Meteorologists Monitor Meteor

September 29th, 2021 |

According to the JPL site, there was a bright meteor (or bolide) on September 29, 2021 over the Gulf of Alaska. (The JPL and a similar NASA site are posted under the GLM tab on this link of links.) This event was seen by both the ABI and GLM on NOAA‘s GOES-17, as well as the AHI on Japan’s Himawari-8. What may be unique about his event is that the imagers monitored the meteor soon after it’s explosion, and not just the resulting plume (as was done in this case over Russia in 2013). This is based on the length of the event, during which the various spectral bands displayed a signature and other information.

Peak Brightness DatePeak Brightness Time (UT)Latitude (deg.)Longitude (deg.)Altitude (km)Total Radiated Energy (J)Calculated Total Impact Energy (kt)
2021-09-29 10:50:5953.9N148.0W2813.7e100.4

Entry from table via the JPL site.


The GLM and ABI observed this event, but given it’s faster readout, the GLM offers much more information than the ABI. The apparent location of the meteor as seen by the ABI is different than the reported location, in part due to parallax. More on the concept of parallax is available here.

Animation of GOES-17 ABI band 12 (9.6 mirometer) mesoscale sector #2 on September 29, 2021.

Hotter brightness temperatures can be seen in the GOES-17 ABI band 12 at 10:50:59 UTC.

Animation of all 16 bands of the GOES-17 imager on September 29, 2021. Note band 12.

Indicative of a short duration event, coupled with how the ABI scans, the meteor signature was only clearly seen at one time in nearly every ABI spectral band (although possibly the ABI band 11 as well). Due to the layout of the focal plane array on the ABI, not all spectral bands observe the Earth at the precisely same time. [Figure a modification from the GOES-R Series Data Book.] A similar loop as above, but as an animated gif, is available here. In addition,. while a bit hard to see, the longwave split window infrared difference also showed a subtle signature of the meteor.

Spectral difference images (over time) can also be useful in the monitoring of meteors. An ABI 10.3 – 12.3 micrometer band difference is shown below. An shortwave minus longwave difference loop.

An animation of the GOES-17 difference image between ABI 10.3 – 12.3 micrometer bands. The brightness temperature range is -5 to +5K.

The GLM on GOES-17 also observed this event. A similar loop as below, but as an animated gif, is available.

ABI band 12 and the GLM Flash Event Group density on September 29, 2021. Credit: CIRA/RAMMB Slider.

The rapid movement of the meteor to the south is clearly evident. As well as the GLM group map and the key (blue is early times and red is later times).

GOES-17 GLM meteor location over time and space on September 29, 2021 with larger circles (color coded to intensity). Credit: Todd Beltracchi.

As well as the changes over time, most likely monitoring the meteor break-ups.

GOES-17 GLM meteor over time on September 29, 2021. Credit: Todd Beltracchi.

More on the GLM’s light curves from NASA AMES.


Both the ABI and Japan’s AHI scan space around the edge of the Earth. However, with the ABI data the process of making calibrated, navigated, and remapped radiance only pixels located on the Earth are included in the Level 1b radiance files. Hence, the ABI may scan meteors in space, but the data are not available to most users.

All 16 spectral bands from Himawari-8 AHI at the same nominal time (10:50 UTC) on September 29, 2021.

A similar loop as above, but as an animated gif, is available here (and an 8-panel AHI image at this same time is available here). This example helps to illustrate that each AHI detector doesn’t sense radiation from the same exact location at the same time.


NOAA GOES17 data were accessed via the University of Wisconsin-Madison SSEC Satellite Data Services. McIDAS-X and Geo2Grid was used to generate imagery. Thanks also to Todd Beltracchi and Scott Bachmeier, and to CIRA/RAMMB Slider images/movies.

Hurricane Sam reaches Category 4 intensity

September 25th, 2021 |

GOES-16 “Red” Visible (0.64 µm) and “Clean” Infrared Window (10.35 µm) images [click to play animation | MP4]

1–minute Mesoscale Domain Sector GOES-16 (GOES-East) “Red” Visible (0.64 µm) and “Clean” Infrared Window (10.35 µm) images (above) showed Hurricane Sam as it intensified from a Category 3 to a Category 4 storm (ADT | SATCON) in the central Atlantic Ocean on 25 September 2021. The eye became cloud-filled during the middle portion of the day, but Visible images revealed the presence of mesovortices within the eye both early and late in the day.

