LMA Data at WFO Huntsville Part II…New Day, New Utility…

So, when are total lightning data useful?  Well, there are many such cases, and we’ve described many of those in this blog…including my post yesterday.  But, they’re also particularly useful when the National Lightning Detection Network (NLDN) data drop out of AWIPS.  That is what happened earlier this evening, in an apparent system-wide outage that lasted for a couple of hours.  When you’re in operations, data redundancies are great to have, particularly for these types of cases.

At about 2247 UTC, a thunderstorm was moving across northern portions of Morgan County, AL, and was approaching southern Madison County (image 1).  Initially, the thunderstorm was expected to remain to the south and east of the Huntsville International Airport, for which we provide Airport Weather Warnings.

Image 1. KHTX radar reflectivity (dBZ, 2247 UTC), overlaid with North Alabama LMA source densities (2246 UTC).  The location of the Huntsville airport (KHSV) and a 5-mile radius ring (blue circle) from the KHSV location are also included.

However, by 2252 UTC, a small shower had developed along outflow to the northwest of the thunderstorm, and had just become electrically active (image 2).  Notice the (albeit small) LMA source density values just to the southwest of the airport radius ring at this time.  Of course, due to the NLDN data outage, we had no idea if the thunderstorm was producing CG strikes, but the total lightning data was sufficient for letting us know that the cell had become electrically active.

Image 2. KHTX radar reflectivity (dBZ, 2252 UTC ), overlaid with North Alabama LMA source densities (2252 UTC). The location of the Huntsville airport (KHSV) and a 5-mile radius ring (blue circle) from the KHSV location are also included.

Armed with this new information, the forecaster issued an Airport Weather Warning for the Huntsville Airport at 2252 UTC.  The next image (image 30) shows that the LMA indicated lightning (at least intra-cloud) within the 5-mile radius ring at 2314 UTC (image 3).

Image 3. KHTX radar reflectivity (dBZ, 2312 UTC), overlaid with North Alabama LMA source densities (2314 UTC). The location of the Huntsville airport (KHSV) and a 5-mile radius ring (blue circle) from the KHSV location are also included.

Knowing that intra-cloud flashes often precede CG strikes, and that thunderstorms in this type of environment will often eventually produce CG flashes, the total lightning data are invaluable, and can help to buy extra minutes of lead time in a rapidly evolving situation.  In this case however, we also see the added benefit when other data are simply unavailable.

LMA Data Back at WFO Huntsville and Quickly Useful for Airport Weather Warning…

The North Alabama Lightning Mapping Array (NALMA) data have been in and out of operations at WFO Huntsville, AL for a while now, due mainly to AWIPS II testing and related issues.   After being unavailable in operations for about a week, we were able to get the data back into operations on the afternoon of Thursday, June 13th…and it couldn’t have happened at a better time.  I was working the Aviation Forecast desk and was assisting in monitoring radar for severe weather operations and the data were of great benefit once again.  The first image below shows a small cluster of thunderstorm cells moving southward from Tennessee across the border into Lauderdale County, Alabama…the very northwest corner of the state, at about 2000 UTC, although the various data in the image range from 1955 to 2000 UTC.  NALMA data overlay the radar data and were being used to monitor for lightning activity in the cells.  Notice that at this time, NALMA data indicated the cell near St. Joseph, Tennessee was electrically active (white-pinkish shading).  Also, notice that the cell to the west and just north of Threet, Alabama was not electrically active yet, according to both the NALMA and NLDN data.

Image 1. KHTX radar data at 1955 UTC June 13, 2013…together with 15-min (2000 UTC) and 5-min (1955 UTC) NLDN data, NALMA source densities (1956 UTC), and METAR observations (~2000 UTC).

A little later, at about 2005 UTC, the cell had moved into Lauderdale County, now a few miles east of the town of Threet, and the NALMA indicated a sudden burst of electrical activity.  At this time, NLDN were not indicating any cloud-to-ground (CG) strikes.  Perhaps more importantly, this developing thunderstorm was moving towards the Muscle Shoals airport, which is located at the observation site (KMSL) in the northeastern section of Colbert County, directly to the south.

Image 2.  KHTX radar data at 2004 UTC June 13, 2013…together with 15-min (2000 UTC) and 5-min (2005 UTC) NLDN data, NALMA source densities (2004 UTC), and METAR observations (~2000 UTC).

The next image (Image 3), valid at about 2015 UTC shows the subsequent CG strikes in the NLDN data (horizontal blue lines).  Given the albeit small, but steady lightning production in this storm and increasing confidence that lightning was possible within 5 miles of the KMSL airport, a lightning warning was issued at 2015 UTC.

Image 3.  KHTX radar data at 2014 UTC June 13, 2013…together with 15-min and 5-min NLDN data (2015 UTC), NALMA source densities (2014 UTC), and METAR observations (~2000 UTC).

The next image shows a CG strike within 5 miles of the KMSL airport at 2025 UTC, as noted by the small blue horizontal line north of KMSL and east of Florence.

Image 4.  KHTX radar data at 2024 UTC June 13, 2013…together with 5-min NLDN data (2025 UTC), and METAR observations (~2000 UTC).

In this case, the LMA data alerted me that the cluster of cells had become electrically active, allowing me to shift my focus on when they might enter a 5-mile radius of the KMSL airport.   With a continuation of electrical activity as observed in the LMA data, my confidence was raised sufficiently, and the warning was issued for the airport.  The LMA data can be a great tool in situations like this, letting a forecaster know when a storm is electrically active and helping him/her to shift situational awareness appropriately, especially when CGs may not be initially present.  The LMA proved to be very beneficial in this case, allowing for some extra lead time with the airport weather warning for this cell.

