IceTrek Research Updates

Final Update: IceTrek Findings Provide Clues to Ice Shelf Disintegration
23 September, 2008

Two years after the launch of their expedition, the IceTrek team has completed their analysis of field and satellite data (for details and photos of the expedition). Detailed results from the IceTrek mission were recently published in the Journal of Glaciology. In this final update, Principal Investigator Ted Scambos explains what the team learned from their observations of icebergs A22A and Amigosberg.

After we finished the field work, the difficult part of the project began: analyzing the data we had collected. The Automated Met-Ice-Geophysics Observation System (AMIGOS) stations on icebergs A22A and AMIGOSberg provided us with daily photos of the icebergs, as well as temperature and location measurements. But in order to create a more complete picture of the icebergs' shrinking, thinning, and breakup process, we combined the AMIGOS data with satellite data and photos taken by astronauts under the NASA International Space Station (ISS) Astronaut Photography program.

The combination of satellite imagery, photos from the space station, and data we collected from the AMIGOS stations helped us understand the factors that ultimately led to the breakup.

Tracking the icebergs
After our work on the icebergs, AMIGOSberg and A22A floated north into warmer waters, eventually drifting thousands of miles from where we had outfitted them with sensors. The AMIGOS systems initially worked well. Both units sent back weather, position, and other data, along with six small images, each day of operation. But as winter came, heavy snowfall buried the AMIGOSberg station, covering its solar panels. Without solar power, the batteries ran down and the station shut off. However, the A22A station stayed active through the winter.

We followed the iceberg trajectories using a combination of satellite data from the NASA Moderate Resolution Imaging Spectrometer (MODIS) and Quick Scatterometer (QuikSCAT), and the AMIGOS on-board global positioning system (GPS). The counterclockwise motion of the Weddell Sea carried the two icebergs out from behind the protective arm of the Antarctic Peninsula into the warmer Southern Ocean and far southern Atlantic Ocean. Figure 1 shows the iceberg paths. Once the icebergs drifted into warmer oceans and air, they began to rapidly melt and break apart. Figure 2 shows the area loss rate for Amigosberg, which accelerated once the iceberg hit open ocean.

iceberg pathsFigure 1. Map showing the drift tracks of icebergs studied. A22A's path is marked by the green line, and AMIGOSberg's track is in pink. —Image courtesy Ted Scambos, National Snow and Ice Data Center

Area loss ratesFigure 2. Area loss rates for AMIGOSberg. Red line represents loss rate; black line with boxes represents area. As AMIGOSberg reached open water, the loss rate increased, and once the firn on top of the iceberg became saturated with meltwater, breakup accelerated significantly. —Image courtesy Ted Scambos, National Snow and Ice Data Center

Surveying the shrinking
At first, the icebergs shrank gradually in a process we refer to as edge wasting, in which warmer water erodes the iceberg at the waterline, causing large chunks of ice above the water to spall, or flake off. This process leaves a border of submerged ice, called a bench, surrounding the iceberg. Eventually, large ice blocks break away from the bench as they try to float up to replace the lost snow and firn. A22A stopped shrinking briefly as winter sea ice again surrounded the iceberg, indicating that warm surface water was a key part of the edge wasting process.

We observed edge wasting through photos provided by astronauts on the ISS. In Figure 3, the underwater portion of the iceberg, where the top has spalled off, appears as a blue border around AMIGOSberg.

A camera on the A22A AMIGOS station gave us another vantage point from which to monitor the breakup process. The camera took successive images of flags planted in the ice (Figure 4). As the iceberg reached open water, the flags began to fall off, one by one. In the winter, sea ice enveloped A22A and halted the disintegration. When spring returned and open water surrounded the iceberg, the flags began to drop off again.

astronaut photoFigure 3. Astronaut photos from the International Space Station were a key addition to the data. In this image of A22A, the faint blue border around the ice edge implies a submerged "bench" feature, indicating that the ice above had broken off due to edge wasting. (High resolution not available) —Image courtesy Ted Scambos, National Snow and Ice Data Center

flag line photo with insetFigure 4. Scientists used flag lines from the AMIGOS station to the iceberg edge to monitor its disintegration. (High resolution not available) —Image courtesy Ted Scambos, National Snow and Ice Data Center

The rapid disintegration we saw as the icebergs reached open water gave us a number of clues to the connections between warm surface temperatures and rapid breakup. As Amigosberg drifted north (a little faster than A22A), warmer air temperatures melted snow on the iceberg surface, uncovering the station’s solar panels and recharging the batteries. In early October, we got a fantastic surprise when the AMIGOSberg station started to transmit data again. 