A DMSP-16 SSMIS Microwave (85 GHz) image at 1918 UTC from the CIMSS Tropical Cyclones site (below) displayed a fully closed eyewall, with several spiral bands wrapping inward toward the storm center.

DMSP-16 SSMIS Microwave (85 GHz) image at 1918 UTC [click to enlarge]

GOES-16 Infrared images with an overlay of deep-layer wind shear at 2200 UTC (below) indicated that Sam was in an environment of low shear — which favored intensification as the hurricane moved across relatively warm water (SST | OHC).

GOES-16 Infrared images, with an overlay of deep-layer wind shear at 2200 UTC [click to enlarge]

During the following nighttime hours, ample illumination from the Moon — which was in the Waning Gibbous phase, at 81% of Full — provided a “visible image at night” using the Suomi NPP VIIRS Day/Night Band (0.7 µm) (below).

Suomi NPP VIIRS Day/Night Band (0.7 µm) image [click to enlarge]

===== 26 September Update =====

GOES-16 “Red” Visible (0.64 µm) and “Clean” Infrared Window (10.35 µm) images [click to play animation | MP4

On the following day, Sam exhibited a similar appearance on 1-minute GOES-16 Infrared and Visible images (above), with a small 7-15 mile diameter eye (containing mesovortices, as seen in Visible imagery). Both Infrared and Visible images revealed repeated pulses of gravity waves propagating away from the storm center. Sam’s intensity peaked at 135 knots late in the day (NHC advisory).

1-minute GOES-16 Visible images with plots of corresponding GLM Flashes (below) showed that Sam exhibited an Enveloped Eyewall Lightning signature (reference).

1-minute GOES-16 “Red” Visible (0.64 µm) images, with 1-minute GLM Flashes plotted in red [click to play animation | MP4]

===== 30 September Update =====

GOES-16 “Red” Visible (0.64 µm) images (credit: Tim Schmit, NOAA/NESDIS/ASPB) [click to play animation | MP4]

Sam was still a Category 4 hurricane on 30 September — and GOES-16 Visible images (above) showed that Sam exhibited a relatively compact eye, with hints of mesovortices within the eye.      

SIFT investigations of an EF-3 tornado that hit Boscobel WI

August 13th, 2021 |
GOES-16 ABI Band 13 (“Clean Window”) Infrared imagery (10.3 µm), 2100-2159 UTC on 7 August 2021 (Click to animate)

An EF-3 tornado moved through the southwest Wisconsin town of Boscobel, in Grant County, late in the afternoon of 7 August 2021 (Preliminary Storm Summary from WFO ARX). The tornado was on the ground from 4:29 to 4:56 PM CDT, or 2129 – 2156 UTC. How did the ABI imagery and GLM data change over this time? The Satellite Information Familiarization Tool (SIFT) can be used to investigate this. Gridded GLM data that can be imported into SIFT (a two-week rolling archive is available) is available at this website. ABI Radiance data can be acquired from NOAA CLASS or from the Amazon Cloud.

The GOES-16 ABI Clean Window animation from 2100 to 2159 UTC, bracketing the times that the tornado, linked to the image above, shows very strong upper-level difluence (consider how the cirrus shield spreads south in the hour of the animation!); one might infer cyclonic motion in the fields as well.

SIFT allows for the identification of regions that can then be investigated. The toggle below shows a polygon that has been defined. Subsequent plots will focus on this region surrounding the storm tops associated with the tornadic storm.

SIFT display of GOES-16 Clean Window (10.3 µm) at 2124 and 2157 UTC on 7 August 2021. The transparent red box defines a region being investigated.

How do the cloud-top brightness temperatures evolve in that region? One way to describe that is a simple bar-graph showing the distribution of temperatures, shown below. There are three distinct cold temperature events: around 2130 UTC, around 2138 UTC, around 2148 UTC. (Recall the tornado is on the ground fron 2129-2156) The time-scale of the changes is such that only 1-minute imagery will be able to capture it accurately.

Distribution of 10.3 µm brightness temperatures within a defined polygon as shown above; 2124-2159 UTC on 7 August 2021

How do the lightning observations evolve in the storm? SIFT will display many different GLM parameters: Average and Minimum Flash Areas, Total Energy, Group (and Flash) Extent and Centroid Densities, Group and Flash Areas. Some are displayed below, again within the confines of the polygon defined above. The first plot compares Average Flash Area (along a constant x axis) and Total Optical Energy (along a varying y axis). The distribution in the plot seems to change during the time when the tornado is on the ground.

