Climate Change: A Photographic Investigation

By Daniel Fagre

Across the globe, mountains have had more than their share of the climate changes that have been so evident in the last several decades. Major glaciers have entirely disappeared from the Andes, the Himalayas have lost a third of their snow, and 99 percent of the glaciers in Alaska are retreating. Particularly at upper elevations, average annual temperatures in mountains have increased between two and three times the global average. At the same time that there has been recognition that mountains are sensitive to climate change, the dependence of humans on mountain resources, especially water, has become ever clearer. Mountains provide 50 percent of the water humans use globally and 85 percent of the water that the relatively arid western United States depends on. Thus, how mountain ecosystems continue to respond to climate change will have direct impacts on human populations.

In 1990, the U.S. Congress passed the Global Change Research Act, directing federal agencies to examine how climate change potentially could affect natural resources of the nation. The National Park System was chosen to be a key player in the U.S. Global Change Research Program because national parks tend to be relatively pristine, making it easier to detect early or subtle changes attributable to climate change. The underlying dynamics of ecosystems also can be investigated more effectively with the nearly intact ecosystems found in many national parks. Against this backdrop, the U.S. Geological Survey has been monitoring and investigating the changes in mountain ecosystems related to ongoing climate change since 1991. In the Crown of the Continent, these efforts have primarily focused on Glacier National Park and have involved numerous collaborators from other federal agencies and universities.

Mountain ecosystems are very complex because of the strong gradients caused by elevation and the spatial variability caused by mountain topography. As every mountain hiker knows, the climate differs greatly as you ascend, go over a pass, or go around a slope to a different aspect. Consequently, plant and animal distributions change quickly over short distances, and there are interactions with the changing climate that complicate the picture. Determining which responses are attributable to a changing climate rather than the inherently dynamic qualities of mountain ecosystems is a daunting challenge. Monitoring small, alpine glaciers provides an elegant solution to this problem. These glaciers, distributed throughout Glacier National Park, are influenced almost completely by temperature and precipitation.

When it is colder and snowier, they accumulate more snow, which eventually forms ice, and the glacier grows. When it is warmer and drier, less snow accumulates and the glaciers shrink. Because it takes several years for the glaciers to reflect changes, one or two exceptionally warm or cold years are averaged out, and the size of the glacier reflects what is happening on the scale of decades. Therefore, changes in the size of glaciers indicate longer-term trends in climate.

The history and potential future of glaciers in Glacier National Park clearly suggest that major mountain ecosystem changes are a reality. Glaciers were present within current park boundaries as early as 7,000 years ago but may have survived an earlier warm period as well, which would make them much older. Tracking past climatic changes over thousands of years, these relatively small glaciers varied in size through time but gradually grew and reached their largest sizes around 1850. Physical evidence left by the glaciers (e.g., end moraines) indicates that there were an estimated 150 glaciers and large perennial snow/ice fields. Tree-ring based climate records and historic photographs indicate the initiation of glacier recession and thinning between 1860 and 1880. Between 1917 and 1941, the coupling of hot, dry summers with substantial decreases in winter snowpack (about 30 percent of normal) produced dramatic recession rates as high as 100 meters per year.

These periodic droughts have occurred on top of a long-term trend of a 1.6 °C increase in annual temperature since 1900. Many smaller glaciers disappeared during this period. Based on 1966 aerial photographs, the first comprehensive map of the region’s glaciers was published by the U.S. Geological Survey in 1968. Only 37 glaciers were named, out of a total of 84 perennial snow-and-ice bodies that survived from earlier in the century. It’s likely that at least some of the other 47 snow-and-ice bodies may have qualified as glaciers. For instance, Glacier National Park documents from the 1970s list “about 50” glaciers. In a 2002 publication, Carl Key, Richard Menicke and I estimated that 99 square kilometers of ice covered Glacier Park in 1850 but only 26 square kilometers remained by 1968.

