Catatumbo Lightning in the Day/Night Band

You may have noticed that many of the recent posts have featured imagery from the VIIRS Day/Night Band (DNB). That’s because the nighttime imagery produced by the DNB is so awesome! The DNB has seen clouds at night, auroras, forest fires, oil and gas flares, and even volcanic eruptions. Many of the previous images shown have included high resolution views of city (and even small town) lights. This post shows another interesting facet of DNB imagery: lightning. More specifically, Catatumbo lightning.

For those of you who don’t know (and didn’t click on that last link), Catatumbo lightning is one of the world’s most frequent lightning displays, with thunderstorms forming over the Catatumbo River in Venezuela an average of 160 nights per year. The lightning displays last up to 9 hours, beginning shortly after dusk. The lightning is nearly continuous and so vivid and reliable that it has been called the “Lighthouse of Maracaibo” or the “Catatumbo Lighthouse”, as fisherman and sailors have historically used it as a navigation aid. It is said that the locals were saved from an invasion by Sir Francis Drake in 1595, as the invading navy could not covertly enter Lake Maracaibo at night due to all the bright lightning. There is even a campaign to make Catatumbo lightning a UNESCO world heritage site. The lightning is so prominent, the state of Zulia in Venezuela has included it in their flag and coat of arms. Two years ago, the storms suddenly stopped for several months, causing mass panic in the streets- I mean, on the river- I mean… um, actually the villagers in this video don’t seem to care all that much.

Earlier this month, when the moon was about 80% full, Suomi NPP passed over Lake Maracaibo at night and, sure enough, a thunderstorm was present right over the mouth of the Catatumbo River.

VIIRS I-05 image of thunderstorms near Lake Maracaibo, Venezuela taken 06:44 UTC 10 May 2012

VIIRS I-05 image of thunderstorms near Lake Maracaibo, Venezuela taken 06:44 UTC 10 May 2012

This image, taken from the high resolution imagery IR-window channel (I-05, 11.45 µm) on 10 May 2012, shows the deep convection over Venezuela and Colombia. The largest thunderstorm near the center of the image formed along the shore of Lake Maracaibo, near the mouth of the Catatumbo River. Here’s what the DNB saw at the same time:

VIIRS Day/Night Band image of thunderstorms near Lake Maracaibo, Venezuela taken 06:44 UTC 10 May 2012

VIIRS Day/Night Band image of thunderstorms near Lake Maracaibo, Venezuela taken 06:44 UTC 10 May 2012

The bright, almost rectangular streaks in the image are lightning strikes. The red arrow points out a lightning strike from the Catatumbo storm – a “Catatumbo lightning” strike, if you will.

The blocky appearance of lightning is due to the fact that VIIRS is a scanning radiometer. As the instrument scans the swath of the Earth that it sees, a bright, transient flash (such as from lightning) will show up in the along-scan direction as an individual streak of light in each sensor. The DNB has 16 different sensors that scan the swath simultaneously, and since lightning typically stretches over a large enough area to be detected by all of them, you get 16 different streaks all lined up next to each other. By the time the sensors have rotated back around for the next scan, the lightning flash has ended, producing abrupt edges in the direction along the satellite track. Compare this with the DMSP Operational Linescan System, which produces much more “streaky” lightning.

In addition to the “Catatumbo lightning”, you can see several other lightning flashes in the two deepest thunderstorms over Colombia. These are far enough away from Lake Maracaibo that they probably don’t count as Catatumbo lightning.

Other interesting features can be seen in these images as well. The moon was bright enough to cast shadows in the DNB image, allowing for the detection of the overshooting tops. These match-up with the coldest brightness temperatures in the I-05 image (which show up as dark blue to pure white in this color scale). A few pixels in the largest storm over Colombia (the one with two visible lightning flashes) have managed to make it to pure white on the color scale, indicating temperatures below 190 K (-83 °C). The dark blue pixels indicate brightness temperatures between 196 and 190 K (-77 to -83 °C). Brrr.

Overshooting tops exist when the convection is so vigorous, it peaks out above the anvil of the storm and penetrates the stable layer above (which is usually the stratosphere in storms this deep). In addition to acting as an indicator for severe weather, overshooting tops are important for energy and chemical transport between the troposphere and stratosphere.

It’s also interesting to see what looks like thin cirrus over the Caribbean Sea near Panama (left center of the image) that show up in the infrared (I-05) image, but not in the DNB. Plus, a number of cold clouds over Venezuela would appear to be optically thick due to their low brightness temperatures in the infrared image (yellow starts at 245 K down to green at 214 K), but they are optically thin enough to see city lights below in the DNB image. Awesome!

VIIRS view of Invest 97S at night

On 5 April 2012, the Joint Typhoon Warning Center was watching an area of the Mozambique Channel for possible development of a tropical cyclone. This area was named Invest 97S. As 6 April 2012 was a full moon, this is a good case to test the capabilities of low-light visible imagery channels for detection of tropical cyclone development at night.

The Operational Linescan System (OLS) aboard the Defense Meteorological Satellite Program (DMSP) satellite F-18 has a low-light visible channel (that inspired the development of the Day-Night Band (DNB) for VIIRS). The image below is from this channel on F-18, taken at 17:22 UTC, 5 April 2012 (courtesy the Naval Research Laboratory).

DMSP OLS low-light visible image of Invest 97S, taken at 17:22 UTC, 5 April 2012

DMSP OLS low-light visible image of Invest 97S, taken at 17:22 UTC, 5 April 2012. Image courtesy Naval Research Laboratory.

The landmass on the right of the image is Madagascar with Mozambique on the left side of the image. A low-level circulation is visible in the clouds just off the coast of Madagascar in the center of the image.

Suomi-NPP passed over the area at 23:02 UTC. The images below are taken from the VIIRS DNB, which is a low-light visible channel (centered at 0.7 µm) with higher radiometric resolution, a higher signal-to-noise ratio and higher spatial resolution. The second image is a zoomed-in version of the first.

VIIRS DNB image of Invest 97S taken at 23:02 UTC, 5 April 2012

VIIRS DNB image of Invest 97S taken at 23:02 UTC, 5 April 2012. Image courtesy Dan Lindsey and Steve Miller.

Zoomed-in image of Invest 97S from the VIIRS DNB taken at 23:02 UTC, 5 April 2012

Zoomed-in image of Invest 97S from the VIIRS DNB taken at 23:02 UTC, 5 April 2012. Image courtesy Dan Lindsey and Steve Miller.

In the nearly six hours that elapsed between the DMSP OLS image and the VIIRS DNB image, you can see that the line of deeper convection to the southwest of the circulation center has moved further south away from the center of the circulation and outflow from these storms has cleared out the low level clouds from where the storms used to be.

Compare these images with the high-resolution infrared window channel (11.45 µm), I-5, from VIIRS, seen below.

VIIRS channel I-5 image of Invest 97S, taken at 23:02 UTC, 5 April 2012

VIIRS channel I-5 image of Invest 97S, taken at 23:02 UTC, 5 April 2012.

The low level circulation is difficult to distinguish, given that there is no significant temperature contrast between the low level clouds and the background (ocean) surface. The deeper convective clouds are easy to spot in I-5, however.

The information provided by the VIIRS DNB near full moon events would be a great help to tropical cyclone forecasting in cases such as this where, typically, only IR data is available at night. Assuming latency issues with VIIRS can be solved, of course.

In the end, Invest 97S failed to develop into a tropical cyclone, which spared Madagascar and Mozambique – both of which had been affected by the cyclones Giovanna and Funso earlier this year.