Puttippoq? Aatsuu

Posted on February 20, 2020 by Curtis_Seaman

For once, I don’t have all the answers. That’s why I said “aatsuu“. That is an Inuit (Inuktitut) word for “I don’t know.” We’re learning Inuit language today because I wonder how they would describe a recent event in Antarctica. You see, I had been told growing up that the Inuit had more than 30 different words for “snow”, so who better to describe the changing surface properties of snow and ice?

But, as it turns out, that is a controversial statement. It has led to what linguists refer to as “the Great Eskimo Vocabulary Hoax.” There are many other blogs and podcasts that have talked about this “myth“. Exactly how many Inuit words there are for snow (or ice) depends on a lot of factors. The two biggest factors are: What is an “Inuit” language? And, what is a word? “Inuit” used here is a blanket term used to describe the native people of the North American Arctic and a few groups in far-eastern Siberia, which includes distinct groups of people that call themselves Inuit, Inupiat, Yupik, and Alutiit, among others, and have a variety of different languages. One commonality is that they all have agglutinative languages. Simply put, they combine root words with modifiers to create complex words that take the place of phrases. It is summarized succinctly in this comic. So, we might describe snow as “wet and heavy” or “light and fluffy”, while an agglutinative language would say “snowwetandheavy” or “snowfluff” to mean the same thing.

If you focus only on the root words, you get a small number of words that is similar to the number of words in English. If you add in all the possible modifiers, you get a limitless number. (Some of these are amazingly specific, such as qautsaulittuq: “ice that breaks when its strength is tested using a harpoon.”)

As part of the International Polar Year 2007-2008, the Sea Ice Knowledge and Use (SIKU) project (“siku” is the Inuit root word for “ice”) combined the efforts of physical and social scientists to better characterize our collective understanding of ice behavior in the Arctic by studying the native Arctic residents’ understanding of ice behavior, in part, through their culture and language. The discussion on the variety of words for snow and ice takes up five chapters of this compilation of SIKU research. That’s where I learned that puttippoq means an ice surface that has become wet due to melting. (You can read their take on the Great Eskimo Vocabulary Hoax here.)

Didn’t think you’d see a discussion on linguistics in a blog about satellite meteorology, did you? So, let’s get to the satellite meteorology. We’ll start with a look at what I previously called the “mystery channel“, although a better name for it is the “snow band”, since it is very sensitive to the properties of snow and ice.

As always, it is best to view this video in full screen mode. What you are seeing is a compilation of VIIRS band M-08 (1.24 µm) images from both S-NPP and NOAA-20 from 12-13 February 2020, and there are two interesting things to note. First, the left half of the image is the high-elevation Antarctic Plateau, which contains a very bright feature that is very stationary. The right side of the image is low-elevation and contains the southernmost tip of the Ross Ice Shelf (outlined on the map). The Transantarctic Mountains (or, more specifically, the Queen Maud Mountains) in the middle separate the two regions. Pay attention to the expanding dark region on top of the ice shelf.

Since it is difficult to focus on more than one thing at a time, let’s focus on the ice shelf first. (Coincidentally, I haven’t found an Inuit word for “ice shelf”, but I did find sikuiuitsoq, which means “ice that doesn’t melt” – used to refer to ice that has been around a long time, which certainly applies to the Ross Ice Shelf.)

Animation of VIIRS M-08 images (12-13 February 2020)

This is an animated GIF that you will have to click to view. This feature shows up in the longer-wavelength bands, M-10 (1.61 µm) and M-11 (2.25 µm):

Animation of VIIRS M-10 images (12-13 February 2020)

Animation of VIIRS M-11 images (12-13 February 2020)

But, see if you can find it in the shorter-wavelength bands, M-07 (0.86 µm) and M-05 (0.67 µm):

Animation of VIIRS M-07 images (12-13 February 2020)

Animation of VIIRS M-05 images (12-13 February 2020)

At the shorter wavelengths, the feature only appears at certain times, suggesting a viewing angle dependence on the reflectance. That means the bidirectional reflectance distribution function (BRDF) is not uniform.

