Return to the Table of Contents for this chapter.

Subsurface soil temperature measurements at McMurdo Station, Antarctica

Mark A. Tumeo, Department of Civil and Environmental Engineering, University of Alaska Fairbanks, Fairbanks, Alaska 99775

Mark A. Cummings, Los Alamos National Laboratory, Los Alamos, New Mexico 87545

Over the past 20 years as the U.S. antarctic research effort has expanded, McMurdo Station, the main U.S. facility in Antarctica, has grown in size and complexity. The increased activity at McMurdo has meant an increase in the introduction of hydrocarbon contaminants through fuel spills on the ice and soil around the station.

When considering movement of fuel spills in an area where the soil freezes intermittently (called the active layer) or has a permanently frozen layer (called permafrost), understanding the freeze-thaw patterns of the soil is essential. Freeze-thaw cycles greatly affect contaminant movement through the processes of exclusion and immobilization (Iskandar and Jenkins 1985; Zukowski, Tumeo, and Lilly 1989; Zukowski and Tumeo 1991; Tumeo and Davidson 1993).

Because soil temperatures in or around McMurdo Station have not been measured, the depth to permafrost or the nature of the active layer was not known. To gain this information, a string of thermistors was installed at the site of an accidental fuel release near the J-1 fuel storage tank at McMurdo. This article reports on the subsurface temperatures found at the spill site and their potential effects on contaminant migration. This project was part of a 3-year study to address fundamental questions that surround the analysis of hydrocarbon contamination movement in polar regions.

Site description and methodology

In mid-February 1991 during a transfer of fuel from tank J-2 to J-1, approximately 11,400 liters of JP-8 was spilled at a site located in the pass between McMurdo Station and Scott Station on Ross Island. In 1993, this spill site was selected as a study site to track the movement of the contaminant plume.

Grain size distribution (ASTM sieve method), percent organics (loss on ignition), and hydraulic conductivity (auger hole method) were used to characterize the soil conditions around the spill site. On 24 January 1994, a thermistor string was installed 120 centimeters (cm) into the ground near the spill to measure seasonal temperature changes in the soil subsurface. Before installation, each thermistor was calibrated to within ±0.3° of 0°C in an ice bath. During installation, ice-rich frozen ground was encountered at a depth of approximately 25 cm. Eight thermistors were mounted at the following depths from the soil surface: 10.2, 20.3, 30.5, 40.6, 61, 81.3, 101.6, and 121.9 cm. Temperature measurements have been taken weekly since the thermistors were installed.


Implications of data on contaminant movement

In ice-rich permafrost, the pores of the permanently frozen soil are filled with ice, thereby creating an effective barrier against downward migration of water and/or contaminants. The fact that the active layer is only approximately 10 cm deep, whereas ice-rich permafrost does not occur until 25 cm means that approximately 15 cm of frozen soil is not saturated, and contaminant can move freely through it. More important, the data indicate that there is probably not a zone of exclusion caused by freezing at the spill site. The fact that a zone of permanently frozen, non-saturated soil lies above the ice-rich zone means that the freezing front does not typically involve the freezing of pore water in the active layer. Instead, water will move downward into the permanently frozen non-ice-rich soil, and flow laterally across the top of the ice-rich layer, until it either freezes or exits the system through exfiltration or evaporation.

The one caveat to this conclusion is that if the water balance in the area is such that excess water is entering the system and adding to the ice-rich layer--in essence moving the ice-rich layer closer to the surface--a time may come when the active layer does have a saturated zone. In such an instance, exclusion could become an important factor in contaminant migration. In the arid climate of McMurdo, under normal circumstances there would probably not be enough excess water to result in an exclusion zone, but water is often applied into the roads in the summer months to control dust. In the future, this excess water conceivably could saturate areas of the active layer.


We would like to acknowledge the assistance and support of Antarctic Support Associates personnel for their consistent and conscientious monitoring of the thermistor string during two long antarctic winters. These dedicated individuals braved extreme cold to record the weekly readings from the thermistor string. Without their efforts, none of the information presented here would have been possible.

This research was supported by National Science Foundation grant OPP 91-19559.


Iskandar, I.K., and T.F. Jenkins. 1985. Potential use of artificial ground freezing for contaminant immobilization. In I.K. Iskander (Ed.), Proceedings of the International Conference on New Frontiers for Hazardous Waste Management (September 15-18), Pittsburgh. Washington, D.C.: U.S. Environmental Protection Agency.

Tumeo, M.A., and B. Davidson. 1993. Hydrocarbon exclusion from ground water during freezing. ASCE Journal of Environmental Engineering, 119(4), 715-724.

Zukowski, M.D., M.A. Tumeo, and M.R. Lilly. 1989. Subsurface transport of BTEX compounds under freezing conditions. In W.S. Ashton (Ed.), Ground water: Alaska's hidden resource, Proceedings of the Alaska Section of the AWRA Annual Meeting (March 16-17) Chena Hot Springs, Alaska. Fairbanks: University of Alaska.

Zukowski, M., and M.A. Tumeo. 1991. Modeling solute transport in groundwater at or near freezing. Groundwater, 29(1), 21-25.