The use of LMA data during severe weather operations at the Huntsville, AL NWS office has a rather long history. Beginning in 2003, these data have been available to HUN forecasters in some form. Very recently, on the morning of March 2nd, 2012, these data again provided utility in helping to assess and anticipate the potential for severe weather. The image below (Figure 1) is a four panel AWIPS display of LMA data at 1442Z (upper left) and other various radar data from HTX (Hytop, AL) beginning at 1441Z March 2nd, 2012, which preceded the initial severe thunderstorm warning in our forecast area that day by about 10 minutes. For your reference, the panels of radar data are as follows…
upper right: (0.5 degree Reflectivity dBZ)
lower right: (0.5 degree Velocity)
lower left: (0.5 degree Correlation Coefficient)
Notice the LMA data (upper left panel) at the start of the event, with source density values generally around 100 or less at 1442Z. The associated storm and area of interest was in northern portions of Lawrence County and SW portions of Limestone County. Reflectivity and velocity data at this time show little storm organization. The next image (Figure 2) is a four panel a couple of mintues later (valid 1444Z for LMA and 1446Z for radar data). Here, the beginning stages of a “spike” in LMA can be seen with source density values up to about 200 at this time. The next image (Figure 3), however, really shows the spike in source density values, up to about 500, at 1446Z. Notice here that there was no change in radar data at the 0.5 degree level because in volume coverage pattern (VCP) 11,
it takes about 5-6 minutes for data between corresponding angles. This is one way in which the LMA data can come in handy. The LMA data are updated about every two minutes, importantly, filling in the gap between successive corresponding radar scans. While the LMA data don’t directly detect severe weather signals, they do indirectly indicate whether or not a storm is strengthening (weakening) due to increases (decreases) in updraft strength. During a situation when a warning forecaster is not ready to issue a severe thunderstorm or tornado warning, these data can help to indicate several mintues in advance that a storm is about to ramp up intensity. This can be especially useful in situations when the warning forecaster is about to the point of being “on the fence” with regards to the warning situation. Such was the case on the morning of Friday, March 2nd. Corresponding radar reflectivity data in the 3.4 degree slice (elevation in northern Lawrence and SW Limesone Counties) was used as one tool to assess the hail threat that morning. Figure 4 below shows the rapid increase in reflectivity values at this level (~20-25 kft) over the area of interest (click the image to bring up the loop…you may have to click it twice). However, dBZ values remained below 55 dBZ.
As the warning operator that morning, although overall storm structure didn’t necessarily indicate large hail or damaging winds, and we had no reports to that point, the LMA data indicated that this storm was on the verge of rapidly increasing in intensity. Due to the significant spike in LMA source density values, and with me nearly “riding the fence” with regards to a warning, I went ahead and issued the first severe thunderstorm warning for our area that morning at 1451Z (851 am CST). At 1505Z (9:05 am CST) we received the first report of quarter size (1 inch) hail in the town of Ripley in Limestone County. A little further downstream, another report was received from law enforcement at 1512Z (9:12 am CST) of golf ball size hail in the Athens community.
Of course, this was the storm that produced the first tornado in the Canebrake community just south of Athens that morning. Overall, the LMA data did show the very typical rapid increase and then decrease in values with the storm evolution, before the tornado. This aspect of the LMA source density data may be followed up in another post. But, I just wanted to show for now the usefulness of these data in warning situations. This was a particularly good case, in which the LMA data helped the warning forecaster to realize that a storm of interest was likely to undergo rapid strengthening, and that a warning was necessary. Figure 5 below is a loop of the storm of the first storm of interest that morning (click on the image to see the loop – you may have to click twice).