Bárðarbunga, the Toxic Tourist Trap

Quick: what was the name of that Icelandic volcano that caused such a stir a few years ago? Oh, that’s right. You don’t remember. No one remembers. (Unless you live outside the U.S. in a place where you might have actually heard someone say the name correctly.) To Americans, it will forever be known as “That Icelandic Volcano” or “The Volcano That Nobody Can Pronounce” – even though it is possible to pronounce the name. Say it with me: Eye-a-Fiat-la-yo-could (Eyjafjallajökull).

Well, back at the end of August 2014 another volcano erupted in Iceland, and there is no excuse for not being able to pronounce this name correctly: Bárðarbunga. (OK, you have one excuse: use of the letter ð is uncommon outside of Iceland. In linguistics, ð is a “voiced dental fricative” which, in English, is a voiced “th”. “The” has a voiced “th”. “Theme” has an un-voiced “th” or, rather,  “voiceless dental non-sibilant fricative“.) Look, you don’t want to offend any Icelanders, so say it right:

“Bowr-thar-Bunga.” See, it’s easy to say. (You may see people who are afraid of the letter ð refer to the recent eruption as Holuhraun [pronounced “Ho-lu-roin”], because Bárðarbunga is part of the Holuhraun lava field. So be aware of that.)

I know what you’re going to ask: “What is so special about this volcano? I haven’t heard anything about it up to this point, so why should I care?” You haven’t heard anything about it because you don’t live in Iceland or in Europe, which is downwind of Iceland. And, why should you care? Let me count the ways in the rest of this blog post.

You probably have heard of Kīlauea (and have no trouble pronouncing that name) and the lava flow that inched its way towards the town of Pahoa. Kīlauea has been continuously erupting since 1983. Bárðarbunga erupted on 29 August 2014 and has been spewing lava ever since, which at this point, is over 100 days of non-stop erupting. It’s Iceland’s version of Kīlauea. (Hopefully, it won’t continue to erupt for another 30 years.)

Just like Kīlauea, Bárðarbunga is attracting tourists from all over the world. It seems every wannabe photographer and videographer has gone (or wants to go) to Iceland to try to come up with the next viral video showing the breathtaking lava flows. Seriously, do a search for Bardarbunga or Holuhraun on YouTube or vimeo and see how many results show up. Here’s a pretty typical example (filmed by someone from Iceland):

Want to join in the fun? Just grab your camera, head to Iceland, hire an airplane or helicopter pilot, and find the most dramatic music you can think of to go along with your footage. Watch out, though – the airspace around the volcano can be rather crowded. As this video shows, it can be hard to film the volcano without other aircraft getting in the way.

If photography is more your thing, here are the latest images of the eruption on Twitter. (Look for the pictures of Beyonce and Jay-Z. If Twitter is correct, they flew over the volcano for his birthday. Viewing the eruption has gone mainstream! You’re too late, hipsters! Good luck getting to the next volcanic eruption before it becomes cool.)

Back to the matter at hand: why you should care about Bárðarbunga. After its first 100 days of erupting, it has created a field of new lava (76 km2) that is larger than the island of Manhattan (59 km2). The volcano has been creating a toxic plume of SO2 for the last 100 days that is making it difficult to breathe. (Here are some of the known health effects of breathing SO2.) SO2 can ultimately be converted into sulfuric acid (acid rain), depending on the chemistry in the air around the volcano. And while it may not be producing as much ash as Eyjafjallajökull did, VIIRS imagery shows it is producing ash, which is a threat to aircraft.

If you follow this blog, you know the best RGB composite for detecting ash is the True Color composite. This is because the visible wavelength channels that make the composite are sensitive to the scattering of light by small particles, like dust, smoke and ash. Iceland is a pretty cloudy place, so it’s not always easy to spot the ash plume, so here it is at its most visible:

VIIRS True Color RGB composite of channels M-3, M-4 and M-5, taken 12:57 UTC 11 September 2014

VIIRS True Color RGB composite of channels M-3, M-4 and M-5, taken 12:57 UTC 11 September 2014. The red arrow points to the location of Bárðarbunga.

Click on the image (or any other image) to see the full resolution version. The red arrow shows the location of Bárðarbunga. In case you’re wondering, the borders drawn inside the island are IDL’s knowledge of the boundaries of lakes and glaciers (jökull in Icelandic). The big one just south of the red arrow is Vatnajökull – the largest glacier in Europe and one of three national parks in Iceland. (If you want to go there, be aware of closures due to volcanic activity.)

