Tag Archives: ice island

Petermann Glacier Ice Islands: Where are they now?

Two large calving events in 2010 and 2012 reduced the floating part of Petermann Gletscher by 44 km (28 miles) in length, 6 Manhattans (380 km^2) in area, and 42 gigatons in mass. But what’s a gigaton? Writing in The Atlantic Magazine, Julio Friedman states that if we put all people living on earth onto a scale, then we will get half a gigaton. So, Petermann’s two ice island weigh more than eighty times as all humanity combined. As a reminder, this is what the break-ups looked like:

Petermann Gletscher in 2003, 2010, and 2012 from MODIS Terra in rotated co-ordinate system with repeat NASA aircraft overflight tracks flown in 2002, 2003, 2007, and 2010. Thick black line across the glacier near y = -20 km is the grounding line location from Rignot and Steffen (2008).

Petermann Gletscher in 2003, 2010, and 2012 from MODIS Terra in rotated co-ordinate system with repeat NASA aircraft overflight tracks flown in 2002, 2003, 2007, and 2010. Thick black line across the glacier near y = -20 km is the grounding line location from Rignot and Steffen (2008).

It turns out that the smaller 2012 ice island is just as heavy as the 2010 island, because it is much thicker, about 200 m, 600 feet, or half the height of the Empire State Building in Manhattan. These thick and thin islands have since left Petermann Fjord and Nares Strait for more southern climes. The thinnest piece reached Newfoundland in the summer of 2011 where it melted away. Most of the thicker, larger, and heavier ice islands from Petermann and Ryder Glaciers now litter almost the entire eastern seaboard of Canada as the two largest pieces have split, broken, and splintered into many smaller pieces. Each of these still represents an exceptionally large and dangereous piece of ice that can wipe any offshore oil platform off its foundation. Luc Desjardins of the Canadian Ice Service now tracks more than 40 segments, some still bigger than Manhattan, some as small as a football field. The distribution along the 1500 km (1000 miles) of coast is staggering:

RadarSat imagery of eastern Baffin Island (bottom, right), western Greenland (top, right), and Nares Strait with Petermann Fjord (top, left) with pieces of Petermann and Ryder Ice Islands identified. [Credit: Luc Lesjardins, Canadian Ice Service]

RadarSat imagery of eastern Baffin Island (bottom, right), western Greenland (top, right), and Nares Strait with Petermann Fjord (top, left) with pieces of Petermann and Ryder Ice Islands identified as green dots. [Credit: Luc Lesjardins, Canadian Ice Service]

What stands out is that most pieces are close to the coast of Canada. This is expected, because often the ocean moves in ways to balance pressure gradient and Coriolis forces as we live on an earth that rotates once every day around its axis. This force balance holds both in the ocean and the atmosphere. We are all familiar with winds around a low-pressure system such as Hurricane Sandy where the winds move air counter-clockwise around the eye (the center of low pressure). This eye of low pressure in our ocean story is permanently near the center of Baffin Bay. Ocean currents then move water counter-clockwise around this eye. This results in a flow to the south off Canada and a flow to the north off Greenland. On a smaller scale this balance holds also, such as Delaware Bay or Petermann Fjord, but I will not bore you with the details of graduate level physics of fluids in motions … as important as they may be.

So, almost all the ice islands we see in the above imagery will make their way further south towards the Grand Banks off Newfoundland. Some are grounded to the bottom of the shallow coastal ocean and may sit in place for a year, or a month, or until the next high tide will lift the ice off the bottom and move it back into deeper water. Some ice islands will keep moving rapidly, some will further break apart, but none will go away anytime soon. If you want to see some of Petermann’s Ice Islands for yourself, take the ferry from North Sidney, Nova Scotia to Port aux Basques, Newfoundland and Labrador and head for the Great Northern Peninsula. That’s what I hope to do one of the next summers.

Johnson, H., Münchow, A., Falkner, K., & Melling, H. (2011). Ocean circulation and properties in Petermann Fjord, Greenland Journal of Geophysical Research, 116 (C1) DOI: 10.1029/2010JC006519

Münchow, A., & Garvine, R. (1993). Dynamical properties of a buoyancy-driven coastal current Journal of Geophysical Research, 98 (C11) DOI: 10.1029/93JC02112

Rignot, E., & Steffen, K. (2008). Channelized bottom melting and stability of floating ice shelves Geophysical Research Letters, 35 (2) DOI: 10.1029/2007GL031765

Shots of Airborne Lasers at Petermann Gletscher, Greenland

If shots of whiskey make you dizzy, shots of laser stun. NASA stunned me this week, when I discovered that they provide millions such shots of Greenland from which to construct detailed images of the landscape. The shots are free, no age-limit. This is better than the usual remote sensing or photography of “just” brightness. The laser gives us height, and not just the perception of it by shadows and fake angles of illumination, but hard and direct measurements of, well, height above sea level. Have a look at several million such shots of Petermann Gletscher taken in 2010 before the glacier broke to Manhattan-sized pieces:

Petermann Glacier surface elevation from laser shots on Mar.-24, 2010 at the site where the Manhattan-sized ice island formed Aug.-6, 2010. The background shows the same scene at the same time at 250-m resolution from MODIS (see below). Colors along the 350-m wide laser track line show height above sea level in meters.

