'll get into the basics of microwave satellite imagery and Utor's weakening before landfall in just a moment.
The 12Z microwave image of Typhoon Utor on August 12, 2013, showing a double eyewall structure of Utor (eyewall thunderstorms depicted in dark red). Larger image. Courtesy of The Naval Research Laboratory.
For now, focus your attention (above; larger image) on the 91-gigahertz (GHz) image of Utor at 1206Z on Sunday, August 11, 3013 (maximum sustained winds were still estimated at 150 mph at this time). For the record, the microwave data for this image were collected by the Special Sensor Microwave Imager/Sounder (SSMI/S) mounted on the F-18 satellite (a polar-orbiting vehicle in the Defense Meteorological Satellite Program). The two red "circles" indicate a double eye-wall structure at this time, an observation that's pivotal to recognizing that a concentric eyewall cycle was well underway (and thus understanding why Utor was not a supertyphoon when it struck Luzon). I'll properly define and discuss the concentric eyewall structure in just a moment.
The SSMI/S is a passive sensor, which collects naturally occurring microwave radiation emitted by the atmosphere and the earth's surface to space (an active sensor transmits pulses of energy and waits for the return signal). Of course, the earth and the atmosphere do not emit nearly enough microwave radiation to warm cold pizza, but the emissions are measurable (and, thus, can be detected in space by passive microwave sensors such as the SSMI/S). For starters, I emphasize that you're not really looking at cold, high cloud tops on this image. In other words, you can't interpret 85-92 GHz imagery the way you interpret enhanced infrared satellite imagery). The scientifically sound interpretation here is that we're looking at the 91-GHz brightness temperatures (in Kelvins) measured by the SSMI/S passive microwave sensor mounted on DMSP's F-18. Grenci, what are brightness temperatures, you ask? And how do we interpret them? Good questions!
The radiometry of microwave imagery (85-92 GHz). Larger image. Courtesy of, and copyright by, the Penn State Online Certificate Program.
Let's start with the schematic above (larger image). Note the relatively wide arrow representing 91-GHz radiation upwelling from the ocean surface (the signal is weak...otherwise, we'd all be cooked). This weak signal of upwelling microwave energy gets mostly absorbed or scattered away from the satellite's field of "vision" by raindrops and cloud droplets that reside below the freezing level in tall thunderstorms (for sake of my discussion, let's assume that the thunderstorms are located in the eyewall or a spiral rain band of a hurricane).
Raindrops and cloud droplets also emit 91-GHz radiation (some of the microwave energy gets emitted upward, of course). This upwelling 91-GHz radiation from the top of the "rain layer" gets scattered and absorbed above the freezing level by precipitation-sized ice particles such as hail and graupel. Higher up in the storm, however, tiny ice crystals in cirrus clouds are virtually transparent to 91-GHz radiation (small ice crystals do not attenuate the signal), allowing the remaining 91-gigahertz (microwave) energy to be transmitted to space (where, in this case, the SSMI/S detected the weak signal). In effect, the Special Sensor Microwave Imager/Sounder detects the "radiating temperature" of the top of the rain layer (more formally, brightness temperature, which I choose not to discuss any further for fear of unnecessarily complicating this presentation).
The microwave energy that gets transmitted to space is so weak, in fact, that the corresponding "radiating temperature" associated with the top of the rain layer in eyewall thunderstorms is typically very low. In more general terms, low "radiating temperatures" (at 85-92 GHz) are consistent with tall thunderstorms because hail and graupel higher up in the storm absorb a large fraction of the upwelling microwave radiation emitted from the top of the rain layer (revisit the vertical distribution of precipitation particles shown on the idealized schematic above). The bottom line here is that 85-92 GHz imagery helps forecasters to identify the structure of eyewall thunderstorms, no matter if the eye is shrouded by cirrus clouds or not. Microwave imagery also determines the structure of spiral-band thunderstorms.
One of the major limitations of 85-92 GHz imagery is that one of the several satellites equipped with a microwave sensor passes over a tropical cyclone, on average, every four to five hours (time lags as brief as 30 minutes or as protracted as 25 hours). So there can be long gaps between 85-92 GHz data for any tropical cyclone. Researchers at the University of Wisconsin devised a creative technique to fill in the time gaps with morphed 85-92 GHz images. Read more about Morphed Integrated Microwave Imagery (MIMIC). To access current and archived MIMIC data, visit the Web site for CIMSS, the Cooperative Institute for Meteorological Satellite Studies at the University of Wisconsin. You can choose animated gif or JAVA loops, which really give forecasters a better picture of changes in cyclone structure and intensity with time.