See the ash plume extending from the red arrow to the east-northeast out over the Atlantic Ocean? Now, try to find the ash plume in this animation of True Color images from 29 August to 14 October 2014:

Animation of VIIRS True Color images of Iceland 29 August - 14 October 2014

Animation of VIIRS True Color images of Iceland 29 August – 14 October 2014

As with most of my animations, I have selectively removed images where it was too cloudy to see anything. Sometimes, the steam from the volcano mixes with the ash to make its own clouds, much like a pyrocumulus. Watch for the ash to get blown to the northwest and then southwest in early October. In case you can’t see it, here’s a static example:

VIIRS True Color RGB composite of channels M-3, M-4 and M-5, taken 12:15 UTC 10 October 2014

VIIRS True Color RGB composite of channels M-3, M-4 and M-5, taken 12:15 UTC 10 October 2014. The red arrow shows the location of Reykjavik.

This time, the red arrow shows Reykjavik, the nation’s capitol and likely only city in Iceland you’ve heard of. The ash plume is pretty much right over Reykjavik!

Over the course of the first 100 days, no place in Iceland has been kept safe from the ash plume. But, that’s not the only threat from Bárðarbunga: I also mentioned SO2. If you recall from our look at Copahue (Co-pa-hway – say it right!) the EUMETSAT Dust algorithm is sensitive to SO2. So, can we detect the toxic sulfur dioxide plume from Bárðarbunga? Of course! But, it does depend on cloudiness and just how much (and how high) SO2 is being pumped into the atmosphere.

If you read my post on Copahue, you should have no trouble picking out the sulfur dioxide plume in this image of Bárðarbunga:

EUMETSAT Dust RGB composite applied to VIIRS, 12:57 UTC 11 September 2014

EUMETSAT Dust RGB composite applied to VIIRS, 12:57 UTC 11 September 2014

This image is from the same time as the first True Color image above, when the plume was very easy to see. Also note the large quantity of contrails (aka “chemtrails” to the easily misled). Those are the linear black streaks west of Iceland. If you’re confident in your ability to see the sulfur dioxide, see how often you can pick it out in this animation:

Animation of EUMETSAT Dust RGB images from VIIRS (29 August - 10 October 2014)

Animation of EUMETSAT Dust RGB images from VIIRS (29 August – 10 October 2014)

Some detail is lost because an RGB composite may contain as many as 16 million colors, while the .gif image standard only allows 256. But, you can still see the pastel-colored SO2 plume, which almost looks greenish under certain conditions due to interactions with clouds. Also note the volcano itself appears cyan – the hottest part of the image has a cool color! Unusual in a composite that makes almost everything appear red or pink.

If you want to see the volcano look more like a hot spot, here are animations of the shortwave IR (M-13, 4.0 µm) and the Fire Temperature RGB composite (which I promote whenever I can). I should preface these animations by saying I have not removed excessively cloudy images but, at least 80% of the days have two VIIRS afternoon overpasses and, to reduce filesizes, I have kept only one image per day:

Animation of VIIRS M-13 images of Iceland (29 August - 15 October 2014)

Animation of VIIRS M-13 images of Iceland (29 August – 15 October 2014)

The Fire Temperature RGB is made up of M-10 (1.6 µm; blue), M-11 (2.25 µm; green) and M-12 (3.7 µm; red):

Animation of VIIRS Fire Temperature RGB images of Iceland (29 August - 15 October 2014)

Animation of VIIRS Fire Temperature RGB images of Iceland (29 August – 15 October 2014)

No surprise, molten rock is quite hot! That area of lava has saturated my color table for M-13 and it saturated the Fire Temperature RGB. As I’ve said before, only the hottest fires show up white in the Fire Temperature RGB and lava is among the hottest things you’ll see with VIIRS. Sometimes, you can see the heat from the volcano through clouds (and certainly through the ash plume)! It’s also neat to watch the river of lava extend out to the northeast and then cool.

To quantify it a bit more, the first day VIIRS was able to see the hot spot of Bárðarbunga (31 August 2014), the M-13 brightness temperature was the highest I’ve seen yet: 631.99 K. The other midwave-IR channels (M-12 and I-4; 3.7 and 3.74 µm, respectively) saturate at 368 K. The Little Bear Fire (2012) peaked at 588 K and that fire was hot enough to show up in M-10 (1.6 µm) during the day, so it’s no wonder that we’ve saturated the Fire Temperature RGB.