Petermann Glacier on March 24, 2010 as seen from MODIS satellite at 250-m resolution with two flight tracks along which laser data are collected. The black box shows the site of the figure above. The color figure on the right shows the slope or gradients of the data shown on left. It emphasizes regions where brightness changes fast. Multivariate calculus is useful!

We see two tracks: the one on right (east) has the ice stick more than 20-m above sea level (yellow colors) while about a mile to left (west) the ice’s surface elevation is only 10-m above sea level (light blue). Since the ice is floating and densities of ice and water are known, we can invert this height into an ice thickness. Independent radar measurements from the same track prove that this “hydrostatic” force balance holds, the glacier is indeed floating, so, multiply surface elevation by 10 and you got a good estimate of ice thickness. The dark blue colors of thin ice show meandering rivers and streams, ponds and undulations, as well as a rift or hairline fracture from east to west. This rift is visible both in the right and left track, it is the line along which the glacier will break to form the 2010 ice island. All ice towards the top of this rift has long left the glacier and some of it has hit Newfoundland as seen from the International Space Station by astronaut Ron Garan:

Last remnant of Petermann Ice Island 2010-A as seen from the International Space Station on Aug.-29, 2011 when it was about 3.5 km wide and 3 km long [Photo credit: Ron Garan, NASA]

Both are images of Petermann ice. The photo measures the brightness that hits the lens, but the laser measures both brightness and ice thickness. The laser acts like flash photography: When it is dark, we use a flash to provide the light to make the object “bright.” Now imagine that your camera also measures the time between the flash leaving your camera and brightness from a reflecting object to return it. What you think happens at an instant actually takes time as light travels fast, but not infinitely fast. So you need a very exact clock to measure the distance from your camera to the object. Replace the flash of the camera with a laser, replace the lens of your camera with a light detector and a timer, place the device on a plane, and you got yourself an airborne topographic altimeter. So, what use is there for this besides making pretty and geeky pictures?

The laser documents some of the change in “climate change.” Greenland’s glaciers and ice-sheets are retreating and shrinking. Measuring the surface and bottom of the ice over Greenland with lasers and radars gives ice thickness. The survey lines above were flown in 2002, 2003, 2007, 2010, and 2011. These data are a direct and accurate measure on how much ice is lost or gained at Petermann Gletscher and what is causing it. My bet is on the oceans which in Nares Strait and Petermann Fjord have increased the last 10 years to melt the floating glacier from below.

There is more, but Mia Zapata of the Gits sings hard of “Another Shot of Whiskey.” What a voice …


Johnson, H., Münchow, A., Falkner, K., & Melling, H. (2011). Ocean circulation and properties in Petermann Fjord, Greenland Journal of Geophysical Research, 116 (C1) DOI: 10.1029/2010JC006519

Krabill, W., Abdalati, W., Frederick, E., Manizade, S., Martin, C., Sonntag, J., Swift, R., Thomas, R., & Yungel, J. (2002). Aircraft laser altimetry measurement of elevation changes of the greenland ice sheet: technique and accuracy assessment Journal of Geodynamics, 34 (3-4), 357-376 DOI: 10.1016/S0264-3707(02)00040-6

Münchow, A., Falkner, K., Melling, H., Rabe, B., & Johnson, H. (2011). Ocean Warming of Nares Strait Bottom Waters off Northwest Greenland, 2003–2009 Oceanography, 24 (3), 114-123 DOI: 10.5670/oceanog.2011.62

Thomas, R., Frederick, E., Krabill, W., Manizade, S., & Martin, C. (2009). Recent changes on Greenland outlet glaciers Journal of Glaciology, 55 (189), 147-162 DOI: 10.3189/002214309788608958

Nares Strait Ice Arches and Petermann Ice Island 2012

Arching barriers of ice locked solidly to land are presently closing off Nares Strait for all ice leaving or entering this ocean passage from the Arctic to the North Atlantic Oceans. Gothic cathedrals have flying buttressing to hold them in place while ice arches have buttressing land that keeps them stable. The sea ice becomes land-fast until these ice arches collapse in June or early July. As the ocean under the ice is still moving, generally from north to south, one often finds very thin ice or even open water to the immediate south of these ice arches. Some of these temperature signals let us “see” large ice structures even in the dark of night which in Nares Strait lasts from early October to late March.