The explanation for this feature is pretty simple. The cold air over the Antarctic Plateau sinks down through the canyons in the Queen Maud Mountains, and as it descends, the air compresses and warms. These are called katabatic winds. In this case, the katabatic winds are aided by the synoptic scale flow as evidenced by the cloud motion. This relatively warm wind is likely melting the top surface of the Ross Ice Sheet, causing a drop in reflectance in the short-wave infrared (IR) similar to what we’ve seen before. In fact, the darkest regions of those canyons are where the howling katabatic winds have scoured away all the snow, leaving behind only the oldest glacial ice. And glacial ice has the largest grain sizes of any of the ice out there, which we know is a big factor on ice reflectivity in the shortwave-IR. (Watch those animations again and note that M-11 appears to provide the strongest signal of blowing snow coming out of those canyons. This is exploited by the Day Snow/Fog RGB.)

For comparison purposes, let’s look at the Natural Color RGB (also known as the Day Land Cloud RGB), made up of M-05 (blue), M-07 (green) and M-10 (red):

Animation of VIIRS Natural Color RGB composite of M-5, M-7, and M-10 (12-13 February 2020)

And, what we are calling the VIIRS “Snowmelt” RGB (M-05/blue, M-08/green, M-10/red):

Animation of VIIRS Snowmelt RGB images (12-13 February 2020)

And, finally, a variation of the “Snow” RGB developed by Météo-France (M-11/blue, M-08/green, M-07/red):

Animation of VIIRS MeteoFrance Snow RGB images (12-13 February 2020)

The inclusion of M-08 makes a big difference on the visibility of this feature. And, in contrast, this is one application where True Color imagery (M-03/0.48 µm/blue, M-04/0.55 µm/green, M-05/0.67 µm/red) is of no help at all:

Animation of VIIRS True Color images (12-13 February 2020)

As for the second region of interest from the original video, “Aatsuu”. We have a region of ice and/or snow in the Antarctic Plateau that is significantly brighter than its surroundings in the shortwave IR. The question is: why is it such a well-defined shape with a distinct edge to it? Here are all the same bands and RGBs as above:

Animation of VIIRS M-05 images (12-13 February 2020)

Animation of VIIRS M-07 images (12-13 February 2020)

Animation of VIIRS M-08 images (12-13 February 2020)

Animation of VIIRS M-10 images (12-13 February 2020)

Animation of VIIRS M-11 images (12-13 February 2020)

Animation of VIIRS True Color RGB images (12-13 February 2020)

Animation of VIIRS Natural Color RGB images (12-13 February 2020)

Animation of VIIRS Snowmelt RGB images (12-13 February 2020)

Animation of VIIRS MeteoFrance Snow RGB images (12-13 February 2020)

We know that smaller particle size leads to increased reflectivity in the shortwave IR. And, fresh snow typically fits that bill. But, fresh snow tends to appear more streaky (technical term) in satellite images. It’s the distinct edges that are so puzzling.

Anyone with more experience about the ice properties on the Antarctic Plateau out there? Or, experts at what makes snow and ice bright in the shortwave IR? If so, feel free to post a comment. (But, any theories involving UUSOs or UUIOs [Unidentified Under Ice Objects] will be placed in this blog’s trash.)

If not, isn’t this what graduate students are for?

Polar Opposites

Posted on August 15, 2019 by Curtis_Seaman

As we all know, the furthest south you can travel is to the South Pole – the Geographic South Pole, not the Magnetic South Pole or the Geomagnetic South Pole. When you get there, try to face east if you can. (This is easier to do at the “Ceremonial South Pole” than it is at the actual South Pole.)

The furthest south you can get by boat is an island off the coast* of Antarctica, called Ross Island. (*The term “coast” is used loosely here, since Ross Island is usually connected to Antarctica by the Ross Ice Shelf.) At the southern tip of Ross Island is the largest “city” in Antarctica: McMurdo Station. McMurdo is the port-of-entry for most visitors to Antarctica. It is also home to a ground station that receives data from NOAA-20 (and many other satellites). So, if you love the lower latency that comes with NOAA-20 VIIRS data, you have McMurdo Station to thank. (S-NPP data is only downlinked at Svalbard – once per orbit – while NOAA-20 is downlinked at both Svalbard and McMurdo.) This is the location of today’s resolved mystery.