Interestingly, the cell that entered eastern Lauderdale County was producing intra-cloud lightning throughout this time, but no CGs were reported by the NLDN.  Nevetheless, a forecaster would want some confirmation that a thunderstorm is in progress, particularly if he/she was involved in real-time weather watches for outdoor events, or for the possibility of lighting that could affect airport operations, as in the example above.

True Color Follow Up on New Mexico Fires

MODIS-VIIRS True Color 240pm June 11, 2013 and supporting ground truth photos.  We are anxiously awaiting more moonlight in the coming days to see how the VIIRS Day-Night Band products perform.

Viewing Fire Behavior Changes with VIIRS DNB Products

The VIIRS DNB Radiance and Radiance RGBs showed an increase in fire activity on the night following record high temperatures and unstable conditions over northern NM.  The Thompson Ridge Fire is nearly 22,000 acres, the Tres Lagunas Fire 10,000 acres, and the newly started Jaroso Fire is approximately 1,000 acres.  The DNB products showed the increased radiance between Monday, June 10th and Tuesday, June 11th, especially for the Tres Lagunas Fire.  The new Jaroso Fire to the north of Tres Lagunas is clearly visible on the nighttime product on the morning of the 11th.  Some cloud cover is also visible on the Radiance RGB to the north of the Thompson Ridge Fire on the 11th.

Nighttime Microphysics RGB Helps With Tennessee Valley Fog Coverage

It’s been a pretty foggy morning across the Huntsville metro area!  But if you’re looking for insight on the extent of the fog from early morning GOES visible imagery, good luck!  Mid-level clouds were obscuring lower-level features for a large portion of the region.

8 June 2013 – 1145 UTC – GOES-East Visible Imagery

Fortunately, the earlier VIIRS pass gave a much better perspective on the fog, despite being on the very edge of the early morning pass.

8 June 2013 – 0828 UTC – VIIRS RGB Nighttime Microphysics

Some nice details stand out in this image, particularly the multiple river valleys that can be picked out across north Alabama and middle Tennessee.  This implies that the fog may be a little more localized than previously thought, and agrees nicely with the rapid improvement of visibility that is being reported after sunrise.

ASTER Imagery for the Moore, Oklahoma Area

Earlier this week, SPoRT obtained cloud-free ASTER imagery over the Moore, OK area. The new imagery clearly shows the tornado damage path across the area. ASTER imagery also shows where the tornado path changes width and direction  as it tracked across the city. The ASTER damage track lines up with the storm survey performed by the NWS forecast office in Norman, Oklahoma. In this false color image the track is of the tornado is quite visible. The track in ASTER imagery begins on the west side of the Canadian River. It then tracks across the river and into the city of Moore.

ASTER False Color Reflectance Image taken on June 2, 2013

This second image is the Normalized Difference Vegetation Index, which emphasizes differences between vegetation and the urban area. The brighter areas of the image show where the vegetation is living and healthy, while the darker areas are a mix of water, urban areas and roads. The tornado track is quite visible in rural areas due to the vegetation being disturbed. The urban areas appear dark as well, with debris producing low NDVI values and a loss of detailed texture in road networks and neighborhoods evident in the non-damaged areas.

ASTER NDVI taken on June 2, 2013

The ASTER damage track can also be compared to the official damage survey and available radar data.  Below is the 1955z radar image and storm survey overlaid on the NDVI imagery. The tornado had been on ground for about 3 miles to this point and had only produced EF-0 intensity damage. The radar hook echo lines up with the track very nicely but NDVI has little to no change.

Zoomed in area just west of Moore with 1955z KTLX radar image overlaid. EF ratings are also outlined, EF-0 are in green, EF-2 in yellow and EF-4 in red.

As the tornado intensified, it is increasingly evident as a “debris ball” defined with higher radar reflectivity. The damage survey confirmed an EF-2 intensity as well as a small pocket of EF4 damage. The NDVI scar is much darker along the survey track where vegetation was damaged and adjacent areas were disrupted.

Once again the same NDVI frame as above but this time the KTLX radar image is 2003z from May 20, 2013.

Additional imagery can be found here on the SPoRT’s website.

SPoRT Images: Wildfire Decision Support Services and also Popular on Facebook

As New Mexico enters a third consecutive year of extreme drought conditions, the wildfire season has been active once again. The Tres Lagunas Fire started on May 30, 2013 and on the afternoon of the following day a second fire, the Thompson Ridge Fire, was started.  Both fires were ignited by downed power lines during strong wind events.

Satellite products received from SPoRT are very useful for graphical products that we share with partners and the general public.  In the early days of the event, it’s important that the public has a heightened awareness of the locations of the fires, associated smoke impacts, and the general character of the fire. Two graphicasts that were recently posted on our home page and Facebook page are included in this post. Note that the bottom right corner is left blank for the graphicast time stamp information. Another graphicast using the VIIRS Day-Night  Radiance product was previously shared on a separate blog entry.

In the first, we used the MODIS true color imagery to show the smoke plume associated with the Tres Lagunas Fire.  Areas along Interstate 25 were impacted by the smoke.

By the following day, a second fire was ignited and both remained active over the weekend.  The high resolution images included in SPoRT’s hybrid 3.9 μm “hot spot” loop not only provide an accurate location of the wildfire, but can illustrate differences in the character of the fire.  In the example below, a continuous hot spot is associated with the Thompson Ridge Fire.  However, the Tres Lagunas Fire had spread more rapidly, and had burned more acres at this time, and the hot spot imagery depicts a burn with an active perimeter. The 3.9μm image was combined with an awips topo map on awips to provide topographical references. Both of these graphicasts were also posted to Facebook, where they received higher than normal shares.