Meteo 101

Meteo 241

Meteo 361

Meteo 410


Page 1


                         THE FIRST LARGE SEVERE WEATHER EVENT OF 2006

                                                     by Philip Lutzak March 2006



After almost a month of no significant severe weather in February 2006, the atmosphere over the United States finally began its inevitable change towards spring-like conditions in the second week of March 2006. On Wednesday, March 8th, it was clear that this shift had arrived and that we would have a significant severe weather event on our hands in the southern and southeastern U.S. It lasted for over 24 hours and produced all of the classic severe weather types, including very large hail, strong, damaging winds and tornadoes.

Figure 1: GOES Visible Satellite image from 03/09/2006 1:45PM CST. Roll your mouse over the image to see the GOES IR4 satellite picture.

On the left is a GOES visible satellite picture  of the southeast U.S. taken at 1:45PM CST on the afternoon of March 9th, 2006 as the outbreak was approaching its peak. The brightest whites represent the areas where the clouds are thickest and the heaviest precipitation was falling. On the matching color-enhanced infrared satellite image from the same time (roll your mouse over the visible photo) notice that the corresponding clouds are colored in bright oranges and reds, denoting the highest cloud tops along and ahead of the cold front stretching from central Missouri down into Louisiana. This is where the tallest thunderstorms were, and there were rainfall rates as high as 2.5 inches per hour at that time and a tornado was spotted on the ground in west central Mississippi one minute after these images were taken (see item 3 in the storm reports from March 9, 2006.) From the afternoon of March 8th through the evening hours of March 9th there were numerous severe thunderstorm and tornado warnings in effect over this area and there was extensive damage from high winds and hail.


  In this discussion, I'll concentrate on the afternoon of March 9th in Mississippi and Alabama. I want to relate the development and progress of these storms to some of the important atmospheric conditions that we can track on the mesoscale level, and how these conditions help us forecast severe weather. Before I discuss the mesoscale conditions that occurred, here's a look at the overall synoptic atmospheric conditions that produced this event.


The Setup for Severe Weather. 

  The synoptic weather systems that triggered this particular event were pretty classic. Although a number of 500mb shortwave troughs, some quite strong, had consistently moved across the U.S. during February, none of the stronger troughs had moved far south enough to interact with any heat and moisture available at lower levels of the atmosphere in the southern U.S. states. But on March 6th, a deep 500mb trough moved in over the west coast, and then emerged into the southern plains states on Wednesday the 8th. This trough provided strong upper level support to a developing area of low pressure over Kansas and its attendant cold front and dryline. An increasingly strong flow of moist air from the Gulf of Mexico had developed ahead of the trough and front at the surface by the afternoon and evening of the 8th. You can easily pick out the dryline running through the middle of Oklahoma from the dewpoint map issued by the Oklahoma mesonet. Upper-air support was becoming stronger as a pool of cold air and vorticity max associated with a 500mb trough began moving in from the west. In addition, a strong jet streak in the southern branch of the jet stream was moving in from the  south and west, adding the major lift needed to make storms severe. On Wednesday morning of March 8th, the SPC Convective Outlook discussion was focused on it, and it became clear that some severe thunderstorms were inevitable. A severe thunderstorm and tornado watch for Texas and Oklahoma went into effect on Wednesday afternoon, and there were numerous severe thunderstorm reports by Wednesday evening. Overnight, the low pressure center and cold front pushed further eastward, with the low centered over northwest Arkansas and southeast Missouri by mid-day, and by Thursday afternoon, March 9th, the peak of the event occurred as numerous severe thunderstorms developed with the low and along the cold front, especially in the lower Mississippi Valley from southern Indiana and west Kentucky down into Mississippi and Alabama, producing extensive areas of damage.

  Because there was so much severe weather occurring in so many areas, I concentrated my discussion on the portion of the line in the deep south that ran from along the west side of Mississippi over into Alabama on the afternoon of March 9th; I did not address the conditions that occurred outside of these areas. Similar mesoscale weather features  existed in far southern Indiana and Illinois, western Kentucky and parts of Louisiana and Arkansas. I would also like to point out that, as you read these charts below, your eyes will quite likely be drawn to the very high values of warmth, moisture and instability in southeastern Louisiana. It might be natural to wonder why this area was not included in the discussion. But this helps point out the need to understand the causes of severe weather at all levels, i.e. at the synoptic as well as the mesoscale. The synoptic level forcing discussed above tells us that southeastern Louisiana simply didn't have the upper-level support (i.e. cold upper level trough and strong divergence from the jet stream or streaks) needed to ignite severe thunderstorms that was so influential a few hundred miles north of it. 

Most of the examples I used were from 18Z thru 21Z, or as close to that time as possible, when the severe thunderstorms along the cold front were at their height in Mississippi and Alabama. The SPC radar reflectivity mosaic in figure 2 above shows the intense thunderstorms that were occurring at this time. The damage was almost entirely from straight-line winds along the intense squall lines, consistent with strong downdrafts.


Figure 2: SPC radar mosaic 03/09/2006 at 1825 UTC or 12:25PM CST.