GLM Average Flash Area v. GLM Total Energy within the defined polygon, 2124, 2127, 2134, 2140, 2149 and 2151 UTC.

SIFT also allows direct comparisons between ABI and GLM data, as shown below: Flash Extent Density is compared to Band 13 (10.3 µm) brightness temperatures at discrete times within the tornado’s lifecycle.

GLM Flash Extent Density vs. G16 ABI Band 13 (10.3 µm) Brightness Temperature within a predefined polygon, 2124, 2127, 2134, 2140, 2149, 2151 UTC

For more information on SIFT, including download instructions for linux, MacOS and Windows, refer to the SIFT website.

Creating and Displaying gridded GLM fields using data from NOAA CLASS

March 23rd, 2021 |

GOES-16 Gridded GLM imagery of Total Optical Energy for the 1 minute ending 1501 UTC on 22 March 2021 (Click to enlarge)

This blog posts describes how to use NOAA’S CLASS (Comprehensive Large Array-data Stewardship System) system (link) that contains Level-2 GLM data, (under the GOES-R Series GLM L2+ Data Product (GRGLMPROD) tab) to create useable GLM imagery. GLM processing produces three Level 2 files each minute, and those files can be processed to produce imagery. First, choose the time range you want in CLASS, and get the global imagery.  For this blog post, I chose GOES-16 data on 22 March 2021 between 15:00 an 15:15 UTC.  On the CLASS website, I clicked the GLM L2+ Lightning Detection Data and didn’t filter by any values (CLASS allows you to filter by minimum/maximum flash, event and group counts, if you want).  This request returned 47 different files, but that is only about 10 Mbytes.  Some of the file names — two minutes’ worth — are shown below: LCFA files from julian Day 081 (that is, 3/22/2021) starting at 15:00:00:00, 15:00:00:20, 15:00:00:40, … etc.

Code to convert these files (that contain raw-ish group, event and flash fields) to gridded GLM fields (that can be displayed with, for example, Geo2Grid, or AWIPS) is within the CSPP Gridded GLM software package that can be downloaded here (free registration may be required; the Gridded GLM tarball to download includes a short and useful README). To create a data file that is properly configured for Geo2Grid (or AWIPS), with software that uses the open-source glmtools software developed by Dr. Eric Bruning at Texas Tech, use this command: ../../data/OR_GLM-L2-LCFA_G16_s20210811501*

That will create a file with a name like this:;

Geo2Grid can then be used to create imagery from the newly-created netCDF file. The Geo2Grid code used is below.

../ TestGridded -75.0 8. 1000 -1000 2000 750 > $GEO2GRID_HOME/TestGridded.conf
../ -r glm_l2 -w geotiff -p total_energy -g TestGridded --grid-configs $GEO2GRID_HOME/TestGridded.conf --method nearest -f /home/scottl/CSPPGeo/GGLM/cspp-geo-gridded-glm-1.0b1/bin/
../ ../../../enhancements/TotalEnergy.txt GOES-16_GLM_total_energy_20210322_150100_TestGridded.tif
../ --add-coastlines --coastlines-resolution=h --coastlines-outline='black' --add-grid --grid-text-size 12 --grid-d 1.0 1.0 --grid-D 1.0 1.0 --add-colorbar --colorbar-tick-marks 250.0 --colorbar-text-size 1 --colorbar-no-ticks --colorbar-align bottom GOES-16_GLM_total_energy_20210322_150100_TestGridded.tif
convert GOES-16_GLM_total_energy_20210322_150100_TestGridded.png -gravity Southwest -fill white -pointsize 24 -annotate +8+30 "1501 UTC 22 March 2021 Total Energy" GOES-16_GLM_total_energy_20210322_1501_Labeled.png

The Geo2Grid package commands above (1) created the grid (‘TestGridded’) onto which the data were interpolated; (2) created the imagery from the netCDF file output from the Gridded GLM package; (3) Added a pre-defined colormap (within ‘TotalEnergy.txt’); (4) Added coastlines, a lat/lon grid, and a colorbar and (5) annotated the image. This last command used ImageMagick.

Note that the GLM image created, shown at top, is mostly transparent. Three areas of GLM observations are apparent, two over South America, one over the Pacific Ocean south of Panama. The transparency is handy if you want to overlay GLM data on top of ABI imagery!