The mystery began with the development of a new website for viewing global VIIRS imagery: Polar SLIDER*. (*Shameless self-promotion: I helped develop that website.) If you click on that link, choose “Southern Hemisphere” from the Sector menu to view Antarctica. With every product, you can zoom in anywhere on the globe* to view the full resolution data. (*Claim is void near the Equator.) Under the Product menu, you can choose between all 22 VIIRS channels (16 M-bands, 5 I-bands, and the Day/Night Band), or from a list of imagery products and cloud products. (And we are always working to add new products.) Since it’s perpetual night down there right now, you’ll notice that the visible and near-IR bands don’t give you much information – except the Day/Night Band, of course, which can provide images like this:

NOAA-20 VIIRS DNB image (14:25 UTC, 14 August 2019)

Ross Island is in the center of that image. That bright light at the southern tip of Ross Island is McMurdo Station. The second bright light south of that is the “airport“. Here’s an annotated image with the map plotted on it:

NOAA-20 VIIRS DNB image of Ross Island and surroundings (14:25 UTC, 14 August 2019)

As always, click on an image to see it in full resolution. Now that we have our bearings, let’s look at the high resolution mid-wave IR band (I-4/3.74 µm):

NOAA-20 VIIRS channel I-4 (14:25 UTC, 14 August 2019)

See that white dot in the middle of Ross Island? What is that? (Hint: it’s not part of the map.)

To make some sense of this, look at the color table plotted on the bottom of the image. White pixels on this scale (not counting the map) are +100°C (+373 K). In contrast, the dark turquoise color surrounding it is in the -25°C to -30°C range (243-248 K). What could be over 100°C in Antarctica in the winter? Did something catch on fire?

It turns out, it is a semi-permanent feature according to this animation collected from Polar SLIDER. (You have to click on the animation to see it play.)

Animation of VIIRS channel I-4 images (13 August 2019)

Looking at Day/Night Band images over the same time period, it also shows up as a bright spot:

Animation of VIIRS DNB images (13 August 2019)

Maybe it’s a nuclear reactor that powers all of McMurdo Station? (Nope. There was a nuclear power plant, but that was de-commissioned in 1972.) Maybe the fact that this bright (in the DNB), hot spot (mid-IR) is on top of a mountain has something to do with it? (Bingo!)

Ross Island is made up of volcanoes, the most prominent of which are Mt. Erebus and Mt. Terror (named for the ships on the original expedition that discovered them). Mt. Terror (the one on the right) is inactive. Mt. Erebus, on the other (left) hand, is the southernmost active volcano in the world. And, what’s relevant here is the fact that it is home to one of only five known lava lakes in the world. So, molten-hot liquid rock exists in an ice-covered environment where temperatures regularly dip down to -50°C or -60°C. And, it’s right next to the largest settlement in Antarctica. Sleep tight. (Since we’re less than a week away from the first sunrise of the spring, get your sleep while you can down there!)

B-31 and the Pine Island Glacier

Posted on December 31, 2013 by Curtis_Seaman

Nope. This post is not about a warplane, an alcoholic beverage or a “New Wave” band from the 1970s. (Those are all B-52s.) And I’m not talking about a county road in Michigan or a New York City bus line. B-31 is the rather bland name given to the massive iceberg that just broke off from the Pine Island Glacier in Antarctica. (Of course, if you tried to name every chunk of ice floating around Antarctica, how long would it take you to run out of names and just switch to random letters and numbers?)

This particular chunk of ice is special, however, as it has been described as the size of a city. Now, as a scientist, I have to say that the size of a city is a terrible unit of measurement. How big a city are we talking about? I suspect people who live in one of the ten largest cities in the world would laugh at what the people of Wyoming call a “city”. And are we talking the size of the greater metropolitan area or just what is within the city limits?

The article that describes B-31 as the size of city mentioned that it was roughly the size of Singapore, or twice the size of Atlanta. Those seem like odd choices for comparison. How many of you have a good idea of what the land area is of Singapore? And twice the size of Atlanta? They could have used New York City, which has just over twice the land area of Atlanta and people are probably more familiar with New York City. In any case, all of these size estimates have errors.