Unstable Air - Deep Moisture Convergence and ThetaE


Deep Moisture Convergence

  When deep moisture is available at lower levels we know that the "fuel for the fire" is being transported into the environment. Below in figure 3 is the "deep layer moisture flux convergence" chart from SPC from the afternoon of March 9th, showing the high amount of moist, unstable air from the surface to 2 kilometers  being brought into the lower Mississippi Valley that day. The chart of surface conditions at 1800 (figure 4) shows the strong surface moisture convergence followed by the chart of 850mb heights, winds and dewpoints (figure 5), showing the low-level jet supplying large amounts of moisture into the area at 5,000 ft. 


Figure 3: This deep layer moisture convergence from the SPC shows the available moisture coming in from the surface to about 6 or 7,000 feet. 03/09/2006 at 1:00PM CST. Courtesy SPC.


Figure 4: Surface conditions with dewpoints. Notice the very strong southerly flow coming from the Gulf of Mexico into Louisiana and Alabama, and dewpoints ahead of the front in the upper 50s and lower 60s. 03/09/2006 at 1:08PM CST. Courtesy RAP.


Figure 5: 850mb heights with wind, temperatures and dewpoints show the deep moisture being transported northward into the lower Mississippi Valley at about 5,000 feet by the low-level jet. 03/09/2006 at 12:00PM CST. Courtesy SPC.




 While deep moisture shows us the volume and depth of moisture available to the storms at hand, thetaE is a valuable tool in quantifying exactly how warm and moist the air is, i.e. how volatile is the fuel being provided for these thunderstorms. ThetaE corresponds to the value of equivalent potential temperature in the environment, which effectively allows us to use one value to assess the temperature and dewpoint together. Where these values are increasing, so is the potential instability of the air. So here we'll take a look at the thetaE of the air that day along with other values that normally show the potential for instability - in this case dewpoints and lapse rates.

  Figure 6 from SPC shows the actual thetaE values for the afternoon of the 9th, and the advection or advancement of this unstable air. We can see values up to 330K up into south central Mississippi and strong advection of thetaE further north into the Mississippi River Valley and Alabama at this time. The high surface dewpoints (60-68 degrees F) in figure 7 match up quite well with the values in figure 6.


Figure 6: Surface values of thetaE and thetaE advection on 03/09/2006 at 1PM CST. Green lines are the actual thetaE values, purple lines are the advection, or amount of advancement, of the thetaE. Courtesy SPC.



Figure 7: Surface dewpoints at 03/09/2006 20:00 or 2:00PM CST. The color key is on the left. CompMap Courtesy SPC.


Wind Shear, Lapse Rates, and Warm Air Advection

  There are three other important variables that we can use from our tools to predict severe weather: wind shear, lapse rates and warm air advection. For this discussion of wind shear we can concentrate on vertical wind shear, the change in direction of wind speed or direction with height. It is profoundly important in determining whether severe thunderstorms will achieve enough spin to develop a mesocyclone and become supercells. On the right of the skewT diagrams in figures 8 and 9 below, we can see the winds turning with height: from southeast to southwest at Birmingham, AL and south to southwest at Jackson, MS. Notice on all of these that the wind is increasing its speed with height as well. This directional and speed shear is a good indicator that supercells can develop, since the other factors that we are discussing here already show that the air is unstable. This skewT forecast for Huntsville, AL for early evening on the 9th (when the front actually arrived) shows the same vertical wind shear pattern. Lastly, look at how well we can see the shear increasing over time on this profiler data from Okolona, Mississippi, as we go from 3AM to 2PM on the 9th. Note the winds turn from southwest to southeast at the surface in time, and also notice how much increase in wind speed shear develops that day.

   Also of great importance in figures 8 and 9 are the lapse rates (the red (temperature) and green (dewpoint) lines moving from bottom to top on the graph). These lines show the actual rates of decrease of temperature and moisture as we go up in altitude. When the rates of decrease in temperature are very close to the moist adiabats (dashed green lines) as they are here, we know that the air may be truly unstable enough for severe weather, if the capping is breached and the air becomes moist enough at those levels. These lapse rates are used in calculating CAPE, which is covered in the next section.

  The last of our list of important tools in this section are hodographs. They are the concentric circles on the upper left of the skewT diagrams below. Hodographs are also great indicators of what the wind is up to in the upper atmosphere. The lines that snake out along the circles of the hodograph are lines of horizontal wind change with height or wind vectors. We can, for this exercise, use them to tell whether the wind is veering or backing as we go higher up in the atmosphere. The winds on both of these hodographs, plus this one from Birmingham, AL on March 9th at 7AM CST, show the wind is veering (changing in a clockwise fashion) as we go up in altitude in the lower troposphere. A veering wind at lower levels of the atmosphere indicates warm air advection is occurring, bringing in more instability below the colder air at higher levels.

Figure 8: SkewT upper-air sounding from Jackson, MS. 03/09/2006 at 7:00AM CST. Courtesy RAP/UCAR.


Figure 9: SkewT upper-air sounding from Birmingham, AL. 03/09/2006 1800 or 12:00PM CST. Courtesy RAP/UCAR.


