CRATER FIRN CAVES Of MOUNT ST. HELENS, WASHINGTON

Charles H. Anderson, Jr.

Christopher J. Behrens

Gary A. Floyd

Mark R. Vining

International Glaciospeleological Survey

547 SW 304 St., Federal Way, WA 98023

Abstract

Systematic observation, photo-reconnaissance, mapping, and sampling have been performed in the crater firn caves of Mount St. Helens, Washington, from 1981 through 1996 by members of the International Glaciospeleological Survey (IGS) in cooperation with the National Forest Service and Mount St. Helens National Monument.

Mount St. Helens is an active dacitic volcano, which is currently in a semi-dormant state after a catastrophic explosive eruption in May, 1980.  The crater is occupied by a dacite dome which plugs the volcanic vent.  The crater area has been progressively covered by a layer of snow, firn, and glacier ice since as early as 1986. Heat, steam, and volcanic gases from the crater fumaroles have melted over 2,415 meters (7,920 feet) of cave passage in the crater ice mass.  The caves are in approximate balance with the present geothermal heat release. Future changes in the thermal activity will influence the dimensions, location, ceiling, wall, and wall ablation features of these caves.  Cave passages are located above fumaroles and fractures in and adjacent to the crater lava dome.  Cave passages gradually enlarge by ablation, caused by outside air circulation and by geothermal sources beneath the ice.  The passages form a circumferential pattern around the dome, with entrance passages on the dome flanks.  Passages grow laterally and vertically toward the surface, spawning ceiling collapse.

The crater ice body has been expanding since 1986, and its mean density is increasing with each passing year.  It possesses at least two active crevasses. Trends and changes in geothermal activity in the crater of Mount St. Helens have been noticeable through cave passage observation and remapping.

Introduction

This paper describes the crater firn caves of Mount St. Helens, Washington, a system of melt passages in firn ice in the crater of an active Pacific Rim volcano.  Observations described were conducted from 1981 through 1996. Glaciologists have made mention of firn, crater, steam, and geothermal caves (Kiver and Mumma, 1975, and Kiver and Steel, 1975), and sometimes have dealt with their origin to a limited degree. No one in the past has provided timely observation of the evolution of geothermal ice cave systems in detail. IGS members are currently conducting these studies in the crater firn caves of Mount St. Helens, and have done so in an ongoing fashion throughout the development of the crater snowpack. This study documents an unique opportunity to capture data on the interaction of geothermal energy and alpine snowpack accumulation from its inception after the eruption of the volcano in 1980.

Charles H. Anderson Jr. began his investigative work in 1981 in the crater of Mount St. Helens.   Yearly surveys began in 1982 with mapping, documentation, and photography of cave passages, snow, firn, and glacier ice (Fig. 1).  Several years of observations and data have been collected, but they have not been systematically interpreted at this point. We have examined temporal relationships of the behavior of the system, thinking in terms of cause and effect. Our conclusions are based on preliminary results and qualitative interpretation.  We intend to offer what explanation we can for the trend of development observed in crater firn caves at Mount St. Helens.

The authors are grateful to IGS members and to staff of the U.S. Forest Service of Mount St. Helens National Volcanic Monument. Special thanks are given to Peter Frenzen (Monument Scientist), James Quiring (U.S. Forest Service), and Don Swanson (USGS).

Geological Setting

Geology of Mount St. Helens has been summarized by Pringle (1993). The Lava Dome consists of dacite, a fine grained igneous extrusive rock.  The west crater wall gives a good cross section of the volcano before the 1980 eruption.  The lower section of the crater wall consists of a dacite dome of Pine Creek age (Approximately 3000-2500 yr B.P.).  This dacite dome is cut by dikes of the Castle Creek age (Approximately 2500-1500 yr B.P.).  These are some of the feeder dikes for the Cave Basalt Eruption that formed Ape Cave, a well-known lava tube cave in Washington.  Above the Cave Basalt is the dacite summit dome of Kalama Age (Approximately 500-200 yr B.P.).  This is the pre-1980 summit dome of Mount St. Helens (Fig. 2).

Eruption History

Mount St. Helens is located in the southwest Cascade Mountains of Washington State, U.S.A.  Mount St. Helens, which has been called the most active and violent volcano in the United States, is once again active.  It erupted at 8:32 a.m. on May 18, 1980, sending billowing columns of ash high into the atmosphere (Tilling et al., 1990).  The eruption was preceded seconds earlier by a  magnitude 5.1 earthquake, which caused a large portion of the north flank of the mountain to slide.  Immediately afterward, an explosive lateral blast was directed northward through the still-moving slide block, closely followed by a summit eruption of ash and steam.

