UPDATE #3 02/15/08
I have received seismic data from MEVO for 10/24/07, and I’m now satisfied that no significant volcanic eruption occured on that date. So what we are seeing is either some meteological effect, such as perhaps a lenticular cloud being shredded and blown downwind, or perhaps a steam release without and eruption. I can’t be sure at this point but it has been an interesting time of investigating it and I learned a tremendous amount in the process.
UPDATE #2 02/14/08
I’m still waiting for the director of MEVO, Dr. Philp Kyle, to provide Erebus seismic data for October 24th. He maintains that no eruption occurred on that day, but the MEVO website stops reporting on October 23rd, and so far he hasn’t provided me with the data I’ve asked for. When/if he does, I’ll post it here.
Also the explanation of “cirrus blowoff” seems to be applied only to thuderstorm anvils, and I have not found any references to it related to volcanic peaks. If anyone has a reference, please let me know.
UPDATE: I’ve received a reply from the director of MEVO (Mount Erebus Volcanic Observatory) My initial concerns about this perhaps being misidentified seem to have been valid, even though the photo evidence and anecdotal evidence appears convincing for volcanic plumes.
Here’s what the director of MEVO has to say. I think he means “serious blowoff” in his email. Others have suggested “cirrus blowoff” which turns out to be correct when he emailed back:
—– Original Message —–
From: Philip Kyle
To: Anthony Watts
Sent: Monday, February 11, 2008 9:11 AM
Subject: Re: Sat image shows possible eruption of Mt. Terror
I was in Antarctica at the time of this unusual phenomena and it has been noted by others.
This is a case of sirius blowoff and due to wind and cloud conditions. There was no eruption from Terror (as much as we would love there to be one).
i made a mistake it is “cirrus blowoff”
I am not a weather person.
In the past when we have seen the phenomenon i was told by a weather person that was the correct term.
you will need to consult either a meterologist or the people who generated the satellite image.
If it has been “noted by others”, why not put it on the MEVO web page to clarify it for everyone? The MEVO website hasn’t been updated since October 23rd 2007, and I queried him on that. His reply:
1. We lost our webmaster.
2. The student who was keeping the site up got distracted and did not do the work.
3. Eruptive activity almost ceased at Erebus
4. We were all in the field and involved with field work.
If the scientific organization responsible for providing the volcanic information about the area won’t keep their own web page updated, how would anyone know that this has been “noted by others” and that his eyewitness account says otherwise?
These images have been looked at by a lot of folks who have reviewed weather satellite imagery, so if we’ve all been misled by it, MEVO should put up a notice on their web page so that other researchers are aware. It shouldn’t be that hard to find someone who can edit a webpage at a University.
Even with this new info, this has been a fun learning exercise, and it just goes to show that when you are chasing a hypothesis, you should get all available information at your disposal, and if you operate a scientific website for public consumption, you should keep it updated. – Anthony
ORIGINAL BLOG POST STARTS HERE:
Like I’ve said before, when your work becomes known, people start sending you things to look at. Such is the case here. An email was forwarded to me by Jim Kingsley that contained a NOAA-17 POES satellite photo and discussion about what it meant. The plume to the left was identified as Mount Erebus, but the plume to the right was simply referenced as “new” with no label as to the landmark. There was also the query “What do you think?”
Initially my first thought was “wow”, but then I started thinking it was probably just another Internet hoax or something misidentified, or simply “volcanic business as usual” there and the folks at McMurdo Base knew all about it.
Here is the photo below that started it all:
My thinking was that this was just another Photoshop trick, where the plume from Mount Erebus had been extracted and repasted to the right, giving the appearance of two eruptions side by side. The photo had a caption at the bottom which read:
I figured it would be a simple matter to prove this a fake by finding the original satellite image. I found it, but to my surprise the source satellite image in the AMRC Ross Ice Shelf sat image archives at the University of Wisconsin SSEC looked EXACTLY the same.
Click this link to see it from the University of Wisconsin server.
