SATELLITE
INTERPRETATION
DISCUSSION
NOAA/NESDIS
 Regional and Mesoscale Meteorology Team
Cooperative Institute for Research in the Atmosphere (CIRA)
Colorado State University   Fort Collins, Colorado

December 20, 2000

Volcanic Ash and Hot Spot Detection Using Multi-spectral Imagery

The detection of volcanic ash and hot spots is easily accomplished using GOES Imager data provided that the ambient atmosphere is basically cloud free.  This discussion covers an interesting case of an eruption of the Popocatepetl (hereafter abbreviated as Popo) volcano on 14 December 2000 (Julian day 349).  Popo has been minimally active for many months but on 13 December started to give off more significant amounts of ash.  These images were collected on the second day of increased volcanic activity.

Two loops of GOES image products are provided.  The first image loop contains visible images from 1315 (sunrise) to 1845 UTC.  These images are enhanced by a solar zenith angle correction that brightens all images so that they appear as if they were taken with the sun directly overhead.  The first image (1315 UTC) shows the day/night terminator, the right side is visible data.  The left side of the image is filled in with an infrared image product that is substituted when visible images are not available at night.  Popo is near the center of the images and is blowing off water vapor and ash that spreads out in both the east and west directions when the images are looped.

Click on images to start loops


Loop 1 
1315 to 1845 UTC

Loop 2 
1145 to 1915 UTC

The second image loop contains infrared images from 1145 (before sunrise) to 1915 UTC.  Each image is a 4-panel composite of image products generated from three GOES infrared channels (channel-2 [shortwave IR, 3.7 µm], channel-4 [longwave IR, 10.7 µm], and channel-5 [longwave IR, 12 µm]).  These channel images are converted into three Principal Component Images (PCIs) that are sums and differences of the three channels.  The fourth panel is a graph of the transformation that is used to convert the three GOES infrared channels into three PCIs.  The first component (PCI-1) is a composite image of all three bands and contains features that are similar to the three bands.  The second component (PCI-2) is a channel-2 minus channel-4/5 difference image. The third component (PCI-3) is a channel-4 minus channel-5 difference image with little or no contribution from channel-2.  The plot is generated for each set of input images and varies with the images as they are looped.  For more information on PCIs see: http://www.cira.colostate.edu/ramm/cal_val/PCI.htm

The volcanic exhalation/plume is visible in all three components, but appears different in each of the component images.  Normally small amounts of volcanic ash are nearly invisible in normal infrared imagery, but due to the larger eruption the ash plume is somewhat visible in PCI-1.  PCI-2, being a shortwave IR minus a longwave IR difference, shows the water vapor portion of the plume, due to differential absorption of water vapor between these two spectral regions.  PCI-3, being a longwave “split-window” difference, shows the ash portion of the plume.  Thin clouds or cirrus normally appear black in this component due to differential transmission through the thin cloud, and the ash cloud behaves similarly.

For verification, a Volcanic Ash Advisory Statement (VAAS) issued by the Washington DC Volcanic Ash Advisory Center (VAAC) is given for 1915 UTC on 14 December.   This is the time of the last image in the PCI loop.  The extent of the ash agrees with that given in PCI-3.  Independent determinations of the flight level of the ash were assigned to different portions of the ash in the VAAS.

Click on images to enlarge

A final figure shows a view of Popo collected at 1932 UTC by a website camera pointed towards the volcano.  The time of this view is shortly after the last image in the PCI loop.  Views such as this are available on a website maintained by CENAPRED (Centro Nacional de Prevención de Desastres [National Center for Prevention of Disasters]) in Mexico City.  A plume rising from the volcano is visible, but the extent of the plume is not apparent from the limited field of view of the camera.

The types of images shown in this discussion are now available at the NESDIS Synoptic Analysis Branch (SAB) for operational use in volcanic ash detection.  GOES images are collected over sectors chosen to cover any of several volcanoes that are currently active, such as Soufriere Hills on the island of Montserrat and the Tungurahua and Guagua Pichincha volcanoes in Ecuador.


Back to top

We welcome your comments and discussion at ramsdis@comet.ucar.edu


Previous RAMMT Satellite Interpretation Discussions

Previous RAMMT Satellite Interpretation Discussions in reverse chronological order
 
 

Information Contact: Lewis Grasso
CIRA/RAMM WebMaster: Hiro Gosden
Author:  Don Hillger
Last Updated: December 20, 2000