The morphed microwave (91 GHz) animation of Typhoon Utor from 00Z on August 11 to 00Z on August 12, 2013, as it approached Luzon and eventually made landfall. Dark red indicates tall thunderstorms (in the eyewall, etc.). Utor was a supertyphoon for a while...note the dramatically small eye early in the movie...but, later in the animation, Utor weakened slightly in response to an eyewall replacement cycle, which was well underway. The ERC was interrupted, however, as Utor made landfall along the Luzon Coast. JAVA link. Courtesy of the Cooperative Institute of Satellite Meteorological Studies.
Check out this MIMIC loop of Typhoon Utor (animated gif above; JAVA link for more control of speed and stopping/starting the movie) as it made landfall along the Luzon Coast (the movie spans from 00Z on August 11 to 00Z on August 12, 2013). Pretty cool, eh? First, note that the dark-red shadings translate to low "radiating temperatures" from the top of the rain layers. In other words, dark red marks eyewall thunderstorms (and tall storms in spiral bands). Early in the movie, note the pinhole eye and the correspondingly small, tight circle of eyewall thunderstorms associated with the supertyphoon status of Utor (sustained winds of 130 knots = 150 mph). As the movie advances, note the larger ring (larger diameter) of thunderstorms that begins to encircle the smaller circle of thunderstorms, indicating that a concentric eyewall cycle had begun. The concentric eyewall cycle is a process that commonly occurs in major hurricanes (Categories 3, 4 and 5) or, more generally, intense tropical cyclones (read more). By the way, I prefer eyewall replacement cycle (ERC) to concentric eyewall cycle, but I initially used the latter because "concentric" is somewhat more descriptive.
During an ERC over the open ocean, the outer eyewall will essentially cut off the inner eyewall's access to moisture. Moreover, the dynamically induced subsidence (compensating sinking air) inside the outer eyewall also promotes the demise of the inner eyewall. Any way you slice it, the inner eyewall collapses during an eyewall replacement cycle. As a result, hurricanes / typhoons weaken as the outer eyewall starts to encircle the inner eyewall. Not surprisingly, Utor weakened as the eyewall replacement cycle was well underway, almost completing by the time the storm made landfall.
After the outer eyewall takes over during an ERC over open seas, it can subsequently contract again (assuming the environment is favorable), and the storm can re-intensify (the radius of maximum winds decreases and there's a corresponding increase in maximum sustained winds). Of course, landfall along the Luzon Coast interrupted the eyewall replacement cycle, so Utor didn't have the opportunity to return to supertyphoon status.
Updated: 12:53 PM GMT on August 22, 2013
By: 24hourprof, 3:27 PM GMT on August 12, 2013
As a follow-up to my last blog regarding the hail-producing HP supercell in northwest Nebraska on August 3, 2013 (I recommend reading The Case of the Unwitnessed Hailstorm before you proceed), the storm that I presented on this 12Z mosaic of composite reflectivity (here's a closer look with counties) transitioned in a little over an hour to a bowing line segment of thunderstorms. For confirmation, check out, below, the 1330Z mosaic of composite reflectivity on August 3 (larger image).
The 1330Z mosaic of composite reflectivity on August 3, 2013. Larger image. Courtesy of Penn State.
Before you jump to any conclusions, supercells transitioning to bowing line segments is a fairly common evolution. And this kind of transition in storm mode (type) typically occurs when the outflow of cool air associated with the rear-flank downdraft (RFD) is both relatively strong and deep. In other words, such transitions occur when there's a strong RFD surge.
By way of review, there are two downdraft regions associated with the structure of supercells (see idealized schematic below). The rear-flank downdraft forms at the rear of the storm (obviously, not rocket science) in the vicinity of the rotating updraft (the mesocyclone) in response to drier air entraining (mixing) into the backside of the updraft. This entrainment of drier air promotes evaporative cooling and, in turn, increases negative buoyancy, which favors downward accelerations. Moreover, when hailstones and other icy hydrometeors (snow and graupel) get tossed into the rear-downdraft region, melting and sublimation of ice hydrometeors also help to cool the air and thereby increase negative buoyancy (melting of ice requires energy, which the surrounding air supplies).