There’s one more interesting way to look at Bárðarbunga using a new RGB composite. When I was first tipped to this event, I saw this image from NASA, which you can read more about here. That image was taken by the Operational Land Imager (OLI) from Landsat-8 and is a combination of “green, near-infrared and shortwave infrared” channels. Applying this to VIIRS, that combination becomes M-4 (0.55 µm), M-7 (0.87 µm) and M-11 (2.25 µm), which is similar to the Natural Color composite (M-5, 0.64 µm; M-7, 0.87 µm; M-10, 1.61 µm) except for a few notable differences. M-4 is more sensitive to smoke and ash and vegetation than M-5. And M-11 is more sensitive to fires and other hotspots than M-10.

The differences are subtle, but you can see them in this direct comparison:

Comparison between VIIRS "Natural Color" and "False Color with Shortwave IR" RGB composites (12:38 UTC 14 October 2014)

Comparison between VIIRS “Natural Color” and “False Color with Shortwave IR” RGB composites (12:38 UTC 14 October 2014)

NASA calls this RGB composite “False Color with Shortwave Infrared,” although I’m sure there has to be a better name. Any suggestions?

Most of my images and loops have come from the first 45 days after eruption. This was a very active period for the volcano, and is where most of the previously mentioned videos came from. (And trust me, you and your browser couldn’t handle the massive animations that would have resulted from using all 100+ days of images.) To prove Bárðarbunga has gone on beyond that, here’s one of the new RGB composites from 17 November 2014:

VIIRS false color RGB composite of channels M-4, M-7 and M-11, taken 13:42 UTC 17 November 2014

VIIRS false color RGB composite of channels M-4, M-7 and M-11, taken 13:42 UTC 17 November 2014

This image really makes Iceland look like a land of fire and ice, which is exactly what it is!

Abafado Bruma Seca

Hopefully, Google Translate didn’t steer me wrong on the meaning of “abafado”. “Bruma seca” is a term used by Portuguese and Spanish speakers that literally translates to “dry mist”. It is typically used to refer to thick haze or the brownish air caused by dust and, more specifically, to the Saharan Air Layer (scroll down a bit on this Weather Underground blog post for nice description of what that is).

We’re speaking Portuguese today because we are re-visiting Cape Verde, an island nation where people speak Portuguese. (Actually, many people speak a creole version called Kriolu kabuverdianu that has Western African elements added to the Portuguese.) Last time we visited Cape Verde, the islands were creating interesting waves and plumes in the atmosphere. This time, Cape Verde is buried under a plume – a plume of Saharan air that is so thick, you can barely see the islands:

VIIRS "True Color" RGB composite of channels M-03, M-04 and M-05, taken 15:07 UTC 30 July 2013

VIIRS "True Color" RGB composite of channels M-03, M-04 and M-05, taken 15:07 UTC 30 July 2013

I had to plot the map boundaries on the image just to see where the islands are. Otherwise, they would be lost in a sea of brown dust. Also, without the map, it’s difficult to find the shoreline of western Africa because the dust looks just like the Sahara Desert where it came from.

This image is (and the images to follow are) a “True Color” RGB composite. (As always, click on the picture, then on the “2442×1920” link below the banner to see the full resolution image.) Unlike many previous true color images shown on this blog, these have been “Rayleigh corrected.” This means the impact of Rayleigh scattering by the molecules in the atmosphere has been removed. The reason for doing this is that it makes the surface easier to see and it better represents what people normally see when looking out of the window on an airplane. Dust particles, on the other hand, are Mie scatterers at visible wavelengths (refer back to that last link) so they still show up. In fact, this is one of the strengths of the True Color composite: it is quite sensitive to particulate matter in the atmosphere like smoke, smog, haze and dust.

The image above was taken on 30 July 2013, one day after the dust really started to be pushed off the African coast. It is not clear if the people of Cape Verde were forced indoors by this dust since I wasn’t able to find any news reports on it. The western edge of the dust plume (between 28° and 29° W longitude) almost looks like it is casting a shadow, which would indicate the dust is lofted pretty high in the troposphere in this image.

This dust plume pushed across the Atlantic Ocean over the following days. VIIRS passed over Cape Verde on 31 July 2013 (14:48 UTC) and captured this image:

VIIRS "True Color" RGB composite of channels M-03, M-04 and M-05, taken 14:48 UTC 31 July 2013

VIIRS "True Color" RGB composite of channels M-03, M-04 and M-05, taken 14:48 UTC 31 July 2013

Here, the dust plume extends from one side of the swath to the other – over 3000 km. On the very next orbit (16:29 UTC 31 July 2013), the plume can be seen on four consecutive data granules, extending almost to the middle of the swath. (The satellite covers a distance of over 2000 km over four granules.)

