Stephenson Glacier Retreat and Lagoon Development, Heard Island

The Australian Antarctic Division manages Heard Island Island and has undertaken a project documenting changes in the environment on the island. One aspect noted has been the change in glaciers. The Allison, Brown and Stephenson Glacier have all retreated substantially since 1947 when the first good maps of their terminus are available. Fourteen Men by Arthur Scholes (1952) documents a year spent by fourteen men of the Australian National Antarctic Research Expedition that documented the particularly stormy, inclement weather of the region. Their visit to the glacier noted that they could not skirt past the glacier along the coast. After crossing Stephenson Glacier they visited an old seal camp and counted 16,000 seals in the area. It is a rich area for wildlife, that should benefit from the lagoon formation overall.  Thost and Truffer (2008) noted a 29% reduction in area of the Brown Glacier from 1947-2003. They also observed that the volcano Big Ben that the glaciers all drain from has shown no sign of changing geothermal output to cause the melting and that a 1 C warming has occurred over the same time period.
Hear Island Map from the Australian Antarctic Division

Stephenson Glacier extends 8-9 km down the eastern side of Big Ben. In 1947 it spread out into a piedmont lobe that was 3 km wide and extended to the ocean in two separate lobes around Elephant Spit. A picture from the Australian Antarctic Division taken in 1947 shows the glacier reaching the ocean and then in 2004 from the same location. Kiernan and McConnell (2002) identified an order of magnitude increase in the rate of ice loss from Stephenson Glacier after 1987. Retreat from the late 19th century to 1955 had been limited. As Kiernan and McConnell observed retreat began that by 1971 the glacier had retreated 1 km from the south coast and several hundred meters from the northern side of the spit. This retreat by 1980 caused the formation of Stephenson Lagoon and by 1987 Doppler Lagoon had formed as well. After 2000 the two lagoons have joined. The first image below shows the terminus locations over the last 60 years from the Australian Antarctic Division 1947-2008.

Here we examine a series of Landsat images from 2001 to 2013 to update the response of this glacier.
In 2001 the glacier has two separate termini, pink arrows,in two different lagoons, Doppler to the south and Stephenson to the east. There are numerous icebergs in Doppler lagoon but none in Stephenson Lagoon, indicating the retreat is underway. In 2008 the two lagoons are well joined, icebergs are even more numerous obscuring in this view the true location of the terminus, orange arrows. By 2010 the glacier has retreated from the main basin of the lagoon, and is at red arrow, and the lagoon is free of ice for the first time in several hundred if not several millenium. In 2013 the glacier has retreated into a narrower valley that feeds into Stephenson Lagoon. The northern arm of the glacier experienced a 1.7 km retreat from 2001 to 2013 and the southern arm as 3.4 km retreat. The period of rapid retreat due to calving of icebergs into the lagoon is over and the retreat rate will now be slower. There is still rapid glacier flow toward the terminus as indicated by extensive crevassing.  The overall glacier slope is steep and accumulation rates high, which would also generate rapid glacier flow.

The AAD has a number of images in their gallery of Heard Island glaciers including Stephenson Glacier. The climate station at Atlas Cove indicates a 1 C temperature rise in the last 60 years.  The AAD will also certainly be looking at how this new lagoon impacts the local seal and penguin communities. The map above indicates the importance of Stephenson Lagoon for wildlife, king penguins and cormorants are noted by AAD.  The retreat of this glacier follows the pattern of glacier retreat at other glaciers on islands in the circum-Antarctic region Cook Ice Cap, Kerguelen Island , Purvis Glacier, South Georgia and Neumayer Galcier, South Georgia.

stephenson map
Map of retreat from superimposed on Google Earth image.

