Roosevelt Glacier Retreat, Mount Baker, Washington

Roosevelt Glacier is on the northwest side of Mount Baker, Washington with its accumulation zone joined with the Coleman Glacier. My first visit to Roosevelt Glacier was in 1984, when the glacier had just completed a 30 year period of advance from 1949-1979.  Since 1979 the glacier has been retreating, image below. roosevelt 79This post examines Google Earth imagery from 1993, 2003 and 2009 along with field observations from the glacier. The purple line indicates the advance moraine that the glacier emplaced during the 1949-1979 period of advance. We mapped the location of this moraine in 1985, when it was still recent and very evident. The red line is from 1993, yellow line from 2003 and green line from 2009. In 1993 the area below the lip of lava flow cliff is thin and stagnant, cliff noted by purple arrow. By 2003 the glacier has retreated to the top of the cliff and by 2009 the glacier has pulled back from the edge of the cliff. By 2012 the lower glacier, viewed from the edge of the Coleman Glacier, is thin and uncrevassed in the lower 350 meters of the glacier, up to the red arrow. The glacier retreated 190 meters from 1979 to 1993 and 220 meters from 1993 to 2009. The rate of 14 m/year has been relatively consistent.roosevelt terminus 1993

roosevelt terminus 2003

roosevelt terminus 2009

roosevelt 2012
The glacier is fed by three principal accumulation zones: 1) A glacier tongue that descends from the summit plateau at 3200 meters, 2) an avalanche fed and direct snowfall region beneath the north ridge, at 2200 meters 3) an avalanche and direct snowfall fed region beneath the northwest face, at 2400 meters. The annual snowline has averaged 2150 meters on Roosevelt Glacier from 1984-2010, which has led to a similar retreat of 14 m/year and mass balance loss -0.52 m/year of all Mount Baker glaciers, Pelto and Brown (2012). Each summer we investigate the retained snow depth retained in the crevasse stratified exposures on Mount Baker in the 2400 meter range. Below are several images from this investigation. In the second image the magenta arrows indicate specific annual layers that have been retained. Typical thicknesses are 1.75 to 2.25 m. The next two images are from when we are in an icefall looking for the best exposure to measure the annual layer stratigraphic thickness. The behavior of this glacier parallels that of Deming Glacier, Boulder Glacier and Rainbow Glacier all on Mount Baker. roosevelt profile

crevasse roosevelticefall strat

icefall strat2

Lewis Glacier Disappears, North Cascades, Washington

In 1985 during my second visit to the Lewis Glacier, was the first time I confronted the idea of a glacier disappearing. We were able to peer down several crevasses and see the bottom of the Lewis Glacier, measurements indicated a maximum depth of 12 meters over an area the size of a football field. This glacier had been selected for the North Cascade Glacier Climate Projects’s mass balance program(Pelto and Riedel, 2001) assessing mass balance on 10 glaciers across the glacier clad mountain range. It was a small glacier in the drier part of the range, near Rainy Pass. This size made it attractive to observe in terms of response to climate change. The USGS map indicates a significant glacier with an area of 0.12 square kilometers in the 1950’s. By 1985 (top image) the glacier had lost half of its mapped area, there were still some significant blue ice areas, and areas of firn, snow several years old that is not yet glacier ice. Return visits each summer over the next few years chronicled the demise of the glacier. By 1988 (middle image) the glacier had shrunk dramatically even since 1985, with no area of blued ice even the size of a basketball court, the thickest ice measured was 5 meters. By 1990 the glacier was gone (bottom image), no blue ice left in the basin, the blue arrows indicate the lateral moraine above the now empty glacier basin. At the time I had not developed the model for forecasting glacier survival (Pelto, 2010). . . Google Earth imagery from 1998 and 2006 indicate the basin does retain snowcover late into the summer during most summers. However, the areal extent is much smaller than the glacier had been. The blue arrows in the 1998 image indicate the moraine marking the extent of the glacier during the Little Ice Age. The blue arrows in the 2006 image indicate the area where the deepest ice was measured in 1985. Runoff observation conducted below the glacier indicate a 70& drop in August streamflow after glacier loss. This is the principal impact of glacier retreat, a reduction in summer runoff, particularly acute for rivers late in the summer when all other snow in a basin has typically been gone. (Pelto, 2008)

Colonial Glacier Retreat and Hydropower

Colonial Glacier is on the southwest side of Colonial Peak in the Skagit River Watershed, North Cascades of Washington. The North Cascade Glacier Climate Project has made six visits to this glacier over the last 25 years. Meltwater from this glacier enters Diablo Lake above Diablo Dam and then flows through Gorge Lake and Gorge Dam. These two Seattle City Light hydropower projects yield 360 MW of power. As this glacier shrinks the amount of runoff it provides during the summer for hydropower is reduced. In 1979 the glacier was clearly thinning, having a concave shape in the lower cirque, but still filled its cirque, there is no evidence of a lake in this image from Austin Post (USGS). The glacier had retreated 80 meters since 1955. In 1985 my first visit to the glacier there was no lake at the terminus. In 1991 the lake had begun to form, second image, but was less than 30 m across. The upper glacier was a smooth expanse of snow. By 1996 the lake was evident, and was 75 meters long. In 2001 the lake had expanded to a length of 125 meters. By 2006 the lake was 215 m in length, and had some thin icebergs broken off from the glacier front. Runoff to the Skagit River is impacted directly by the climate change and the resultant retreat of the glaciers. Three notable changes in North Cascade streamflow have occurred.
1) Alpine runoff throughout the North Cascades is increasing in the winter (Nov.-Mar.), as more frequent rain on snow events enhance melting and reduce snow storage Streamflow has risen 18% in Newhalem Creek and 19% in Thunder Creek despite only a slight decrease, 1% in winter precipitation at Diablo Dam, within 5 km of both basins. These basins are on either side of Colonial Glacier.
2)Spring runoff (April-June) has increased in both basins by 5-10% due to earlier alpine snowpack melting.
3)Summer runoff has decreased markedly, 27%, in the non-glacier Newhalem basin with the earlier melt of reduced winter snowpack. In Thunder basin runoff has in contrast increased negligibly, 4%. The difference is accounted for in part by enhanced glacier melting. The observed net loss of -0.52 meters per year in glacier mass spread over the melt season is equivalent to 2.45 cubic meters per second in Thunder Basin, 10% of the mean summer streamflow. This trend of enhanced summer streamflow by reduction in glacier volume will not continue as the extent of glaciers continues to decline.

The lower portion of Colonial Glacier is not moving. GPS readings on both rockpiles on the lower glacier indicated no movement from 1996-2006. In the picture above the lake is still small in 1996, lower right corner and the lower rock pile distant from the terminus. The first two images below are from 2006, the lower rock pile is near the terminus and the last image is 2007 the lake has expanded back to the lower rockpile. Additional rock outcrops have appeared in the midst of the upper glacier that were not present in 1991, indicating this glacier does not have a persistent accumulation zone and will not survive current climate.