Analyzing MODIS Imagery of North Alabama Tornado Tracks

We could analyze and discuss the meteorological, sociological, and personal impacts of the April 27, 2011 tornado outbreak for years to come.  It’s certainly an event that many Alabamians will never forget.  The outbreak brought the first EF-5 tornado to Alabama since the April 8, 1998 Oak Grove tornado in Birmingham, and the first EF-5 tornadoes in north Alabama since the April 3, 1974 “Super Oubreak”.  The tornado tracked for more than 130 miles, touching down in Marion County, Alabama southwest of Hackleburg, and remaining on the ground (with very few brief skips) until lifting near Huntland, Tennessee.

The tornadoes brought power and connectivity problems to the Huntsville NWS office and to NASA SPoRT.  So until recently, we have been unable to view quality MODIS imagery, and then clouds obscured the damage paths over the weekend and early this week.  This morning’s 1640 UTC pass marked the first time we were able to view the damage paths in AWIPS.  The 500-meter color composite from AWIPS is included below.

500m MODIS Color Composite (5-11-11 1640 UTC)

500m MODIS Color Composite -- 5-11-11 1640 UTC (Click to enlarge)

The long EF-5 path is really only readily apparent for the first 60 miles or so, when it was tracking across more heavily-forested portions of Marion, Franklin, and Lawrence Counties in Alabama.  The track fades as the tornado moved into the red clay farmland surrounding the Tennessee River (less of a change in reflectance), and is only briefly evident as it reemerges in northwestern Madison County, Alabama (close to where many NWS and SPoRT staff members live).  Even that portion of the track is obscured slightly due to a few pesky cumulus.

Other tornado paths are more evident in the 250-meter visible image, including all three tornado paths in Marion County, Alabama, and the last 15-20 miles of the EF-4 that moved across Cullman and Marshall Counties.  All of these paths are noted with arrows on the image below.

250m MODIS Visible Image - 5-11-11 1640 UTC

250m MODIS Visible Image - 5-11-11 1640 UTC (Click to enlarge)

Let’s focus again back on the EF-5 tornado.  One of the advantages of MODIS imagery is that it can help target storm surveys–or verify what has been found.  Zooming in on the widest part of the EF-5 track yields an image like this:

250m MODIS Visible Image (Marion, Franklin, and Lawrence Counties) - 5-11-11 1640 UTC

250m MODIS Visible Image (Marion, Franklin, and Lawrence Counties) - 5-11-11 1640 UTC (Click to Enlarge)

At its widest in eastern Franklin County, the path is approximately 8 pixels wide, and 4 to 6 pixels wide closer to the Marion-Franklin County line (near where Hackleburg and Phil Campbell are located).  This suggests a width of 1-1.5 kilometers (2/3-1 mile) in these areas, and 2km (almost 1.25 miles) wide in eastern Franklin County.  This closely matches what the NWS storm survey teams found (myself included): original estimates of the path width in Phil Campbell the day after the outbreak were 3/4 mile, and a follow-up survey yielded a 1.25 mile wide swath in the Oak Grove community of eastern Franklin County.

SPoRT continues to work to acquire and process even-higher resolution imagery from the ASTER (Advanced Spaceborne Thermal Emission and Reflection Radiometer) instrument aboard Terra.  Hopefully this imagery will yield even more information, which can be used for future research…and hopefully new ways to limit such killer outbreaks.

One thought on “Analyzing MODIS Imagery of North Alabama Tornado Tracks

  1. Pingback: MODIS Detection of Tornado Paths, One Year Later « The Wide World of SPoRT

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s