As warm temperatures continued, water from the melt collected under the snow surface. At this time, the AMIGOS unit on the berg took a series of images of its base. We had set up this photo angle to keep track of the wires and solar panels, and we never expected it to provide scientific results. But on November 7, we saw a pool of water in the pit that we had dug for battery units, showing the extent of surface melt on the iceberg (Figure 5).

After November 7, AMIGOSberg began to rapidly disintegrate (Figure 6). Warm slush and perhaps some surface fractures caused the AMIGOSberg station to fall over a few days later, but it kept transmitting until November 21, 2006. On that day, GPS data showed it teetering on the edge of the crumbling iceberg.

A22A survived for several more months, shrinking gradually by the edge wasting process, but never reaching the rapid disintegration state that AMIGOSberg did. We finally lost contact with the A22A unit in January, 2007, as it too fell into the sea. By August, 2007, A22A had shrunk so much that we could no longer track it by satellite.

photo time seriesFigure 5. This series of photos from the AMIGOSberg show the accumulation of meltwater that preceded the rapid breakup of the iceberg. Scientists think that water on the surface of sea ice can weaken it and hasten disintegration. —Image courtesy Ted Scambos, National Snow and Ice Data Center

satellite image time seriesFigure 6. MODIS images show the rapid breakup of AMIGOSberg in November of 2006. —Image courtesy Ted Scambos, National Snow and Ice Data Center

A model for climate change
Watching AMIGOSberg and A22A break up provided us with a model of how ice shelves, large plates of ice still attached to the Antarctic continent, can collapse. Like ice shelves, icebergs float on the ocean surface, but unlike the ice shelves, they are free to roam. As they drift from the poles into warmer water, they essentially undergo rapid climate change. Following the icebergs, and watching how they evolve in various climates, can tell us a lot about ice shelves might respond to a warming environment

When the ice shelves break up, the feeder glaciers for the former shelf accelerate and begin to put more ice into the ocean, raising sea level. Knowing that a breakup may be on the way is important. If we know that the shelves are near the point of collapse, we can predict what will happen to the glaciers surrounding the shelf, and therefore, how sea level will change.

Scambos is planning another Antarctic field mission monitoring both ice shelf breakup and the glacial acceleration that follows. Seventeen researchers and staff will conduct a series of studies at the Larsen B embayment (site of the largest ice shelf disintegration event in 2002). Scambos and a research team will bring new, upgraded AMIGOS units to the glaciers and the remaining ice shelf for further study. For more information, see the project announcement, Effects of Climate Change to be Investigated During IPY. To access the recently published paper on IceTrek, visit the Journal of Glaciology Web site.

From a Distance: the AMIGOSberg and A22A Stations Send Data Back to Civilization
3 April, 2006

The IceTrek team returned to their various homes after an intensive field expedition (see Mission Log for more details). Principal Investigator Ted Scambos writes this update on the expedition wrap-up and the current status of the floating Antarctic iceberg science stations.

Even on the commercial flight back, the Icetrek team witnessed some amazing scenes related to climate change and the icy parts of the world. Relatively clear skies over southern Patagonia allowed me to get a couple of pictures of two very famous and active glaciers that flow into western Argentina: Perito Moreno Glacier and Uppsala Glacier. You may recall that the team visited Perito Moreno (see the 4 February entry in the Mission Log).

Figure 1 documents the aftermath of an event that occurred March 13, 2006: the breakdown of the ice dam on Perito Moreno Glacier that had previously separated Lago Argentino into two portions. This caused the lake levels to drop on the southern portion of the lake (right side Figure 1) as it flowed into, and filled, the northern side (left side of Figure 1). Uppsala Glacier is retreating and its ice flows rapidly, as image mapping from Landsat and ASTER satellite images have shown. The image from the airplane shows the ice has retreated about 1 km since 2001 (see Figure 2). Pedro Skvarca will continue to research the changes in both glaciers.