In late September 1998, aerial photographs were acquired of all GNP glaciers. The glacier area measurements from these photographs were made by Michele Manly and were the first for all glaciers since 1966. The overall glacier coverage for Glacier Park was reduced to 17 square kilometers. Using criteria of a 0.1 square kilometers minimum area, and/or visual evidence of crevasses in the ice surface indicative of down slope movement, only 27 glaciers existed of the original 150. Other former glaciers appeared to have shrunk to the point of being miniscule and stagnant ice masses. Between 1993 and 1998, glaciers ranging in size from 0.15–1.72 square kilometers became 8 to 50 percent smaller. The relative rate of shrinkage was greatest for the smaller glaciers. Red Eagle Glacier, for example, was reduced to half its size between 1993 and 1998 and no longer meets the 0.1 square kilometer criterion for being considered a glacier.

Using Global Positioning System (GPS) technology, a survey of the glacier margin was completed for Grinnell Glacier in 2001 and showed a loss of 0.17 square kilometers, or 19 percent reduced area, from 1993 to 2001. The margin survey of Grinnell was repeated in 2004 and a further loss of 0.4 km2, or 5.6% reduced area, had occurred in three years. An additional 9 percent reduction occurred by 2006. Grinnell Glacier will be measured again in 2009 and will definitely be smaller yet again. We’ve witnessed several places where ice has collapsed into the lake, leaving icebergs that almost cover Upper Grinnell Lake.

Many watersheds of Glacier Park no longer contain glaciers, and glacial coverage in any of the remaining watersheds does not exceed 3%. Furthermore, glaciers have thinned by hundreds of meters and, like Grinnell Glacier, may have less than 10% of the ice volume that existed when George Bird Grinnell first explored GNP in 1887. Park area covered by ice and permanent snow was reduced from 99 km2 in 1850 to less than 16 km2 by 2005, and there are only 25 glaciers that meet our size and other criteria.

New aerial photography is scheduled to be acquired in 2009, and a new estimate of the remaining ice and permanent snow areas will be made next winter. Field measurements of other glaciers have been completed in the last few years, including Swiftcurrent, Chaney and Boulder, and these show that the loss of glaciers continues. Sperry Glacier—a glacier potentially more reflective of climatic change because it lacks a glacial lake such as the one at the base of Grinnell Glacier—was chosen as an index glacier for annual surveys and other measurements in a collaboration with Joel Harper and Blasé Reardon at the University of Montana. Sperry Glacier shrank from 0.89 square kilometers in 2003 to 0.86 square kilometers in 2005, according to precision GPS surveys of the margins at the end of the summer melt season. This represents a 3.6 percent loss in two years. Sperry Glacier will be monitored for mass balance, movement and ice depth in addition to its area. A climate station and automated camera have been installed, and GPS surveys of its margins and other features will continue annually.

To better understand the connection between changing climate and shrinking glaciers, all available data were used to build a computer-based projection of glacier dynamics now and into the future. With colleague Myrna Hall, this projection focused on the Blackfoot-Jackson Glacier Basin of Glacier National Park, where ice cover had decreased from 21.6 square kilometers in 1850 to 7.4 square kilometers. Using the temperature records from nearby weather stations, the climatic causes of glacier retreat in the Blackfoot-Jackson Basin were analyzed, the melt rate (change in glacier area/decade) was determined, and the topographic influences on the spatial pattern of melt were examined. Analysis of glacial area extent per decade from 1850 to 1979 versus a variety of climatic drivers reveals that annual precipitation and summer mean temperature together explain 92% of the loss over time. Using this information, potential future glacier behavior under both a “climate as usual” and a “global warming” scenario was predicted per decade until 2100.

These images of changing glaciers and landscapes were displayed for each climate scenario as an animated time series and indicated that all glaciers in the basin will disappear by the year 2030 if current trends of increasing temperatures continue under the “global warming” scenario. Even if no further warming occurs, the glaciers were predicted to be all but gone by 2100. The results were confirmed by several other computer models that also estimate that all glaciers will be gone between 2030 and 2050 at current warming rates in the northern U.S. Rocky Mountains. If the largest glaciers in Glacier Park will be gone by 2030, it is likely that the smaller glaciers will likely be gone as well.