Surface temperature in degrees Celsius for Nares Strait on Nov.-10, 2012 from MODIS Terra. Thick ice is blue (cold) while thin ice is red (warm).

Surface temperature at the northern entrances to Nares Strait with the Arctic Ocean to the north.

Surface temperature at the southern entrance of Nares Strait with the North Atlantic Ocean to the south.

Southern entrance of Nares Strait as seen from RADARSAT showing ice arch formation in more spatial detail than MODIS temperatures do. Note the embedded ice island PII-2012 from Petermann Gletscher at the north-eastern edge of the ice arch. [Credit: Luc Desjardins, Canadian Ice Service]

These ice arches usually form in December or January, but this year they form a little earlier than usual. In some years such as 2006/07 or 2009/10 and 2010/11 they did not form at all and thick multi-year ice left the Arctic via a passage that is now closed. This leaves only Fram Strait to the east of Greenland for such export this year.

It appears that the large ice island that broke free from Petermann Gletscher earlier this year provides some stabilizing support to the southern ice arch as it is anchoring its north-eastern corner where it is possibly grounded. The depth of the ice island PII-2012-A1 is about 180 to 200 meters thick. I derived this estimate from both NASA’s Airborne Topographic Mapper (ATM) and the University of Kansas’s Radar Depth Sounder both flown concurrently on a DC-8 plane that surveyed Petermann Gletscher on May-7, 2011 with PII-2012-A1 still attached:

Profile of Petermann Glacier from laser (red) and radar (black) measurements on May-07, 2011. The 2012 break-up is indicated by a spike of the red under-ice topography near km-22. Bottom profiles from laser assume hydrostatic balance of floating ice.

The ATM is a scanning laser that measures the distance from the DC-8 to the surface within 0.2 meters (about 6-7 inches). If I know both the true sea level surface (I do, it’s called the geoid) and if the ice is floating undisturbed, then I can convert the surface elevation into a bottom draft. The red curve outlines the “theoretical” bottom of the glacier. This curve is masked by a thicker black curve that is a radar-derived image of the under-side of the glacier. Nothing theoretical about that one. These radar measurements agree closely with the red curves indicating an almost perfectly balanced floating glacier. This “balance” breaks down at two important points: (1) Near -20 km the glacier bottom is shallower than the red draft curve and it is here that the glacier sits on land as it is not floating. (2) Near +22 km we see a large red spike. This is the location of the 2012 break-up.

So, the 2012 ice-island that is anchoring the ice arch in southern Nares Strait is the piece of the glacier to the right of the red spike and with these data I can now conclude that PII-2012 was 11 km long, 15 km wide, and about 200 m deep. This Manhattan-sized ice-cube weights about 30 gigatons (10^12 kg), but “… that doesn’t mean much — who goes to the store and buys a gigaton of carrots? For a sense of perspective, a gigaton is about twice the mass of all people on earth …” [James Fallows writing for The Atlantic]. Hence this little ice-cube weights 50 times as much as do all people living on earth today. Incidentally, it is also the amount of CO2 that all humanity adds each year to the atmosphere. Coincidence.

Front of Petermann Glacier Aug.-11, 2012. View is from a small side-glacier towards the south-east across Petermann Fjord with Petermann Gletscher to the left (east). [Photo Credit: Erin Clark, Canadian Coast Guard Ship Henry Larsen]

Petermann and Ryder Glacier Ice Island

Ice island from 2010 and 2012 calvings litter Nares Strait and northern Baffin Island, Canada. All these glacier fragments originate from Petermann and Ryder Gletscher in north-west Greenland. The image below is a composite that Luc Desjardins of the Canadian Ice Service compiled from RadarSat imagery. He painstakingly identified 25 segments in these imagery.

Ice Islands and fragments from Petermann and Ryder glacier 2010 and 2012 calvings. [Credit: Luc Desjardings, Canadian Ice Service]

The largest piece is PII-2012-A1 and it covers an area a little less than 2 Manhattans (100 km^2). We see it in Kane Basin for several weeks now as it pivots back and forth with the tides around the point where it is stuck to the bottom of the ocean. The second largest piece is RII-2012 roughly half the size of Manhattan (33 km^2) and it originates from Ryder Gletscher which is to the north by north-east of Petermann Gletscher. Trudy Wohleben identified this piece when it was entered Nares Strait from the north about 4 weeks ago and together we traced it back to Ryder Gletscher where it had lingered for several years. RII-2012 is now moving rapidly south and is about exit Nares Strait to enter Baffin Bay:

Two of these ice island send their position several time each day with the data made available at for PII-2010-B-a (9 km^2) and for PII-2012-A2 (13 km^2). The last piece broke off from Petermann on July 16, 2012 and it entered Nares Strait in August when we passed it during our explorations of Petermann Fjord on Aug. 10/11, 2012 aboard the CCGS Henry Larsen:

Canadian Coast Guard Ship Henry Larsen at the entrance to Petermann Fjord on Aug.-10, 2012. The ice island PII-2012 is in the background with puddles on sea ice in the foreground. Polaris Bay, Greenland is in the far back. [Photo Credit: CCGS Henry Larsen and Jo Poole.]