An idealized schematic of a supercell thunderstorm. Courtesy of, and copyright by, the online Penn State Certificate Program.
Although I really can't say for sure what exactly caused the RFD surge on the backside of the HP supercell in northwestern Nebraska on August 3, I'm betting that the partial (or complete) melting of some hailstones probably played a role in enhancing the depth and strength of the rear-flank cold pool. The bottom line here is that the leading edge of the surging deep and strong RFD outflow initiated new thunderstorms, thus paving the way for the HP supercell to transition into a bowing line segment of thunderstorms early on August 3.
Below is the 1216Z image of storm-relative velocities (larger image; twin image of base reflectivity) I showed in my previous blog. This time I circled the footprint of the rear-flank downdraft...greens indicate negative (inbound) velocities. While speeds near 30 knots might not seem impressive, remember that the HP supercell was approximately 64 nautical miles from the radar site, placing the footprint of the RFD at roughly 6000 feet. Most cold pools have a depth of one kilometer, although cooler air flowing outward from thunderstorms have been as shallow as a few hundred meters and as thick as four kilometers (the depth of a cold pool largely depends on the longevity of the parent convection). Nonetheless, in the case of August 3, the footprint of the RFD at roughly 6000 feet (about two kilometers) was sufficiently large to convince me that rear-flank winds at lower attitudes were much stronger (keep in mind that the radar beam tilts upward and that the radar routinely misses strong low-level winds when the storm is relatively far from the radar site).
The 1216Z image of storm-relative velocities from the radar at North Platte, Nebraska (KLNX), on August 3, 2013. The white circle indicates the footprint of the rear-flank downdraft of the HP supercell that allegedly produced hail as large as four inches in diameter (or greater). Larger image. Courtesy of NOAA.
So, based on this image of storm-relative velocities, I believe there was an RFD surge associated with the HP supercell on August 3, 2013, and that the surging rear-flank gust front initiated new thunderstorms that paved the way for the supercell to transition into a bowing line segment of thunderstorms in a little over an hour.
Here endeth the lesson.
Updated: 10:31 AM GMT on August 16, 2013
By: 24hourprof, 1:57 PM GMT on August 09, 2013
On Saturday morning, August 3, 2013, a high-precipitation supercell (HP supercell; see 12Z mosaic of composite reflectivity below; larger image) developed near the Nebraska-South Dakota border, likely producing very large hail that did not appear on the display of SPC storm reports (even though hail might have exceeded four inches in diameter).
The 12Z mosaic of composite reflectivity on August 3, 2013. At the time, a mesoscale convective system was affecting parts of South Dakota and Nebraska, with a high-precipitation supercell producing large hail near the state border. Larger image. Courtesy of Penn State.
That's because the large hail was not observed in this sparsely populated area of the country. I'll have more to say about how I arrived at this estimate of hail size in just a moment, but I just have to get this off my chest...records for hail size, rainfall, snowfall, etc. are probably not really records in the grand scheme of weather. Extreme weather happens all the time, and, when it occurs, there's not always instruments or weather observers around to measure or document the "event."
Why did I refer to this storm as a high-precipitation supercell? Yes, the answer to this question is a bit of a digression, but I want all my readers "on the same page." There's a spectrum of supercells, ranging from a low-precipitation (LP) supercell, to a classic (CL) supercell, to a high-precipitation (HP) supercell. I don't want to get into all the gory details in this blog, but it stands to reason that it rains hardest in the general vicinity of the rotating updraft (mesoscyclone) inside an HP supercell (this assertion is not exactly rocket science, wouldn't you agree?). In turn, the more abundant rain associated with an HP supercell readily gets wrapped into the mesocyclone's circulation, tending to "camouflage" visual clues that a mesocyclone is present (one of the reasons forecasters look at images of storm-relative velocities). Without an obvious hook echo on reflectivity images (example of a supercell with a hook echo on radar reflectivity), the overall appearance of HP supercells often resemble kidney-beans.