VIIRS "True Color" RGB composite of channels M-03, M-04 and M-05, taken 16:29 UTC 31 July 2013

VIIRS "True Color" RGB composite of channels M-03, M-04 and M-05, taken 16:29 UTC 31 July 2013

Hold on. What’s that strip of white-colored stuff extending north-northwest from 50° W longitude label? Some kind of white dust? That happens to be in a straight line? Nope. It’s what is called “sun glint” and it’s the same basic phenomenon as the glare you see looking out over a body of water without polarized sunglasses.  The dust is all the brown stuff on the right side of the image. That’s South America and the Lesser Antilles on the left side of the image.

If you click to the full resolution version of the image above, you may find that the image doesn’t seem very big considering it is made of four granules. (Its pixel size is 1600×1536. In contrast, the image above that is only two granules, yet is 3200×1536 in size.) That’s because I had to reduce the resolution of the data in order to plot it all without running out of memory on my computer. VIIRS has twice the resolution of what is shown in the latter image. (And this high resolution requires a lot of computing power to display!)

On 1 August 2013, the plume pushed even closer to the Lesser Antilles (although they are off the left side of this image).

VIIRS "True Color" composite of channels M-03, M-4 and M-05, taken 16:10 UTC 1 August 2013

VIIRS "True Color" composite of channels M-03, M-4 and M-05, taken 16:10 UTC 1 August 2013

Again, the resolution has been degraded by a factor of two. It is interesting to note that one granule covers an area of the Earth about 3040 x 570 km in size (1.7 million sq km, or 669,000 sq mi), so four granules is about 6.9 million km2. That’s 2.6 million square miles. In comparison, the size of the lower 48 states is about 3.1 million square miles (3.7 million square miles if you add on Alaska and Hawaii).  Now notice that the dust covers most of the last image. If you add on the area of the dust plume that stretches all the way back to Africa, you are talking about an area well over the size of the United States! By the time it arrives in the Caribbean, that dust better learn to speak Antillean Creole. It is a long way from Cape Verde.

So, what does all of this mean? It is often claimed that the presence of Saharan dust layers is bad for hurricane formation. Evidence for that claim is provided here and here. However, there are also scientists who refute that claim, which you can read about here. Scientists at the U.S. Geological Survey (USGS) have found that Saharan dust may be harmful to people and to coral reefs. According to this article in Nature, the dust is beneficial for the Amazon rainforest.

This event was also discussed on the Weather Channel. Compare his visible images to mine, which use only one color of the visible spectrum to my three color images. So, whether Saharan dust is good or bad, I think we can all agree that VIIRS is good!

UPDATE (5 August 2013): Remember the “split window difference”? It was mentioned the last time we visited Cape Verde. Here’s is a split window difference product produced at CIMSS that highlights the plume as it traveled across the Atlantic. This loop starts on 29 July and ends on 2 August 2013 and is made of data collected by the geostationary satellite MSG-3.

UPDATE (19 August 2013): Here’s another animation of the dust plume, made using observations from the Ozone Mapping and Profiler Suite (OMPS), one of the new instruments aboard Suomi NPP alongside VIIRS. (Actually, it’s on the opposite end of the satellite from VIIRS, so it’s not literally alongside VIIRS, but you get the idea.)

Cape Verde Waves and Plumes

Cape Verde is an island nation off the west coast of Africa, located in the North Atlantic. The islands are a popular initiation point for tropical storms. The original capital of the 10-island archipelago was sacked twice by Sir Francis Drake, the same one who, in his later years, would fail to sack the villages along Lake Maracaibo in Venezuela due to Catatumbo lightning. That guy really got around, and I mean that literally: he circumnavigated the globe between 1577 and 1580, sacking nearly every village and boat he came across. But, this isn’t about Francis Drake – it’s about the Cape Verde islands and the amazing view of them captured by VIIRS.

False color RGB composite of VIIRS channels I-1, I-2 and I-3 taken 14:41 UTC 6 June 2012

False color RGB composite of VIIRS channels I-1, I-2 and I-3 taken 14:41 UTC 5 June 2012

Can you see the 10 major islands? One of them (Santa Luzia) is almost obscured by clouds. If you click on the image, you’ll see each of the major islands identified. Go ahead and click on it. It will help for later.

The image above was made from the RGB composite of VIIRS high-resolution imagery channels I-01, I-02 and I-03. While it technically is a false color image (uses reflectance at 0.64 µm [blue],  0.865 µm [green] and 1.61 µm [red]), it looks realistic in many situations, so that we refer to it as “pseudo-true color”. Snow and ice show up as an unrealistic blue, however, which is the main difference between it and a “true color” image. You might also notice a few more differences between the “pseudo-true color” image above and the “true color” image below.