Note that the link is by day of the year, not calendar date, so I had to use this online day of year calculator to find the right image for October 24th, 2007, which was day 297. I looked at sat images several days before and several days afterwards, seeing no other plumes, it appeared this was a one day event.
Being unfamiliar with Antarctic landscape, it took me awhile to get the “lay of the land” to really understand what I was looking at. I found this map that helped me understand the landscape:
zoomed, click for original source image
Based on the scale key on the map above, and also Google Earths measurement tool, Mount terror appears to be about 20 miles (33 kilometers) to the East of Mount Erebus.
There were no landmarks on the NOAA-17 satellite image at all, I couldn’t even locate McMurdo base. The infrared sat image was low contrast, and had quite a bit of obscuring clouds, so I wanted to be sure these plumes weren’t from some other source, like some experiment where maybe scientists were releasing some smoke to track wind flow or something equally odd but not an eruption.
I figured there had to be another explanation. After all, the reference I found on Mount Erebus, and its companion to the right, Mount Terror, all said that Erebus was active and that Terror was extinct. I wondered why they called it “Terror”, if it was extinct, and soon learned Mount Terror was named in 1841 by it’s discoverer Sir James Clark Ross for his second ship, the HMS Terror. His first ship was named the “Erebus”.
Puzzled, I figured I’d better check to see if in fact Mt. Erebus was erupting at that time, but more importantly, I figured that if Mount Terror had erupted at the same time, there would surely be some mention of it someplace, since Antarctica is crawling with scientists.
To my surprise, I found nothing indicating any awareness of an eruption.
The authority for volcanism on Ross Island, the New Mexico Tech, operates the Mount Erebus Volcanic Observatory. But while there was a plethora of info about eruptions of Erebus, including photos, movies, a seismic network, and a live cam, there was no mention of anything about Mount Terror. Curiously, the last eruption they recorded on Erebus was on October 23rd, 2007, and there were quite a number of eruptions leading up to that. None are listed since, but given the previous frequency of Erebus eruptions listed there, it seems likely that may be simply a web page gone stale. Other searches on the Internet for a reference to an eruption on Mount Terror proved fruitless, but I did find this below.
From International Volcano Research Centre they write:
“As of the 1st of February, the Mt. Erebus Volcano Observatory (MEVO) reports that the volcano continues to have frequent Strombolian eruptions with infrequent ash eruptions.”
“The Mt. Erebus volcano in Antarctica was successfully forecasted by INTLVRC’s programme ERUPTION Pro 10.7 to erupt in 2008 with 100% probability.”
Clearly though, from a graph of eruption frequency of Mt. Erebus, there appears to be an increase in activity over the last couple of years, though it is not clear if the lack of the updated entries for late 2007 after October 23rd into 2008 is lack of eruptions or a reporting problem:
I also noted that in the live cam image, you can’t see Mount Terror, some 20 miles away. But this makes sense, why monitor an extinct volcano?
So the next thing I did was to try to ascertain that the plume seen in the original NOAA-17 satellite image was in fact coming from Mount Erebus. The original photo had low contrast, cloud cover, and no good landmarks, so it was hard to be sure where the plumes originated. Once I was sure that the leftmost plume was in fact Erebus, it then should be easy to correlate the position of Terror.
Having some skill in computer graphics, I decided that I’d try to enhance the contrast of the original satellite image to see if Ross Island would become visible. I was successful:
click for a larger image: ross-ice-shelf-satellite-10-24-07-contrast-enhanced.jpg
The characteristic “bulb” on Ross Island which is Mount Bird, became clearly visible, identifying the source of the plumes as being on Ross Island. I’ve created this image to help you identify the “bulb” in the contrast enhanced image above.
But I still wasn’t happy with the result, so I tried a different technique. I located a satellite photo of a clear day in the Antarctic summer, the NOAA-17 photo from December 15th, 2007 which clearly shows the Ross Island outline and its features. Note though it appears this photo is using a different IR channel than the winter photo, so the picture looks different. This is probably to help spot iceberg calving better.