The 12Z NAM model analysis of 850-mb streamlines on August 3, 2013, indicates upslope flow over parts of western Nebraska and western South Dakota. The upslope flow was part of the anticyclonic circulation associated with a high centered along the border between North Dakota and South Dakota (the blue "H" marks the center of an 850-mb high). Larger image. Courtesy of Penn State.
With my caveat about weather records and my digression about the spectrum of supercells out of the way, let's get back to the "unwitnessed hailstorm" on August 3. For starters, the supercell developed behind a cold front (12Z surface analysis) as winds around a ridge of high pressure forced relatively moist air up the sloping terrain. To see the upslope flow of air, check out the 12Z NAM model analyses of 850-mb streamlines (above; larger image) and compare it to the pocket of upward motion at 700 mb (below; larger image). Note that the pocket of upward motion at 700 mb at 12Z coincides with the mesoscale convective system (12Z mosaic of composite reflectivity) that included the supercell.
The 12Z NAM model analysis of 700-mb vertical velocities (in microbars per second) on August 3, 2013. Negative values indicate upward motion. Larger image. Courtesy of Penn State.
By the way, folks that insist on explaining the development of severe thunderstorms as a result of the "clashing of air masses" should really start to question the wisdom of this overly simple phrase...the supercell formed back in the cool air (far to the north of the stalled cold front; revisit the 12Z surface analysis; 12Z mosaic of composite reflectivity and the 12Z NAM model analysis of surface temperatures). And to readers who insist that all severe thunderstorms develop in concert with a 500-mb trough, I note that this HP supercell was initiated below a 500-mb ridge of high pressure (here's the NAM model analysis of 500-mb heights at 12Z on August 3, 2013). There, I feel much better now!
The 1216Z image of base reflectivity (Level III data) from the radar at North Platte, NE (KLNX...lower right) on August 3, 2013. At the time, an HP supercell was producing large hail. Larger image. Courtesy of NOAA.
To get a sense of the HP supercell's structure, check out, above (larger image), the 1215Z image of base reflectivity (Level III data) from the radar at North Platte, NE (KLNX). For reference, 1216Z is 8:16 A.M. CDT, and the storm was roughly 64 nautical miles from the radar site at this time. Yes, I can see the resemblance to a kidney bean. On closer inspection, note the values of 75-80 dBZ that presumably mark the hail core of the supercell. Here's the corresponding image of storm-relative velocities...I annotated the velocity couplet to confirm the storm's rotating updraft (the altitude of the couplet detected by the Doppler radar at North Platte was roughly 6000 feet at this time).
Okay, I'm getting close to where the rubber meets the road. There's no doubt that such high vales of reflectivity indicate hail (nothing new here). But how am I estimating the diameters of hailstones in excess of four inches? Good question! Below is the 1214Z analysis of hail size (larger image) derived from the National Severe Storms Laboratory's (NSSL)
Warning Decision Support System – Integrated Information (WDSS-II)).
The 1214Z WDSS-II estimates of hail size associated with the HP supercell in northwest Nebraska on August 3, 2013. At the time, WDSS-II indicated hail as large as four inches in diameter, perhaps larger. Larger image. Courtesy of NSSL via SPC.
In a nutshell, WDSS-II integrates data from a variety of sources...radars, satellites, models, observations, etc. in an effort to help weather forecasters analyze, diagnose, and predict severe weather (read more). For the specific case of large hail, algorithms estimate hail size, incorporating maximum vertical reflectivity (composite reflectivity), the height of the wet-bulb-zero (wet-bulb temperature), the height of the maximum reflectivity, vertically integrated liquid (VIL), and several other data sources. Perhaps I can write a series of short blogs on each topic in the future (example).
At any rate, the 1214Z WDSS-II analysis on August 3 indicated hailstones with diameters greater than or equal to four inches. Of course, we'll never know for sure because reliable reports are difficult to come up in such sparsely populated regions, but, had the supercell been near a more heavily populated area on the morning of August 3, reports of large hail would likely have been listed on SPC's Storm Reports Web page. Such cases are quite sobering because it emphasizes my point that some of the state, national and world weather records are likely not records at all. Indeed, many extreme weather events simply fall through the cracks in our observational network.
Updated: 11:54 PM GMT on August 11, 2013