Potential for Instability - Mean Layer CAPE  and the Lifted Index


Mean Layer CAPE

  The mean layer CAPE values at 1900 or 1PM CST on March 9th during this event were representative of the values for the entire afternoon, with the caveat that the highest values gradually shifted eastward and died down by evening. Figure 10 shows the values from that hour. Mlcape is quite reliable as an indicator of instability when the lowest layers of the boundary layer are well-mixed, and we can see from the prior skewTs from Jackson, MS and Birmingham, AL (figures 8 and 9 above) that the layer is well-mixed up to about 820mb, so we can use 100mb mlCAPE as a valid guide here.

  The mean layer CAPE values shown here are within the range of marginally unstable (0-1000J/kg), which effectively means that very strong thunderstorms, and supercells, can occur here. And we know that they did occur, with a very large area of wind damage along the path of the cold front. With high thetaE and deep moisture convergence values, and CAPE values like these at the same time, it was a pretty sure bet that the strong cold front with upper level support and a 300mb jet streak coming in overhead (see synoptic section above) would be more than enough to cause severe thunderstorms.  

   Notice that the CIN (Convective Inhibition) values are fairly low here along the Mississippi River Valley, and thus the air offered little resistance to the approaching cold front with its strong dynamic forcing. However, it's also interesting to note that some of the most severe cells that afternoon were clustered in northwestern Alabama (see storm reports for 03/09/2006 in figure 14 below), where there is higher CIN, not long after this 1900 chart was compiled, and it may have been that the CIN there was strong enough to allow for some more impressive resistance to the upward forcing of the approaching cold front, and thus contributed to the especially strong thunderstorms in that area. We can also see this capping on the skewT from Birmingham (figure 9 above) from 820-790mb, with a small isothermal profile of the environmental temperature from 640-600mb. In addition, it's interesting to note the large wedge of dry air from about 800-380mb. This probably enhanced the damaging straight-line winds in the area, as the incoming precipitation got evaporatively cooled in this wedge, giving the downdrafts increasingly negative buoyancy and thus accelerating their downward momentum.   


Figure 10: 100mb mean layer CAPE and CIN on 03/09/2006 at 1PM CST. Red lines are the actual mlCAPE values, blue lines are CIN values. Courtesy SPC


Lifted Index and CINH

  Lifted Index (LI) is a measure of real buoyancy at 500mb, rather than potential buoyancy. But it is also a conditional indicator of instability, meaning that it does not guarantee the eruption of thunderstorms (severe or not) on its own. LI is rather a valuable indicator of how unstable the atmosphere is, and therefore how severe the thunderstorms may be, IF storms erupt. The highest values I saw during the outbreak in the Mississippi Valley area were from -1 to -3, in the marginal to probable category (see figure 11 below), and, once again, with all of the other indices at marginal to high values, the LI was another clue that these storms would become severe.

  The CINH, another measure of convective inhibition, but derived using the same parcel as is used in the LI calculation (i.e. lifted to 500mb), is insignificant in the vicinity of the advancing front and squall line in western Mississippi at 1900 UTC. See the end of the prior section on mean layer CAPE regarding the higher values in western Alabama.

Figure 11: Surface-based Lifted Index and surface-based CIN on 03/09/2006 19:00 or 1PM CST. Courtesy SPC.



Doppler Radar and Severe Weather

 The following radar page illustrates some Doppler radar images taken during the event in order to show some of the radar features that were frequently observed during this phase of the severe weather outbreak, and how some of the severe weather warnings came to be issued.




  We can see from our examination of this outbreak that we had all of the classic ingredients required for severe weather. Strong thetaE advection at the surface and deep moisture convergence at low-levels allowed for plenty of warm, moist air to converge ahead of the cold front. The air was quite unstable, as witnessed by the high values of mlCAPE due to unstable lapse rates and Lifted Index. Any capping that existed due to inversions was small in most places, and where it was larger was still no match for the powerful forcing caused by the cold front. The strong upper level 500mb trough along with the 300mb jet streak provided tremendous upper level support to the surface low and cold front, in effect providing the match to ignite the volatile air along the front.    

  And finally, here's an excerpt from the Convective Outlook issued by the SPC at 10:35AM CST on March 9th:

Their assessment was correct. The upper level jet stream energy shifted more northward along with 
the 500mb upper level trough, while the cold front continued eastward during the late evening hours and
overnight without its upper level support, so that it gradually weakened into a more typical late winter system
with heavy rain showers and weaker, smaller squall lines with embedded thunderstorms.  

  The chart below shows the storm reports for the first of the two days of the event, March 8, 2006, compiled by the Storm Prediction Center.  To see the March 9th reports, roll your mouse over the image.


              Figure 12: Roll your mouse over the image to see March 9th storm reports.




Reflections on Writing Project 2





1) For an interesting treatise on the basic requirements for the development of squall lines and bow echoes, see this detailed primer on squall lines from the NWS forecast office in Louisville, Kentucky.

2) On damaging winds in thunderstorms: Damaging Winds from SPC's Jeff Evans