So I combined these two NOAA-17 images:
The process I used was 66% transparency for the 10-24-07 image placed on top, and did a 3x contrast enhancement to help bring out details. This is the resultant image:
While there was some slight orbital perspective misalignment of the coastline (or perhaps shelf movement) from the 10-24-07 picture to the 12-15-07 picture, from the characteristic outline of Ross Island it is clear that the plumes are originating from Ross Island and near the summits of Mount Erebus and Mount Terror.
See the zoomed sector below:
I also used Google Earth to help me pinpoint the two given their published lat/lons (Mount Erebus -77.53° 167.17° Mount Terror -77.5167° 168.533°). The Google Earth Image below is rotated to the right:
Ok, so I’d proved the plumes are coming from very near or at the summits of Mt. Erebus and Mount Terror.
UPDATE: Alert reader Dave D. provided a link to a TERRA/MODIS image that showed much greater detail at 500meter resolution. It is in bands 3,6 and 7 (land/cloud/aerosols properties) at 500m resolution taken at 19:40 UTC 10/24/07
But I still had some issues with the images I’d adjusted and composited:
This is an infrared image, thus warmer things should show up as darker (like the sea) so why do the plumes look white?
What else could be causing the plumes? High wind driven blowing snow from the summits or simple orographic lifting creating a cloud trail perhaps? Contrails?
If this were a volcanic eruption on Mount Terror, why hasn’t this been reported on an continent crawling with scientists, seismic monitors, live cameras and satellite observations?
So I looked further trying to answer these questions.
1. This is an infrared image, thus warmer things should show up as darker (like the sea) so why do the plumes look white?
Normally that would be a dead giveaway as it being a “cold” plume on a GOES satellite image, but this is POES, and a visit to the NOAA-17 page provided a clue under the AVHRR sensor characteristics table:
|Medium Wave IR||3B||3.55-3.93||sea surface temperature, volcano, forest fire activity|
I’d noted that in the December 15th, 2007 image, the IR channel appears to have changed from the image taken on 10-24-07 that showed the plumes. Unfortunately, there is no reference in the University of Wisconsin imagery that says what channel is used, or when it might be switched. It seems plausible though that they would switch channels depending on what they were looking for, such as iceberg calving, which seems to be the main mission of the Antarctic Meteorological Research Center at the university of Wisconsin where the sat images are processed and archived.
So its possible they were running on IR channel 3B at the time, which by the AVHRR sensor description should be sensitive to such events. Another thought I had was that the plumes weren’t ash or steam, but rather ice crystals from steam that perhaps froze in the cold, high, thin air where the steam was ejected.
UPDATE: Barry Wise sends me a link at the European Space Agency to this photo of Mt. Etna erupting, showing a similar image in IR with an explanation that sound much like my question and conjecture.
2. What else could be causing the plumes? High wind driven blowing snow from the summits or simple orographic lifting creating a cloud trail perhaps?
Fortunately, The University of Wisconsin operates a number of Automatic Weather Stations around Mount Erebus, and I thought that would give me a clue as to the likelihood of that possibility. Fortunately they take pictures of their weather stations (hmmm where have I seen that before?) every year.
AWS at Willie Field in 2002
They also conveniently provide an overlay of weather station plots on the NOAA-17 satellite imagery from the Ross ice shelf. Here is the image from 10-24-07 with those plots:
The surface winds, while strong, didn’t appear to correlate well directionally with the “volcanic” plume directions, and by looking at cloud patterns there appears to be a curvature to the right for cloud movement, so the case for wind driven snow off the summits forming downwind plumes doesn’t seem strong. Nor does the case for orgraphic lifting creating a cloud trail, which would likely be different and divergent from peak to peak. It seems from the straight and nearly identical flow, that the plumes had reached stratospheric height, something not uncommon for a volcanic eruption but unlikely for wind driven summit snow or orographic lifting.
