My RaDAR Paper for SatRad GR6753

A RaDAR Analysis of the 2007 Groundhog Day Tornado Outbreak, Central Florida

 

Introduction

          In the early morning hours of 2 February, 2007, ground temperatures were already rising to levels well above average; coupled with massive amounts of moisture and a powerful mid-level jet stream overhead, synoptic conditions were favourable for severe weather.  The instability and wind shear led to a system of squall lines and supercells across Central Florida which would produce multiple tornadoes, killing 21 people and causing over $200 million in damage.  One of the most dangerous aspects of these storms is that they began at night locally.

          Following is a description of multiple mesoscale events affecting Florida that day.  I describe these events using BR, CR, BV, SRV, VIL, and Tops products provided by NOAA’s HDSS Access System and displayed in GRLevel3.  When viewing these images, please note I use mostly 4-panel windows, with images designated in order from top right to top left, bottom left, and bottom right, the same as a mathematic Cartesian plane.  Also note, there are so many RaDAR signatures associated with this event, however only a few important aspects will be discussed.

RaDAR Analysis

          Below in ‘Img01’, an image of development in the proximity of KJAX Jacksonville, FL at 0224Z, which given the time of year and time difference makes it 2124L (Local).  The area of concern is the area approximately 60nm from the RaDAR unit, or halfway out, in the 3^rd range ring.

Img01; BR0.5, BR1.5, BR2.5, CR; 0224Z

 

  

 

The majority of the reflectivity returns for the southern storm are in BR0.5 and 1.5 tilt angles, which at 60nm means they are at around 13,000 ft.  At a closer look in ‘Img02’, there is also a tight reflectivity gradient with what appears to be a BWER and inflow notch on the southern edge of this storm.  I would be watching for development within this storm.  This is shown by the reflectivity returns as well as relatively high inbound BV returns.  With the Mesocyclone algorithm marks shown for reference, the SRV0.5 product shows what appears to be cyclonic rotation on the WNW side of the storm, and also on the ESE side.  Oddly enough, just to the left of the mesocyclone marks, you can see decreased inbound velocities.  Also, in the reflectivity you see a WEC to the WNW of the main storm, which appears to be causing the entire system to rotate cyclonically.  So basically we have a rotating system with relative rotation within the system itself.  I would be very wary of this storm.

          ‘Img03’ shows what appears to be an MARC at 0319Z.  The BV and SRV both show high velocity differences over a relatively short distance, and given their tilt angles and proximity to the RaDAR unit,  puts this feature at approximately 10,000ft., which is consistent with MARC locations.  At this stage I would be issuing severe storm warnings for Saint Augustine and the surrounding areas due to the threat of powerful winds and downbursting winds at the surface, as well as the potential for hail, given the presence of an MARC signature.

Img02; BR0.5, BV0.5, SRV1.5, CR; 0224Z

  

Img03; BV0.5, BV0.5, SRV0.5, CR; 0319Z

 

Shifting now to KMLB in Melbourne, FL a few hours later, ‘Img04’ shows a WER coming in from the Gulf of Mexico onto the W coast of FL at 0725Z.  There is a tight leading edge reflectivity gradient and a vertical tilt as is evident by the BR1.5, VIL, and Tops products.  Upon seeing this come into viewing range, I would immediately issue a severe thunderstorm warning for Citrus and Sumter Counties, FL due to the potential for flash flooding, hail, downbursts and straight line winds, and possible future BWER development and tornadic formation.

Img04; BR0.5, BR1.5, VIL, Tops; 0725Z

Unfortunately, this storm does indeed intensify rapidly, as is shown by using the Lemon Technique in’Img05’ at 0803Z.  Looking at several levels of the BR product for the same storm you can see its layers vertically, which give a much clearer indication of the storm’s vertical structure, indicating such things as tilt, core location, and rotation.  It is now what appears to be a HP supercell.  You can see by the VIL  that the amount of precipitation aloft has increased dramatically.  There is also a slight vertical tilt, and an apparent hail core aloft.  This is shown by the BR2.5 product, and given its proximity to the RaDAR, approx.. 85nm, puts this hail core at between 25,000-30,000ft aloft.  At this stage, I would be issuing a severe thunderstorm warning for Sumter, Lake, and Marion Counties, FL, citing flash flooding, strong surface winds, and hail all as serious threats.  Additionally, due to the fact that the precipitation at the lowest tilt seems to be migrating slightly to the SW of the precipitation aloft, I would anticipate the imminent development of a BWER, subsequently erring on the side of caution and issuing a tornado watch for the same counties.

Img05; BR0.5, BR1.5, BR2.5, VIL; 0803Z

 

In ‘Img06’, this storm has now become a well-defined supercell at 0811Z, exhibiting all of the signs, the tight reflectivity gradient on the inflow, or southern side of the storm, a BWER alongside the inflow notch, a hook echo, a slight velocity couplet shown in the SRV product, and even a v-notch formed from the shear striking the updraft and fanning out towards the NE.  You can also make out the FFD and RFD.  At this stage I am upgrading to a tornado warning for Sumter, Lake, Marion, and Volusia Counties, FL.  I would expect flash flooding, hail, powerful winds, and a tornado to either form or have already formed with this storm.

Img06; BR0.5, BR1.5, SRV0.5, CR; 0811Z

Img07; BR0.5, BV1.5, SRV1.5, CR; 0849Z

 

In ‘Img07’, you can now see it is very evident there is a tornado present due to the BWER, hooked and tightly-packed gradient nature to the reflectivities, and rotation couplets in both the BV and SRV products as shown by the localized area of inbound/outbound returns shown by the green and red returns, respectively.  I would have already issued an extended tornado warning for Volusia Co., FL, as well as Sanford Co, FL due to the risk of tornadic formation due to spinoff from the flanking line in the RFD of the supercell.

          Switching now to an event later that day, at around 1630Z, or 1130L.  ‘Img08’ shows a storm which appears to be a localized area of rotation, however it seems to have a lack of moisture.  If this were not Florida, I would go so far as to call this system a LP Supercell.  If you look closely, you can make out a slight BWER in the BR product.  You can also notice the slight relative rotation in the SRV product.  I would watch for future development with this storm, issuing a severe thunderstorm warning for Okeechobee Co., FL.

Img08; BR0.5, SRV1.5, BV0.5, CR; 1631Z

 

In ‘Img09’, you can now see a well-defined BWER on the S edge of the storm.  The rotation is more pronounced in the SRV and BV products as well.  The CR product shows that there is yet much more precipitation aloft.  I have also included an image, ‘Img09a’, of the BR product at 0.5, 1.5, and 2.5 tilt angles, as well as VIL to show the lack of substantial moisture in the vertical column above this storm as well as the SE tilt of this storm.  At this stage I would issue a severe thunderstorm warning and tornado watch for Okeechobee, Indian River, and St. Lucie Counties, FL.



Img09; BR0.5, SRV1.5, BV0.5, CR; 1700Z

In ‘Img10’, at 1738Z, you can see that the storm has hit the ocean, and gaining a water source, gained low-level moisture, caused further convection linearly along the N, and now has a well-defined area of inflow from the rear as shown in both SRV and BV products.  The storms died off soon after this, with the passage of a cold front.

Img09a; BR0.5, BR1.5, BR2.5, VIL; 1700Z

Img10; BR0.5, SRV1.5, BV0.5, CR; 1738Z

.BT