Ice Thickness in Nares Strait 2008 and 2009

[Editor’s Note: Undergraduate Julie Jones of the University of Delaware summarizes her work that was supervised by Helga Huntley as part of an NSF-funded summer internship.]

Three years ago in 2009 Andreas Muenchow left from Delaware for Greenland with students Pat Ryan and Berit Rabe to recover instruments that recorded salinity, temperature, current velocities, and ice thickness in Nares Strait since 2007.  This summer, I used those observations to estimate ice thickness for April through June in 2008 and compare them to estimates for the same spring period in 2009.  An ice bridge had formed in 2008 but not in 2009.  Working as a group, we wanted to investigate the effect of ice arches on the ice thickness.  Allison Einolf, another summer intern who focused on ocean currents during the same time periods and Andreas produced these maps that introduce the study area, spatial ice cover, and mean ocean currents:


Nares Strait MODIS satellite imagery of the study area and ice arch April 21, 2008. Red dots are instrument locations. Arrows show current velocities.


Nares Strait MODIS satellite imagery of the study area and ice arch April 22, 2009. Red dots are instrument locations. Arrows show current velocities. Note the lack of the southern ice arch, but the presence of one north of the study area.

I used Matlab for most of the data processing, more specifically the Ice Profiling Sonar (IPS) Processing Toolbox for Matlab provided by the manufacturer of the instrument that collected the data: ASL Environmental Sciences, Inc. First I transformed the data from the IPS instrument into a range time series.  I then manually “despiked” the data, taking out any data points that were likely due to bubbles or fish within the acoustic path from the sensor system to the ice above and back.  In a second step I wrote a function using sound speed data from Andreas, atmospheric pressure from Dr. Samelson at Oregon State University, and pressure (depth) data from the IPS instrument to get a time series of the thickness of the ice.  In a third step I applied a Lanczos raised cosine filter that was taught as part of a 2012 Summer Intern Page Workshop. Hence I finally had some nicely filtered data for the periods of the April-June of 2008 and 2009.

Now the results:  Just as we expected, there was much thicker ice in the 2008 spring with a southern ice arch present than there was in the spring of 2009 when no such ice arch was present:

Histogram for April – June 2008 ice. There is a peak at 3 meters, with almost 25% of the ice that thick.

Histogram for April – June 2009 ice. The ice does not get thicker then 2 meters with most of the ice thinner than one meter.

The histograms show thicker ice in 2008, about 2-6 meters on average and with some ice even reaching 10 meters.  In 2009, the ice doesn’t get thicker than 2 meters with most of the ice being thinner than 1 meter.  More specifically, the mean ice thickness for April – June 2008 (2009) is 3.8 (0.58) meters with a standard deviation of 1.8 (0.29) meters.  This further shows that there was thicker ice in 2008 than there was in 2009.  I attributed the cause for the thin 2009 ice to ice flowing freely through Nares Strait all winter and spring as no ice arch in the south blocked such flow.  The ice, thus, did not spend enough time in the high Arctic to thicken.

I noticed something else in my histograms when the 2008 ice bridge collapsed.

April 2008 ice thickness

May 2008 ice thickness

June 2008 Ice Thickness

The monthly histograms show that the ice in April and May is thicker than the ice in June.  We know that the 2008 ice bridge collapsed near June 6th, so it is interesting and it makes a lot of sense that the ice in June would be thinner than the ice two months earlier.

The mean ice thickness for April 2008 was 4.6 meters with a standard deviation of 2.40 meters.  In May 2008 the mean ice thickness was 3.5 meters with a standard deviation of 1.40 meters.  Lastly, in June the mean ice thickness was 3.5 meters with a standard deviation of 1.30 meters.  The ice thickness decreased after April and the variability decreases in June, which helps detect the bridge collapse in the data.

Lastly here are the filtered time series of April – June of 2008 and 2009.

Filtered time series for April – June 2008

Filtered time Series for April – June 2009 with the same scale as 2008 (above figure)

Filtered time series for April – June 2009 with a different scale to see the variability over time more clearly.

Hopefully we can see more interesting and exciting results from the instruments that the Nares Strait team picked up this summer even though they were hit hard by the 2010 Petermann Ice Island!

Two Ice Profiling Sonars (IPS) aboard the CCGS Henry Larsen in Aug.-2012. The protective stainless steel frame was bent by the 2010 ice island that hit both instruments in Sept.-2010. [Photo Credit: Andreas Muenchow]