True color RGB composite of VIIRS channels M-3, M-4 and M-5 taken 14:41 UTC 6 June 2012

True color RGB composite of VIIRS channels M-3, M-4 and M-5 taken 14:41 UTC 5 June 2012

The true color image uses moderate resolution channels M-3 (0.48 µm, blue), M-4 (0.55 µm, green) and M-5 (0.67 µm, red), which actually observe radiation in the blue, green and red portions of the visible spectrum. Apart from differences in resolution, the vegetation on the islands shows up a bit better in the “pseudo-true color” image. The islands just look brown in the true color image.

What is particularly interesting about these images are the visible effect that the islands have on the local atmosphere. Downwind (southwest, or to the lower left) of Sal, Boa Vista, and Maio, you can see singular cloud streets, much like the flow of water around a rock. In the photograph in that link, you can see how the water dips downward on both sides of the center line downstream of the rock, and upward in the middle (along the center line). The islands are acting like rocks in the atmosphere, causing upward motion behind them, and this lift was enough to form cloud streets. On either side of these cloud streets there is downward motion and, as a result, clear skies.

Downwind of São Nicolau, São Vicente and Santo Antão, the cloud streets highlight von Kármán vortices and vortex shedding, which you can see in more-controlled lab conditions here and here.

Many of the islands appear to be producing their own aerosol plumes (i.e. dust), and if you zoom in on the area between Boa Vista and Santiago, you can see gravity waves present in some of the plumes (highlighted by the arrows in the image below).

False color RGB composite of VIIRS channels I-1, I-2 and I-3 taken 14:41 UTC 5 June 2012

False color RGB composite of VIIRS channels I-1, I-2 and I-3 taken 14:41 UTC 5 June 2012

A common way to detect dust is the “split-window difference”: the difference in brightness temperature between the 11 µm channel and the 12 µm channel. On VIIRS, this means subtracting M-16 from M-15 which, when you do that, gives you this image:

Split-window difference from VIIRS (M15 minus M16) from 14:41 UTC 5 June 2012

Split-window difference from VIIRS (M15 minus M16) from 14:41 UTC 5 June 2012

The color scale goes from -0.16 K (black) to +4.0 K (white). For some reason, the dust or aerosol plumes don’t produce a strong signal here. It may be that the dust is too low in the atmosphere and the lack of temperature contrast with the surface prevents a strong signal. Maybe water vapor absorption effects in M16 are washing out the signal. Or, there could be some other explanation waiting to be discovered.

The plumes are highly reflective in the 3.7 µm channel (M-12), as are the clouds, which show up as warm spots in the image below (not as warm as the islands, however):

Moderate resolution 3.7 µm image (M-12) from VIIRS, taken 14:14 UTC 5 June 2012

Moderate resolution 3.7 µm image (M-12) from VIIRS, taken 14:41 UTC 5 June 2012

Here, just to throw you off, the color scale has been reversed so that dark colors mean higher values. The scale ranges from 295 K (white) to 330 K (black). When you take the difference of this image and the 10.6 µm brightness temperature (M-15), the clouds and aerosol plumes really show up, along with the gravity waves and vortices:

Brightness temperature difference between VIIRS channels M-12 and M-15 from 14:14 UTC 5 June 2012

Brightness temperature difference between VIIRS channels M-12 and M-15 from 14:41 UTC 5 June 2012

In this case, the M-12 brightness temperatures are always greater than the M-15 brightness temperatures (due to the combination of Earth’s emission and solar reflection in M-12 as opposed to just surface emission in M-15), so the scale varies from +5 K (black) to +30 K (white). Higher (brighter) values on this scale show off where the most solar reflection occurs at 3.7 µm – the liquid clouds and aerosol plumes.

There are much more sophisticated ways of identifying dust and aerosol plumes. To find out more, check out this article written by one of our resident experts, Steve Miller, who is currently working on applying dust detection algorithms to VIIRS.

If you are more interested in the von Kármán vortices, NASA has put together a great page that you can visit here. If you take the original image in this post, zoom out and rotate it a little bit, you can get a sense of just how far the vortices extend from their parent islands:

False color RGB composite of VIIRS channels I-1, I-2 and I-3 taken 14:41 UTC 5 June 2012

False color RGB composite of VIIRS channels I-1, I-2 and I-3 taken 14:41 UTC 5 June 2012. This image has been rotated from the previous images to highlight the length of the vortex streets.

Coincidentally, this image has been cropped to a size that makes it suitable for use as a desktop wallpaper, should you happen to have a 16:9-ratio monitor and a desire to stare at this image all day. (You have to click on the image, then click on the “1920 x 1080” link below the header to get the full resolution image.)