The original estimate came from this NASA MODIS image and associated caption, which put the size of B-31 as 35 km x 20 km. Now, that’s 700 km² assuming the iceberg is a perfect rectangle, which you can see in the image that it isn’t. Singapore has a land area of 714 km², while New York City is 768 km² and Atlanta is 341 km² (these are “within the city limits” numbers, not the size of the greater metropolitan area). Since the iceberg is actually smaller than the 35 km x 20 km rectangle based on the widest and longest dimensions of the iceberg, maybe “twice the size of Atlanta” is the most accurate estimate.

Anyway, MODIS is not the only satellite instrument out there capable of viewing B-31. Landsat-8 saw it in much higher resolution in another post from NASA. And, of course this entire blog is about what VIIRS can see. Now, VIIRS doesn’t have the resolution of Landsat or the highest-resolution channels on MODIS, but VIIRS has the Day/Night Band, allowing us to see the iceberg both day and night (at visible wavelengths).

To show why that is important, take a look at the infrared image (M-15, 10.7 µm) below. Images in the “infrared window” (the N-band window, according to this site) used to be the only way to detect surface features and clouds at night. At these wavelengths, the amount of radiation detected by the satellite is a function of the temperature of the objects the instrument is looking at. As always, to see the high resolution version of the image, click on it, then on the “1660×1706” link below the banner.

VIIRS IR image (M-15) taken 23:34 UTC 7 November 2013

See that slightly darker gray area near the center of the image? That’s open water in Pine Island Bay, which is only slightly warmer than the ice and low clouds surrounding it. Otherwise, there isn’t much detail in this picture. What really stands out are the cold, high clouds that are highlighted by the color scale. Contrast this with a visible wavelength image from the same time (M-5, 0.67 µm):

VIIRS visible (M-5) image, taken 23:34 UTC 7 November 2013

The open water in Pine Island Bay shows up clear as day because, well, it is daytime and the ice and snow reflect a lot more sunlight back to the satellite than the open water does. Icebergs can easily be distinguished from the low clouds now. You can even see through some of the low clouds to identify individual icebergs that are not visible in the infrared image. The difference in reflectivity between the ice and water at visible wavelengths is a lot greater than the difference in brightness temperature in the 10-12 µm infrared wavelengths, and that contrast is what makes things more easily visible.

Now, it is summer down there and at these latitudes, the sun is up for most of the day (actually, all day for everywhere in this scene on the Summer Solstice, which occurred on 21 December 2013), so you could say that using the VIIRS Day/Night Band to look at this stuff is unnecessary. But, since VIIRS is on a polar-orbiting satellite, it views the poles a lot more frequently than where you or I live: every 101 minutes on average, instead of every 12 hours in the low and mid-latitudes. That means it may occasionally capture a nighttime image here or there during the short nights and will frequently capture images where the day/night terminator crosses through the scene and we still want to be able to see what’s going on then. And you need the Day/Night Band to do that.

For the first time on this blog, however, we’re not going to show the Day/Night Band data exactly. We’re going to show the Near Constant Contrast imagery product, which is produced from the Day/Night Band. You can read up more on the Near Constant Contrast product and how it’s related to the Day/Night Band here. At this point, we’ll refer to NCC and DNB rather than having to type out Near Constant Contrast and Day/Night Band all the time.

Here’s a NCC image from 7 November 2013 at 20:15 UTC where the Pine Island Glacier has been identified. B-31 is still attached to the glacier – it’s sticking out into the bay and, if you look at the high resolution version of the image, you may be able to see the crack where it has started to calve.

VIIRS Near Constant Contrast image from 20:15 UTC 7 November 2013. The Pine Island Glacier is identified.

Keep your eye on that spot as you watch this zoomed-in animation of NCC images starting from the above image to 03:06 UTC 18 November:

Animation of VIIRS NCC images of the Pine Island Glacier from 7-18 November 2013

I should say that the above animation does not include images from every orbit. I’ve subjectively removed images that were too cloudy to see anything as well as images where the VIIRS swath didn’t cover enough of the scene. This left 25 images over the 11 day period. Even so, VIIRS captured the moment of B-31 breaking free quite well.

