Physical Glaciology

Glacier stake height measurements, McMurdo Dry Valleys, Antarctica (1993-2020, ongoing)


As part of the Long Term Ecological Research (LTER) project in the McMurdo Dry Valleys of Antarctica, a systematic sampling program has been undertaken to monitor glacial mass balance and meltwater flow. This data package includes stake height measurements to the surface of six glaciers (Canada, Commonwealth, Hughes, Suess, Howard, and Taylor) in Taylor Valley and one glacier (Adams) in Miers Valley, all of which are located in the McMurdo Dry Valleys of Antarctica. Most measurements began during the 93-94 field season. Adams measurements were established during the 14-15 field season. Measurements are ongoing except at Hughes and Suess Glaciers where monitoring ceased following the 08-09 field season. Monitoring the changes in these measurements over time provides a record of mass balance, and aids in determining the role of glaciers in the polar hydrologic cycle.

LTER Core Areas: 

Dataset ID: 


Associated Personnel: 


Short name: 



The goal of each summer season is to take measurements in early spring (Oct/Nov) and late summer (late January). This provides a measure of seasonal winter/summer changes of glacier mass. In some circumstances we have the opportunity or need to measure the glaciers in mid-season (Dec). View the dates the measurements were made at the following URL:

Data sources: 



Stakes were drilled into the snow and ice. Measurements are taken from the top of the stake to a round board that sits on the surface. The board, about 25 cm in diameter, is needed to average the surface height over a larger area because local roughness, especially on ice surfaces, can overwhelm the change in height since the last measurement. Roughness typically results from an "ablation well" around the stake, caused by enhanced solar absorption and subsequent longwave radiation from the stake. Measurements against the stake became meaningless. Methodology evolved to taking 4 replicated measurements at each stake. Height to the surface was then calculated from the sum of the distance from the top of the stake to the board, the board thickness, and subtracting the depression of the board into the snow If snow covered the ice surface, the snow depth was also measured. Procedures evolved to acquiring 4 - 8 replications around each stake. Typically, the final value of snow depth was the average of all measurements. In some cases, values were discarded because measurements were taken in hidden holes in the ice surface. Snow depth is important to record because the ice surface could have ablated prior to a snowfall. Therefore, to assess the mass balance at that stake, both ice surface change and snow depth is required.


Data for the original stake height and snow depth measurements, as well as the averages and standard deviations was submitted by Andrew G. Fountain to the data manager in August, 1997. The column showing "file name" identifies the original file containing that record. These are ascii text files and can be found on the PC in the McMurdo LTER data manager's office at INSTAAR.
Once submitted to INSTAAR, the data manager used Microsoft Excel and Access software to produce files that were in more of a relational mode. Information was assigned to different files according to type of data it dealt with. Separate files were generated for:
                        * stake height replications,
                        * snow depth replications, and
                        * averages/standard deviations calculated from stake height and snow depth replications.
 In March, 2000, Thomas Nylen submitted data from the 1997-98, 1998-99 and 1999-2000 seasons. Instead of lumping all of the stake height and snow depth values for each glacier under one file, they were separated into files for each glacier. Metadata fields were also updated at this time.
In April, 2000, Denise Steigerwald added fields for dataset code and glstkid. Dataset code would allow the data to be linked to the  metadata in a relational database. Glstkid is a code that ties stake records to points on a GIS base map being developed by Michael Prentice at the University of New Hampshire.
In addition, Denise divided the 'stake' field to 'stake' and 'stk replcmnt' values. The 'stake' value is an ID for a stake in a given location  over time, while the 'stk replcmnt' value shows whether the stake is the original (A), the first replacement stake installed (B), the second stake installed (C), and so forth. (Melting on the glacier over time makes installation of replacement stakes necessary.) Since the replacement value is only relevant for the stake heights and snow depths datasets, division of the 'stake' value into separate fields made it more straightforward for grouping data across different datasets (eg. stake heights and snow densities).
In 2016, data and metadata were enhanced to comply with PASTA's unreasonable requirements.

Additional information: 

In some circumstances we have the opportunity or need to measure the glaciers in mid-season (Dec).   View the dates the measurements were made at the following URL:
About the SURFACE variable: The surface is either snow or ice. Sometimes a thin layer of snow (~a few crystals thick) is referred to as trace or dusting.
About the stakes:
A stake with an H or V indicates a stake along the ice cliff that forms the boundary of the  glacier terminus. An H is a horizontal stake placed into the vertical wall of the glacier terminus, and V is a stake placed into the ice apron at the base of the ice cliff. The vertical stakes were installed to support the horizontal stakes and provide a nearby measure of ablation for a surface with a much different slope.


Subscribe to RSS - Physical Glaciology