Time Series of Glacier Retreat

Science Center Objects

The retreat of glaciers in Glacier National Park, Montana, has received widespread attention by the media, the public, and scientists because it is a clear and poignant indicator of change in the northern Rocky Mountains of the USA.  In 2017 the USGS and partner, Portland State University, released a dataset which describes the areas of the 37 named glaciers in Glacier National Park and two glaciers on the U.S. Forest Service’s Flathead National Forest land. The areas are described for 1966, 1998, 2005 and 2015/2016, marking 49 years of change for most of the glaciers and 50 years of change for a few.  The difference in record length is due to adequate satellite data not being available for a few glaciers in 2015.

Glacier Margin Time Series Results:

·         All glacier areas reduced between 1966 – 2015

·         Average area reduction between 1966-2015  = 39%

·         Largest area reduction (Boulder Glacier) = 85%

·         Smallest area reduction (Pumpelly Glacier) = 10%

·         In 2015, 26 glaciers remain large enough to be considered “active” in GNP

Siyeh Glacier (upper – 2015 satellite photo,  lower - 2005 aerial photo)
Siyeh Glacier (upper – 2015 satellite photo,  lower - 2005 aerial photo). The glacier margin is more clearly identified in the 2015 image  because seasonal snow is persisting on the ice and rock debris atop the ice can be more clearly seen than in the heavily shaded 2005 image.  (Public domain.)

The glacier areas were determined by digitally mapping the perimeters of the glaciers in late summer when seasonal snow has melted to reveal the extent of the glacial ice.  Digital aerial photography and satellite imagery were used with a Geographic Information System (GIS) to conduct the mapping using terrestrial photographs taken from nearby ridges and summits as references. Site visits were made to a number of the glaciers over several years to investigate portions of glaciers that were covered by rock debris which made delineation from aerial photographs and satellite images difficult.  These finalized mapped measurements of glaciers complement ground surveys of glaciers using Global Positioning Systems (GPS) and a repeat photography project that involves re-photographing historic photos of glaciers taken early last century.

The introduction of satellite imagery to the most recent time series analysis provided unsurpassed resolution and clarification of glacier margins.  In several cases, the high-resolution 2015 satellite imagery was used to inform margin characterization from previous aerial analyses where rock debris covering ice had been mistakenly excluded from the ice perimeter or where heavy shading had made margin determination difficult.  One example of this is the identification of a large portion of Siyeh Glacier that was previously unidentified.  The aerial photos of 1966, 1998 and 2005 were heavily shaded.  The only identifiable ice  was mapped as a small horseshoe shaped glacier along the headwall of Mount Siyeh.  However, 2015 satellite imagery allowed scientists to discover that rock debris obscured more than half of the actual  glacier area, accounting for a much larger glacier footprint than previously mapped.  A site visit confirmed these findings and the area increase was updated for the previous years, which meant that Siyeh Glacier was large enough to be considered an active glacier for all years in the time series.  The previous number of glaciers in GNP had been documented at 25, but satellite imagery has improved the accuracy of analysis and revealed that 26 glaciers are large enough to be considered active in 2015.    The availability of satellite imagery for the purpose of tracking landscape change, such as glacier area, will allow the USGS to update margin data more frequently and with greater accuracy.  


Appropriate use of these data include comparing Glacier National Park's glacier areas  between the years included in the data release, and from different parts of the world. The change in glacier area can be used to test models of glacier-climate interactions, and to estimate glacier contributions to streamflow.

Interpretation of the glacier area data comes with caveats, inherent to characteristics of small alpine glaciers, such as those in Glacier National Park.  The first is that the glacier margins, and calculated area, reflect only the footprint of the glaciers. There are no ice depths for the majority of the glaciers so ice volume cannot be directly calculated. Thus, in theory, a small glacier that is thick could have the same volume as a larger glacier that is thinner. Although in general, larger glaciers will have more volume, the relationship between area and volume is variable enough such that a measured reduction, or lack thereof, in area for each glacier through time must be interpreted with caution. One example will serve to illustrate this latter point. Several of the smaller glaciers in Glacier National Park formed on benches or ledges on steep slopes. The footprint of these glaciers was limited by the space afforded by the benches. Thus, in the past, the glaciers could not grow any larger in area because accumulated snow and ice would fall over the edges. However, these glaciers gained in volume by growing to the maximum thickness that could pile up on the bench. Once warming temperatures initiated accelerated ice melt, these small glaciers first lost thickness and volume without their footprint changing much at all. As these small glaciers became thinner, the continued ice melt caused them to retreat from the edge of the bench where they were located. Therefore, the relatively minimal reduction in areas for these small glaciers between a period like 1998 and 2005 does not necessarily indicate that there was little loss of ice.



Gem Glacier (left) and Blackfoot Glacier (right) in Glacier NP
Gem Glacier (left; J Scurlock, 09/15/16) is perched on a bench that limits its area expansion, whereas Blackfoot Glacier’s (right; J Scurlock, 09/01/09) area is relatively unlimited by its broad basin topography.(Credit: J. Scurlock. Public domain.)


Another caveat is that, for some glaciers, continued warming temperatures begin to have less influence on glacier size once the (now smaller) glaciers retreat to the upper confines of a cirque basin. These glaciers are now positioned under steep slopes and headwalls where there is more shade that slows melting, greater occurrence of snow avalanches which add mass to the glacier, and wind deposition of snow from other areas. Although these factors were contributing to ice accumulation in the upper portions of the glacier when the glacier was much bigger, they were minor factors in overall ice formation and melting. Now these factors are impacting a much larger percent of the glacier area even as the overall glacier area shrinks. The net result is that ice retreat and glacier area contraction can slow down, because of these compensating factors, even as regional air temperatures continue to increase. This complicates the simplistic assumption that more warming equals faster ice melt. 

Other variables, such as elevation, aspect, ice flow, and the presence of a meltwater lake at the glacier’s edge, influence glacier retreat to varying degrees. The last caveat is that, because of the complicating circumstances described above for small glaciers, the best measure of glacier response to warming temperatures is to look at the entire population of glaciers in Glacier National Park.

This dataset of "named glaciers" constitutes a subset of the total snow and ice inventory of GNP and does not represent a comprehensive accounting of all glaciers or permanent ice features in the park. A subsequent dataset will be published by USGS that will include all glacier and perennial snow features and will better represent the hydrologic contribution of ice and snow features of the park. 


Back to Retreat of Glaciers in Glacier National Park

Grinnell Glacier in Glacier National Park - 2016
Grinnell Glacier (J Scurlock, 09/15/16) has retreated to the mostly shaded, upper confines of the basin.   (Credit: J. Scurlock. Public domain.)