As for the two auto-camera AMIGOS stations, one on AMIGOSberg and one on iceberg A22A, they are functioning well and collecting images daily. An image of the A22A tower for April 3 is shown in Figure 3; the IceTrek team deployed a separate camera opposite the main tower so that we can monitor the state of the tower as the iceberg drifts north and begins to break down. Solar panels supply energy to the tower and can be seen as dark rectangles in the image.

The A22A setup had an extra instrument attached, a radio-echo-sounder for measuring ice thickness. The device is very simple in concept: a transmitter uses a cascade release of energy from capacitors to send out a very brief but very powerful radio pulse at about 10 MHz wavelength. A receiver begins to listen as soon as the loud pulse is triggered. It sees first the direct pulse through the air and then the reflected pulse from the bottom of the iceberg—if we're lucky.

A typical radar profile from A22A is shown in Figure 4. Unfortunately, the metal structure of the tower may be causing some problems, creating a lot of reflections in the early pulse. The return pulse from the base of the ice may actually be obscured by the loud direct pulse. The earlier test on the Chip/Tempanito iceberg is shown for comparison (Figure 5); at Chip, there were no large metal structures around. We're planning to analyze the radio pulses from A22A extensively to determine if the data can be filtered and made useable.

Lago Argentino and Perito Moreno GlaciersFigure 1. Image of Lago Argentino and Perito Moreno, previously split into two separate bodies, now one body of water following the breakdown of an ice dam on March 13, 2006.

Uppsala GlacierFigure 2. Image of Uppsala Glacier from plane, showing 1 km in retreat since 2001.

A22A towerFigure 3. Image of A22A tower taken by separate nearby camera.

A22A radar graphFigure 4. A22A radar profile showing disturbance, possibly from the metal tower itself.

Chip-Tempanito comparisonFigure 5. Comparison image of Chip/Tempanito iceberg radar profile, showing no disturbance. The IceTrek team did not deploy a station on Chip/Tempanito, so this profile was taken during their reconnaissance mission.

"Chip" and AMIGOSberg Deployments: What We've Learned so Far
10 March, 2006

Following successful missions to both the "Chip," or "Tempanito," iceberg and AMIGOSberg (see Mission Logs for more information), Principal Investigator Ted Scambos sent this message with an update on the research the team has conducted, so far, and how it fits in with current scientific understanding of icebergs.

On-the-Ground Research
I would like to explain in a bit more detail what kind of things we are seeing on the icebergs and what we hope to learn from them. With Tempanito and AMIGOSberg visits under our belt, we have a lot of observations that allow us to build up some ideas that we can continue to test with the satellite and on-site AMIGOS stations.

AMIGOS is an acronym; it stands for Automated Met-Ice-Geophysics Observing Station. The idea is that these stations can supplement our satellite data with high-resolution, continuous observations of various parameters on ice sheets, glaciers, and icebergs. In this case, we're setting up two stations (AMIGOSberg, which is done, and A22A, which we hope to attempt soon).

But in addition to the data that the stations will take, this expedition gives us the opportunity to do on-the-ground research. In the flights we take to the bergs and back and in our time hiking around on the surface, we're making a lot of observations of the ice, trying to figure out what is going on and how it might relate to global warming on ice shelves and ice sheets.

Seeing History in Layers
We can tell a lot about the history of the ice in the iceberg by looking at the edge (see Figure 1). Layering shows us the yearly history of snowfall; but in warm summers, this snow melts, leaving blue layers of refrozen ice. This formation of melt on the surface is thought to be a big part of the process of ice shelf disintegration. We can see that AMIGOSberg, which came from the southern Larsen C (about 450km to the south of where we met up with it), seems to show an increase in the intensity of melting in the past few years.

The last year or two would be after the berg broke free, and drifted north. But the top 8 meters of the berg represent about the last 15 years of time. The face is about 25 meters tall, so we can look back in time by looking at the layers in the ice. The layers give us an indication that climate is warming well to the south on the Peninsula.