Imagine the sound that this 600+ km² chunk of ice made as it broke free. I bet it sounded something like this glacier calving event in Greenland:

One of the articles linked to above mentioned the importance of tracking such a large iceberg, because it could impact ships in the area. (Just this week a ship got stranded in ice off the coast of Antarctica.) So, I decided to see if VIIRS could track it. The results are in the MP4 video clip linked to below. You may need an appropriate browser plug-in or add-on (or whatever your browser calls it) to be able to view the video.

Animation of VIIRS NCC images from 7 November – 26 December 2013 (.mp4 file)

That’s 50 days of relatively cloud-free VIIRS NCC images (7 November – 26 December 2013), compressed down to 29 seconds. Go ahead, watch the video more than once. Each viewing uncovers additional details. Notice how B-31 doesn’t move much after 10 December. Notice how ice blocks the entrance to Pine Island Bay at the beginning of the loop, then clears out by the end of the loop. Notice all the icebergs near the shore that are pushed or pulled or blown out to sea from about 20 December through the end of the loop. Notice that B-31 isn’t even the biggest chunk of ice out there. Notice the large ice sheet on the west side of Pine Island Bay that breaks up right at the end of the loop. In fact, here’s another zoomed-in animated GIF to make sure you notice it:

Animation of VIIRS NCC images from 20-26 December 2013

That area of ice is much larger than B-31! (Dare I say, as large as the state of Rhode Island? Probably not, because then you’ll just think of how Rhode Island is the smallest US state, so it can’t be very impressive. It’s also not very accurate since that estimate is based on eye-balling it and thinking it looks like it could be four times the size of B-31.)

Of course, we are heading towards the middle of summer in the Antarctic when the ice typically reaches its minimum extent. So the ice breaking up isn’t unusual. Plus, large calving events occur on the Pine Island Glacier every few years. But, the B-31 event is noteworthy because Pine Island Glacier holds about 5% of the total freshwater contained on Antarctica. It’s also the site of an ongoing field experiment where researchers are investigating glacier-ocean interactions. You can read up on what it’s like to install instruments on a glacier while living in a tent on the coldest continent 1000 miles from any other human settlement in this article. (That article doesn’t say if any instruments are still stuck in B-31 and floating out to sea, though.) And, if you’re curious, Pine Island Glacier has its own Twitter account. So far, the conclusions are that Pine Island Glacier is thinning, receding and speeding up. Large calving events are just one piece of the puzzle, but an important piece to understand since they contribute to sea level rise.

The calving process of B-31 was first noticed by NASA researchers noticing a crack forming in Pine Island Glacier while flying over the area in October 2011 – before VIIRS was even launched. But, VIIRS was there to capture the end result of that crack two years later!

UPDATE (22 April 2014): B-31 has continued to drift towards the open ocean. Researchers at NASA have been monitoring the movement of the massive iceberg since it first calved, and have put together their own video here, which tracks B-31 from the time of my video above into mid-March 2014.

Aurora Australis from the Day-Night Band

Posted on October 10, 2012 by Curtis_Seaman

How fast does an aurora move? I “googled” it, and got answers ranging from “fast” to “very fast”. Not very scientific. It also doesn’t help that the majority of aurora videos on the Internet are time-lapse footage, and there’s no way to know how fast the footage has been sped up. Although, I did find this video that claims to be real-time footage:

When the camera is still, you could try to calculate the speed of some of the aurora elements if you knew where the cameraman was, what stars were in the view (and how far apart they are), and how high up (or how far away) the aurora was at that time. All information that I don’t have.

What if I said we could estimate the speed of the aurora by examining VIIRS Day/Night Band (DNB) images?

Here’s a DNB image of the aurora australis (a.k.a. Southern Lights) over Antarctica, taken on 1 October 2012:

VIIRS DNB image of the aurora australis, taken 00:22 UTC 1 October 2012

Compare this image with the images of the aurora borealis shown back in March 2012. Something doesn’t look right. Far from looking like smooth curtains of light, the aurora (particularly the brightest one) has a jagged appearance, like a set of steps. (This is easier to notice if you click on the image to see it in higher resolution.) This is because the aurora wouldn’t stay still, and we can use this information to estimate the speed it was moving.