UPDATE: new imagery from TERRA/MODIS (above in orange) shows the plumes casting shadows on the cloud deck below it, lending credibility to the idea of the plume originating clouds reaching the stratosphere. This could also be related to a lenticular cloud phenomena, but more study is needed.
3. If this were a volcanic eruption on Mount Terror, why hasn’t this been reported on a continent crawling with scientists, seismic monitors, live cameras and satellite observations?
This didn’t seem to make sense. But, as I began to get a “lay of the land” by studying the landscape, the placement of sensors and cameras, looking at the research focus, and finally going back to look at other satellite images for months either side of the eruption, it started to become clear how such an event could be missed. Here is my reasoning:
The scientific attention is focused on Mt. Erebus. Mt. Terror is “extinct” so why watch it?
Mt. Terror is apparently not visible from McMurdo Base, but it is visible (see photos further below) from Scott base just two miles away on the other side of the ridge. Mt. Terror is lower in elevation that Mt. Erebus. So, if there weren’t any eyes near McMurdo at the correct vantage point, the Mt. Terror plume might go unnoticed. Scott base is less populated than McMurdo.
Scott base with Mt. Erebus, Mt Terror is beyond to the right (see photos further below).
Here is a YouTube Video tour of McMurdo Station and a second video tour.
Erebus erupts regularly, the closest eruption reported on October 23rd, 2007 by MEVO was routine. It is still unclear why MEVO has not reported any eruptions since then though they appear to have regular frequency. Thus there didn’t seem to be any cause for alarm or to go looking at Mt. Terror. The question remains as to why MEVO has dropped the ball on updating their web page beyond October 23rd, 2007 when there is anecdotal evidence of eruptions of Mt. Erebus from the International Volcano Research Centre
- Weather appears to also be a complicating factor. A review of satellite images on the day shows cloud cover in the area. A review of additional satellite images prior to and after the 10-24-07 satellite image showing dual plumes show that cloud cover in that area is common, further complicating a visual sighting. This it seems unlikely that human eyes would have seen an eruption or plume emanating from Mount Terror.
Here is a photo showing Mt. Terror as seen from Scott base, notice the cloud deck at summit. On the day in question it was cloudier than this according to the satellite imagery. It would be difficult to discern a plume even with a “good” weather day as shown above.
Click for a large panorama of Mt. Erebus, and Mt. Terror seen from Scott Base.
- In checking satellite images bracketing 10-24-07, it appears this was a one day event, maybe even just a few hours and the satellite got lucky to be in the right place at the right time. Something so short could also easily be missed by ground observers.
Thus my conclusion:
I’ve spent several days looking at this issue, scanning and creating imagery, plus gathering all other supporting evidence I could find, including seismographic records, surveys, weather/winds at the time, and additional satellite imagery and absent any other explanation that fits, it seems that yes, on October 24th, 2007, on Ross Island in Antarctica, Mount Erebus and Mount Terror (said to be extinct) both let loose with some sort of an event, which may be steam, or steam and ash.
End note: I’ll be the first to admit that I am not a vulcanologist, and that my analysis could be wrong, and that the plumes seen on October 24th, 2007 have some other explanation. However, this entry defining “extinct volcano” with historical anecdotes in Wikipedia gives me some encouragement:
…volcanoes may remain dormant for a long period of time and it is not uncommon for a so-called “extinct” volcano to erupt again. Vesuvius was thought to be extinct before its famous eruption of AD 79, which destroyed the towns of Herculaneum and Pompeii. More recently, the long-dormant Soufriere Hills volcano on the island of Montserrat was thought to be extinct before activity resumed in 1995.
I feel I’ve done a reasonable job investigating this and looking at alternatives given the time and resources available to me. Just writing this blog entry took me six hours of fact checking, link collection and imagery double checking. I now turn this over to others for falsification or confirmation. Hopefully scientists close to the issue will take an interest and do a site survey of Mt. Terror to determine if any evidence of recent activity exists.
If Mt. Terror did in fact erupt, it may indicate a resurgence of volcanic activity near the Ross Ice Shelf, which is worrisome to many. I welcome comments.