stephenson 2001
2001 Landsat image

stephenson 2008
2008 Landsat image

stephenson 2010
2010 Landsat image

stephenson 2013
2013 Landsat image


Spørteggbreen Separation and Retreat, Norway

Spørteggbreen is a small ice cap just east of Jostdalsbreen, Norway’s largest ice cap.  The Norway Glacier atlas chronicles the area of the 10 glaciers that comprise the ice cap at 23.8 square kilometers in 2006 (Andreassen and Winsvold, 2012). The Norwegian Water Resources and Energy Directorate (NVE) has an excellent glacier monitoring program that monitors 10 glaciers that drain Jostedalsbreen from 2008-2013 the net change was retreat on all 10. This is exemplified by Tunsbergdalsbreen. In 2014, 33 of 38 glaciers observed in Norway retreated and 5 did not change. Laumann and Nesje (2014) examined Spørteggbreen during the Holocene.  They noted the glacier disappeared from 7300 years before present and reformed approximately 5400 years before present.  Their model simulates minor changes from 2011 to 2050. Here we focus on changes to the southwest side of this ice cap using Landsat imagery from 1988-2014 and implications for the near future.

sporteggbreen 2010Norway Glacier atlas image looking southwest toward Jostdalsbreen.  The foreground is the main Sporteggbreen, the new lake is in midground and then the separated southwest extension.  There is no retained snowcover evident on the southwest extension and it is thin.

In 1988 at the southwest side of Spørteggbreen the main ice cap joins the southwest extension of the ice cap in a small lake at the red arrow. At the purple and yellow arrow are small peripheral proglacial lakes. The green arrow indicates a narrow section in the southwest extension. In August 1998 there is substantially more snow and lake ice, which makes identification of change difficult. In 2010 the southwest extension is well separated by a lake from the main ice cap, red arrow. This lake is m across, with the glaciers reaching the shore at both ends. The lakes at the purple and yellow arrow have also expanded. The width of the narrow section of the southwest extension has narrowed from 600 m to 300 m. In 2014 the main Spørteggbreen Ice Cap no longer reaches the eastern shore of the lake, red arrow. The lake is 750 m across. The lake at the purple arrow is now equivalent to the glacier that ends in the lake and is m long. The width of the glacier at the narrow section is down to 200 m. The most noteworthy item is that in this mid-September image the ice cap has 25% of its area still covered in snowpack, note the areas marked with pink letter A. The other areas are either firn from previous years or bare glacier ice.  By the end of the melt season at the end of the month this will likely be closer to 20% . Remember a glacier typically needs 55-65 % snowcover to be have an equilibrium balance.  This will lead to a large negative balance and thinning right to the top of the ice cap. This was the case in 2013 as well. Hence, the Spørteggbreen Ice Cap is currently experiencing significant volume and area losses. The retreat is not as large in terms of distance as on nearby Tunsbergdalsbreen.

sporteggbreen 1988a













1988 Landsat image.

sporteggbreen 1998

1998 Landsat image

sporteggbreen 2010a

2010 Landsat image

sporteggbreen 2014a
2014 Landsat image

Anderson Glacier, Olympic Mountains, Washington Disappears

Anderson Glacier was the headwaters of the Quinault River in the Olympic Mountains of Washington. A century ago the glacier was 2 km long, and a half kilometer wide. Retreat of this glacier in the first half of the 20th century exposed a new alpine lake as the glacier retreated 1 kilometer. From 1950-1980 the glacier diminished slowly. From 1959 to 1990 the glacier thinned and retreated from the shore of the lake trapped behind the Little Ice Age moraine. The 1959 picture below was donated to me by Austin Post. Since 1990 the glacier has begun to shrink rapidly. The Google Earth image from 1990, indicates Anderson Glacier has retreated 200 m from the 1959 terminus position near the lake shore, green arrow to the 1990 position, pink arrow. The red arrow indicates a future location of a bedrock outcrop.
1959 Austin Post image
anderson Glacier 1990
1990 Google Earth image