The stripes that you see in the image are a caused by the 16 detectors that comprise the DNB which, for various reasons, don’t have exactly the same sensitivity to light. (This condition is given a super-scientific name: “striping”.) The DNB senses light from the Earth by having a constantly rotating mirror reflect light onto these detectors. One rotation of the mirror (particularly the part that occurs within the field of view of the sensor) comprises one scan. Each detector comprises one row of pixels in each scan, each with 742 m x 742 m resolution at nadir. There are 48 scans in one “granule” (the amount of data transmitted in one data file), and it takes ~84 seconds to collect the data that make up one granule. That means it takes ~1.75 seconds per scan.

If you watch that video again, you’ll notice that the aurora can move quite a bit in 2 seconds. Now, let’s zoom in much more closely on one of the aurora elements:

Zoomed-in VIIRS DNB image of an aurora, taken 00:22 UTC 1 October 2012

This image has been rotated relative to the original image, in case you were wondering why it doesn’t seem to match up with the first image. The brightest pixels are where the brightest aurora elements were located. The “steps” (or “shifts” as they are typically called) occur every 16 pixels, which mark out the end of one scan and the beginning of the next. If you count the number of pixels that the brightest aurora elements shifted from one scan to the next, it varies from about 6 to 10 pixels. Assuming a constant resolution of 742 m per pixel along the scan (which isn’t exactly true, the resolution degrades a little bit as you get closer to the edge of the scan but not by much), that means this particular aurora element moved somewhere between ~4.5 and ~7.5 km in ~1.75 seconds from one scan to the next. Doing the math (don’t forget to carry the 1), that comes out to somewhere between 9000 and 15,000 km h^-1 (rounded to account for possible sources of error), which I guess counts as “very fast”. But, it’s not as fast as the coronal mass ejections that create auroras. They have an average speed of 489 km s^-1 (1,760,000 km h^-1)!

So, what looks like an oddity in the VIIRS image, actually contains some interesting scientific information about the speed of an “active aurora“.

But, we’re not done yet. Let’s get back to the striping. Along with “stray light”, it’s one of the few remaining issues in VIIRS imagery. Stray light, which you can see evidence of in the lower right corner of first aurora image, is a particular problem in the DNB. It occurs when sunlight is reflected onto the detectors when the satellite is on the nighttime side of the Earth, but close to the edge of the day/night “terminator“. Our colleagues at Northrup Grumman have been working on a correction to stray light that also reduces the striping. This correction allows for much better viewing of auroras, which have a tendency to occur right where stray light is an issue.

Here is an image of another aurora over Antarctica, taken on 15 September 2012, corrected for stray light and striping:

VIIRS DNB image of the aurora australis over Antarctica, taken 18:56 UTC 15 September 2012. The data used in this image was corrected for stray light and striping by Stephanie Weiss (Northrup Grumman).

This was the night of a new moon, so the only light in the scene (once the stray light is taken out) is the aurora. (OK, there may be some “air glow” and starlight. But, it doesn’t show up on this brightness scale.)

This aurora was a lot less “active” so it looks more like smooth curtains of light. Although, when you zoom in on the brightest swirl in the upper right corner, you can see it did move 3-5 pixels between scans:

VIIRS DNB image of the aurora australis, taken 18:56 UTC 15 September 2012

VIIRS DNB image of the aurora australis over Antarctica, taken 18:56 UTC 15 September 2012. This image has been zoomed in and rotated relative to the previous image of the same aurora. The data used in this image was corrected for stray light and striping by Stephanie Weiss (Northrup Grumman).

This translates to 4000 to 8000 km h^-1, which still counts as “fast” even if it doesn’t count as “very fast”. See, Google was right! Auroras do move anywhere from “fast” to “very fast”. But, now we at least have an estimate to quantify that speed.

And, in case you were wondering, these estimates of the speed of auroras are consistent with earlier observations. According to the book Aurora and Airglow by B. McCormac (1967), the typical speed of auroras is between 0 and 3 km s^-1 (up to 10,800 km h^-1). So, it appears that VIIRS does give a reasonable estimate about the speed of an aurora. We just happened to catch one “typical” aurora and one “faster than typical” aurora.

JPSS (SNPP and NOAA-20)

VIIRS Imagery and Visualization Team Blog

Tag Archives: antarctica

Puttippoq? Aatsuu

Polar Opposites

B-31 and the Pine Island Glacier

Aurora Australis from the Day-Night Band