More recent measurements of the Blackfoot-Jackson Glaciers indicate that the area in 1998 (2.94 square kilometers) was substantially less than the computer model predicted for 2000 (3.89 square kilometers) and was only slightly greater than the area predicted for 2010 (2.44 km2). This indicated that the glaciers were being reduced to specific areas nearly 10 years earlier than predicted. Myrna Hall compared the predicted temperature increase used in the model for 1990–2007 against the actual temperature increase in Glacier Park for the same period. The actual increase was twice as much as the model predicted. Precipitation was variable but did not have a net increase. This leaves the temperature increase as the cause of accelerated glacier retreat. It also means that the model was too conservative in predicting the demise of glaciers by 2030. Without a significant reversal in the upward trend in temperatures, the glaciers will continue to disappear, perhaps as early as 2020.

In 1997, my staff and I initiated the Repeat Photography Project with photographs repeated from historic images of the Grinnell and Boulder glaciers in Glacier National Park. The images revealed dramatic glacial recession and became, for many people, some of the first visual representations of the effects of climate change. Since then, repeat photography has proved to be a critically important tool for documenting and analyzing the retreat and disappearance of glaciers at Glacier Park. Of equal importance has been its function as a compelling communication medium for educating the public and policymakers about the dramatic transformation of the park over the last century of warming temperatures. Because humans are predisposed toward visual information, photographic evidence often trumps other types of data in convincing people that fundamental changes have occurred.

The earliest photographs taken of the area that was destined to become Glacier Park date from 1861, when a joint U.S.-British survey expedition was marking the boundary between the U.S. and Canada along the 49th parallel. These photographs were taken by the British surveyors and are of poor resolution, but they document a cold and stark landscape. Photographs do exist of Dr. Lyman Sperry and his team on Sperry Glacier in 1887 and by G.B. Grinnell at various times in the 1890s on Grinnell and other glaciers. However, the earliest photograph deemed useful for documenting glacier retreat is from 1900 by W.C. Alden, which shows a panoramic view of Grinnell Glacier. Repeat photography in Glacier Park has been used to effectively show other types of environmental change in mountains. Alpine treeline changes, geomorphological events, the aftermath and long-term recovery from wildland fire, and changes in grasslands have all been documented with repeat photography.

Beginning in 1997, a systematic search was made of the archives at Glacier Park and elsewhere to locate appropriate historic photographs of glaciers. To date, photographs have come from sources as diverse as personal collections to the National Park archives in Washington, D.C.

Photographs and negatives are digitally scanned, and copies are taken into the field to locate the photo site where the historic photograph was taken. A modern photograph is taken (i.e. repeated), and photo pairs are that focus on the glacier in that landscape are digitally created. Some of the photo sites have required multi-day hikes to relocate, and glaciers’ photographs must be taken late in the summer when snow has melted enough to reveal the ice and extent of the glacier. Fair weather, good air quality and the absence of late-season forest fires are also required. Thus, there is a narrow window of opportunity to repeat photographs each year. At this time, 44 photo points, or camera stations, overlooking 19 glaciers have been located where repeat photographs can be taken.

Dramatic changes have taken place over various time periods. I previously reported that 13 of 17 glaciers showed obvious reductions in size when comparing historic to repeated images. However, from 2005 to 2007, even glaciers that seemed to resist retreat, such as the Gem and Sexton glaciers, have begun retreating.

The paired photographs of the Boulder Glacier ice cave are significant from both a cultural and natural-resource perspective. The 1932 image illustrates the attraction that early tourists to Glacier Park had for glaciers as charismatic geological phenomena. The tourists are part of a guided horse-packing trip to the Glacier Park backcountry, and the furry chaps of the guide are visible on the figure closest to the ice cave. A mere 56 years later, all of the ice is gone, and vegetation has become established in the forefield of the glacier. This repeat photograph garnered the most media attention in 1997 and provided the impetus for establishing our current project.

Shepard Glacier clearly illustrates that glaciers have complex boundaries and features in these mountain environments. In 1913, the upper glacier portion in the wide cirque has crevassing indicative of fairly thick ice. The glacier flows down to the bench, where ice previously broke off the glacier front and fell to the valley below. By 2005, however, only a remnant of debris-covered ice (darkly streaked) remains in the upper left part of the cirque, and bedrock is showing elsewhere. Shepard Glacier is less than 0.1 km2 in area in 2005 and is thus considered by some scientists to be too small to be defined as a glacier. The modern photograph underscores this point compellingly.

Results from the study