Investigating this glacier in 1992 we measured its area at 0.38 square kilometers, down from 1.15 square kilometers a century before. Ten years later the glacier had diminished to 0.28 square kilometers, but had thinned even more, leaving it poised for a spectacular change, over the next five years. Large outcrops of rock appeared beginning in 2003 and further exposed in 2005 and 2007 in the middle of the glacier. Note the outcrops in the 2007 image from Kathy Chrestensen. The 2009 Google Earth image indicates the 1990 terminus position, pink arrow, and the fact that there is no longer a ribbon of snow that is even 50 m wide. The snow patches have insufficient size or thickness to be classified as a glacier.  The largest outcrop at the red arrow had been beneath the ice in 1990, giving a scale to the thinning.  The glacier at this point no longer exists. In 2014 an Eric Hovden image indicates some seasonal snow in the basin, but the thin ribbon of snow has numerous holes in it as well, indicating the thin nature of the remaining snow patches, with a month left in the 2014 melt season.

Kathy Chrestensen Image

anderson glacier 2009
2009 Google Earth Image

anderson glacier 2014
2014 Eric Hovden image.

This glacier had become a series of small disconnected relict glacier ice patches in 2005 and by 2009 had disappeared. It is not the only glacier that is disappearing, which has led to a visual model for forecasting glacier survival (Pelto, 2010). The key is observed retreat of the margin of the upper portion of the glacier and emerging rock outcrops in the upper part of the glacier where snow should accumulate and be retained through the melt season. If a glacier does not have a significant persistent accumulation zone it cannot survive. Anderson Glacier was not the only glacier feeding the Quinault River, all the others are retreating as well. The result of this glacier retreat is reduced late summer and early fall streamflow, impacting salmon runs at that time of the year. This is primarily the fall Coho, Chum and Chinook salmon and Steelhead summer run. During the spring and early summer runoff increases as snowmelt still occurs, but is not retained in the glacier system.To get a sense of the special nature of this area Out of the Mist is an excellent start

Nizina Glacier Retreat, Lake Formation, Alaska

If you have heard of Nizina Glacier in the Wrangell Mountains of Alaska it is probably because you have contemplated a float trip down the Nizina River from Nizina Lake.  In 1990 there was no lake, since 2000 the lake has provided a good location for float planes to land.  In 2014 the lake has reached a new maximum in size and minimum in icebergs on its surface.  Here we examine Landsat imagery form 1990-2014 to identify changes in the Nizina Glacier.  The main tributary of the Nizina Glacier is Regal Glacier indicated by the dark blue flow arrows.  The light blue flow arrows are from the Rohn Glacier tributary that no longer reaches the terminus area.

nizina ge

Google Earth image

In each image the yellow arrow marks the 1990 terminus, red arrow the 2014 terminus location and pink arrows the summer snowline.  In 1990 the glacier had narrow sections of fringing lake evident, though the glacier reached the southern shore of the developing lake at yellow arrow.  By 1995 the lake had developed to a width of 100-300 m fringing the shoreline around the terminus of Nizina Glacier.  In 1999 the main lake has developed and is 1.6 km long and 1.3 km wide though it is still largely filled with icebergs.  In 2013 there are a few icebergs left in the lake.  In August, 2014 the lake is free of icebergs for the first time, which does mean more will not form. The lake is 1.4 km wide and 2.3 km long.  The glacier has retreated 2.1 km from 1990 to 2014, a rate of 150 m per year, red arrow marks 2014 terminus. A close up view of the terminus in Google Earth from 2012 indicates numerous icebergs but also substantial rifts, green arrows, that will lead to further iceberg production and retreat.  The snowline in this late July or early August images is typically at 1800-1900 m, pink arrow, with a month still left in the melt season. The retreat of this glacier is similar to that of glaciers in the Talkeetna Range to the west South Sheep Glacier and Sovereign Glacier and Valdez Glacier to the south.