Warping of an Iceberg
We also wanted to look at the structure of the edge of the iceberg. Models predict, and satellite measurements show, that the iceberg edge is flexed. In cold water, the main force is a downward bending of the ice. It's a subtle effect, but think of it as the same kind of force you get on the top of a cork that is pushed halfway into a bottle. The force of buoyancy is 'squeezing' the submerged sides of the ice, warping the iceberg. This effect is so subtle, you really have to be at the edge, looking at it in comparison to the water horizon, to see it. And, ideally, you would have to measure it with GPS.

On AMIGOSberg, we didn't have time to do a precise GPS profile, but we used the line of flags that we were
setting out, as we walked, as a guide to measure the structure of the iceberg. As flags disappeared or reappeared behind us in our line of vision, we knew we were going up or down. We found that AMIGOSberg is flexed like a cork, humped with a downward slope of as much as 10 meters--more than we measured with the satellite. See Figure 2; note that the berg edge is sloped over Ted's left shoulder, but the waterline is flat in the other small berg in the distance.

On the other hand, in warmer water (by that I mean without sea ice on the surface--the water is still darned cold!) another effect takes over, and the edge profile is sloping upward. We're expecting to eventually see this on A22A, the edges of which are currently sticking out from the iceberg. What happens in warm water is that wave action erodes the waterline rapidly, and the top part of the berg breaks away. But below the surface, the erosion is not so fast, and so there is a kind of 'bench' formed in the ice. This 'bench' wants to float up, and so it lifts the ice at the edge upward. We got a great photograph of an iceberg that shows this process (Figure 3). It has recently tilted, showing what used to be underwater.

The other thing you'll notice is that the top of that small tilted iceberg is covered in cracks. This seems to happen to all the small icebergs that make it to open water. We currently think that this may be the the effect of the very gentle ocean swell that begins to appear near the boundary of where sea ice occurs. We think the swell is flexing these icebergs. But south of the boundary of where sea ice occurs, there is no ocean swell. There, it's always calm water because the sea ice on the surface of the water absorbs the ocean's wave energy by bumping into each other.

Push and Pull of the Tides
One last thing we're looking at is how icebergs drift, and what forces are pushing them around. In a word: it's tides! This was a surprise to scientists until a few years ago. Most scientists previously thought that a combination of winds and currents push the icebergs. However, it turns out that iceberg motion is a bit like having a ball on a flat board, and then tilting the board around. The tilting represents the coming and going of the Earth's tides, resulting in a very very gentle slope on the ocean--a slope on the order of only 1 meter height in 1,000 km. The icebergs simply slide "down-hill." Currents and winds have a secondary, smaller effect.

In the last picture, we've plotted the first 76 hours of iceberg motion on a zoomed satellite image of AMIGOSberg (Figure 4). The GPS sensor, located on the tower, is at the site labeled 'acamp' on the image at the time of the photograph. From the motion, shown with a thin line, we can see that the motion "loops" and shows a pattern caused by tides.

So, even though our Mission Log entries may make it sound like we're merely constructing towers to take data that we'll analyze later on, each of our outings during the expedition provides us with crucial on-the-ground observational research. This hands-on research helps us add to the current scientific understanding of how icebergs work.

Snowfall layersFigure 1. Layering shows the yearly history of snowfall. In warm summers, the snow melts, leaving blue layers of refrozen ice. Photograph courtesy Marcelo Gurruchaga.

Iceberg edgeFigure 2. The iceberg edge is sloped over Ted's left shoulder, but the waterline is flat in the other small iceberg in the distance. This slope shows that the iceberg is being flexed by the forces around it.

Iceberg with benchesFigure 3. Warmer water washes against the iceberg and causes benches to form. The benches push the iceberg edges upwards; here, a bench floats above the water line and forces ice that used to be underwater into view.

Figure 4. This satellite image shows the first 76 hours of AMIGOSberg's motion after we deployed the GPS and weather station. The GPS, which is sending back data on the iceberg's current position, is located at "acamp." The thin line indicates where the GPS station, in its fixed position on the iceberg, has moved because of tidal forces.