nizina glacier 1990

1990 Landsat image


nizina glacier 1995

nizina glacier 19991995 Landsat image

nizina glacier 1999

1999 Landsat image

nizina glacier 2013

2013 Landsat image

nizina glacier 2014

2014 Landsat image

nizina lake 2012

Google Earth image 2012


Roncagli Glacier Retreat, Tierra del Fuego, Chile

The Cordillera Darwin in Tierra Del Fuego, Chile is a remote area that is notorious for stormy, cloudy weather that makes for only a few good satellite images. Roncagli(Alemania) Glacier is the focus of this post and is an update to a previous post. The glacier has a terminus adjacent to the Beagle Channe(BC) and a secondary terminus in Lago Martinic (LM), 5 km upglacier. Velocity profiles by Melkonian et al (2013) indicate the highest velocities directed toward the LM terminus, making this the primary terminus. They also found that the glacier thinned by 5-10 m along most of its length from 2000-2011. Here we examine Landsat imagery from 1997 to 2014.

Roncagli ge
Googel Earth image
In 1997 the BC terminus at the pink arrow is at a narrowing of the valley. The LM terminus is at the yellow arrow with two primary glacier branches encircling the nunatak at the red arrow. In 2000 the terminus positions are relatively unchanged with the LM terminus actively releasing icebergs into Lago Martinic. Upglacier a single area of bedrock is emergent through the glacier, purple arrow. In 2001 the BC terminus remains unchanged, the water level in LM has declined exposing more bare rock surfaces around the LM terminus. By 2008 the LM terminus has separated, both still ending in the lake, the lake again is at a full stage on the date of the imagery. The lake experienced periodic filling and draining episodes during the 1997-2008 period. There are now two upglacier areas with exposed bedrock now. By 2014 the BC terminus has retreated 1 km along the southeastern margin and 200 m along the northwest side. This retreat from the pinning point that restricted calving at the pink arrow, suggests further retreat will occur in the near future. Lago Martinic has largely drained. The LM terminus has separated into two tongues and the former nunatak is no longer surrounded by glacier ice, red arrow. The retreat at LM terminus is 1500 m on the west side, orange arrow, and 800-1000 m on the east side. Upglacier both areas of bedrock that are emergent are expanding, purple arrows, indicating the thinning observed by Melkonian et al. (2013). The continued upglacier thinning indicates reduced flux to the terminus and continued retreat. The degree to which Lago Martinic can refill is uncertain, MODIS imagery from late 2014 shows the lake is still not filled. I have not seen imagery indicating even a nearly full lake in the 2011-2014 period. The rate of retreat is less than on Marinelli Glacier to the north or Glaciar Steffen.
racongli 1997
1997 Landsat image
racongli 2000
2000 Landsat image
racongli 2001
2001 Landsat image

roncagli 2008
2008 Landsat image
racongli 2014
2014 Landsat image

Marinelli Glacier Retreat, Chile

Marinelli Glacier, Chile is the largest glacier of the Cordillera Darwin Icefield. This ice cap is in Tierra del Fuego, a region famous for cloudy, stormy weather. Fernandez et al. (2011) indicate that rapid retreat particularly since 1945 has led to high erosion and sedimentation rates. They also provide an excellent diagram of the glacier from three time periods. The glacier extended to the Little Ice Age-Neoglacial moraine at the red arrow. Koppes et al (2009) indicate a retreat of  13 km from 1960 to 2005, 300 m/year.

marinelli ge
Marinelli Glacier in Google Earth

marinelli ela
Cross section of glacier from Fernandez et al (2011)
Melkonian et al (2013) note widespread thinning with a peak on Marinelli Glacier. They also note frontal velocities of 7.5 m/day to 10.5 m/day from 2000 to 2011. They note approximately a 4 km retreat during this period and an average accumulation area ratio (AAR) of 38 (Melkonian et al, 2013). A non-calving glacier needs an AAR over 50 and typically over 60, since calving is an additional loss, calving glaciers typically need an AAR above 70 (Pelto, 1987).

marinelli velocity
Change in thickness on Marinelli Glacier from Melkonian et al. (2013)

Here we examine Landsat imagery from 1998 to 2014. In 1998 Marinelli Glacier had a main calving tidewater terminus and a land based terminus, red arrow. The tidewater terminus extends beyond the land based terminus. The land based terminus is connected to a tributary at the pink arrow. A tributary from the east is connected to the main glacier at the purple arrow. The yellow arrow is the 2014 terminus position. By 2001 the tidewater terminus has retreated up fjord of the land terminating terminus. The tributary on the west is still connected with the land terminating section of the glacier. By 2008 the main terminus has retreated exposing a new island in the center of the calving front. The land terminating section is now separated from the main glacier and with no supply of new ice will melt away, orange arrow. The tributary from the west is separated from the land terminus now at the pink arrow. The east tributary sill has a connection at the purple arrow to the main glacier. By 2014 the island at the main terminus has expanded in size as the glacier has retreated. The east tributary at purple arrow is separated from the main glacier. The isolated stagnant former land based terminus section between the red and orange arrows continues to melt away. The tidewater terminus of the glacier has retreated about 3.75 km from 1998 to 2014.  This is a rate of less than 300 m/year the long term average. The glacier will not stop retreating until its AAR rises and the calving margin reaches a pinning point. In this case there is no lateral pinning point apparent, hence it will have to be a rise in the elevation of the base of the glacier. The velocity and thickness change profile indicate such a location may exist 3-4 km behind the current calving front.  This glacier is retreating faster than the other glaciers of this icefield and is more in line with glaciers in the Southern Patagonian Icefield such as, Onelli Glacier, Glaciar Steffen, Glaciar Chico and Jorge Montt Glacier.

marinelli 1998
Landsat image 1998

marinelli 2001
Landsat image 2001

marinelli 2008
Landsat image 2008

marinelli 2014
Landsat image 2014

Glaciers have the Wrong Business Model

This is the story of a glacier that recently went to a bank for a loan, the Sholes Glacier, North Cascade Range, Washington. … “A recent resolution of mine is to work to change my future. I first had to go to the trouble of getting registered as a business so the bank would even recognize my existence, simply having existing on a map was not sufficient. This despite the fact that the water I store and release each summer is valuable to many businesses. I was ushered into the loan office where my basic need was explained, I need to replenish my main asset snow and ice, otherwise the water resource service I provide to others will diminish. The documentation requested included the state of my overall sector.  The World Glacier Monitoring Service, collects data on glacier mass balance and terminus change from around the globe, showed that my sector had lost net assets for 25 consecutive years, see below.  This graph showed not only that global glacier mass balance has declined 25 years in a row, but that North Cascade glaciers have lost an equivalent amount of volume, the 2014 data is preliminary.

Global Glacier Mass Balance

In fact auditors, glaciologists, have examined my asset sheet each of the last 25 years, and this data was not helpful. I then provided my own net asset sheet indicating a 25% asset loss in the last 20 years.  The increased stream of liability from me was eating the long term assets, that were literally no longer frozen. The bank officer, took a hard look and pointed out that, “banks loaned money with the expectation that there would be a return on their investment, improved assets of the loan recipient being crucial”. Given the recent history in the glacier sector I was told, “that our business model was wrong. We cannot expect after 25 consecutive years of loss that a positive asset trend is possible”.   I noted that the business model was hard to change and that is was the overall “business” climate that was wrong. This yielded a final rejection, “that maybe true, but until the business climate changes, you still have an unsustainable business model, and any loan would likely simply melt away, so to speak”.   So I ask for advice. What can I use for a business model? Will the business climate change in time for my business… How about the other businesses I supply too? I am afraid Kickstarter is not an option. Examine the other glaciers and their stories to see that my story is not unique.

sholes overvieew 2014

Sholes Glacier, North Cascades of Washington assets melting away.

sholes 2013 August melt

Snow melt from August 4th to Sept. 12th, 2013 on Sholes Glacier.

ba north america

Annual balance of glaciers in western North America all losing assets.