On May 19, 1980, when the extent of damage to the volcano was revealed, the entire north flank was gone.  The greatly enlarged and deepened crater was now a horseshoe shape, open to the north, with the south of the mountain being the highest remaining part of the volcano at approximately 2502 m (8,211 feet) above sea level.  The lip of the open crater on the north was estimated to be about 1890 m (6,200 feet) in elevation.  Mount St. Helens remains a potentially active and dangerous volcano, even though it now (1997) appears quiescent.  In the last 515 years, it is known to have produced four major explosive eruptions and dozens of lesser eruptions.

On June 15, 1980, the formation of a small lava dome on the floor of the crater was evident. The dome measured about 185 m (600 feet) in diameter and was less than 37 m (120 feet) in height.  By June 23, 1980, it had grown to be 200 m (660 feet) long and 60 m (200 feet) high.  From May 1980, to October 1986, there have been a series of 16 dome-building eruptions, constructing the new 305 m (1,000 feet) high and 915 m (3,000 feet) wide lava dome, in the crater formed by the May 18, 1980, eruption (Swanson and Holcomb, 1989).  The nearly 1.6 km (1.0 mile) wide, 3.2 km (2.0 mile) long, 610 m (2000 feet) deep crater is so large it makes the lava dome seem small (Fig. 2).  The Washington Monument placed in the crater would only be half as high as the lava dome.

Sufficient time has elapsed since the last dome-building eruption in October 1986 for magma in the conduit beneath the dome to crystallize and form a plug.  The pressure needed to overcome this blockage may exceed that required on May 18, 1980.  Therefore, the next eruption may be initially explosive owing simply to blockage of the conduit. This eruption is likely to be as large as, or larger than the eruption of May 18, 1980 (USGS, 1994).

Effects on Pre-existing Glaciers of Mount St. Helens

From analysis of USGS topographic maps, the May 1980 eruption removed all of the Loowit and Leschi Glaciers and parts of the Ape, Forsyth, Nelson, Toutle, Shoestring, Smith and Wishbone Glaciers. This represents more than 70 percent of the pre-eruption ice volume.  Only two unnamed glaciers on the south side suffered no net volume loss of ice during the eruption.  The Forsyth and the Shoestring Glaciers lost about 75 percent of their volumes of glacier ice, and their zones of snow accumulation were removed by the eruption. As a result, the Shoestring Glacier has suffered significant ablation.

In 1981, following the great eruption, a surge occurred in Shoestring Glacier.  Apparently, the weight of volcanic debris added to a fairly heavy snow load in the winter of 1980-81 to produce a sudden budget overbalance, in spite of removal of a substantial portion of the original ice volume.  The surge behavior was not repeated the following year.

Snow, Firn, and Ice in Mount St. Helens Crater

A large volume of snow and ice is presently accumulating in Mount St. Helens Crater.  The accumulation is protected by the shade of the high, steep crater walls to the south and west.  Crater ice volume increased from approximately 28 million m3 of uncompacted snow and firn in 1988 to over 55 million m3 in 1995.  These figures are derived from published USGS figures modified by consideration of thickness data collected from direct cave observation. As of late 1996, the crater was estimated to have 58.5 million m3 of ice, firn, and snow (Fig. 3).

During the winters since 1982, snow and ice avalanches from the crater walls have contributed to the formation of a snow and ice field on the south (interior) side of the lava dome.  The accumulation of avalanche material from the crater walls has helped form an ice field approximately 60 m (200 feet) thick, based on crater firn cave surveys.

The New Crater Glacier

A glacier is a system of flowing ice that originates on land through the accumulation and recrystallization of snow (Fig 4).  The necessary conditions for the development of a glacier are simple: more snow must accumulate each year than is lost by melting and evaporation.  Under these conditions, a new layer of snow accumulates each year, and over many years, the mass of ice eventually becomes dense and thick enough to flow under its own weight.  Water enters the system primarily in the upper parts of the glacier, where snow accumulates and is transformed into ice.  The ice then flows out of the zone of accumulation, generally moving a few centimeters per day.  At the lower end (or terminus) of the glacier, ice leaves the system by melting and evaporating.  The crater icefield is an incipient glacier that continues to grow.

The snow and ice on the south crater wall behind the lava dome have crevasses and flow features that show that a new glacier is forming (Fig. 4).  The ice mass is showing signs of ice flow around both sides of the lava dome and flowing out to the front of the dome.  The rate of its advance may be greater than any other glacier in the contiguous U.S. in recent centuries.  The new glacier is forming between the south crater wall and the lava dome (Fig. 5). The snows stacking higher each year have locally compressed the lower layers into the dense crystalline ice form known as glacial ice.  There are at least two radial crevasses in the permanent ice field caused by glacier movement over an uneven surface.  One crevasse is located on the northwest side and another is on the northeast side of the crater near the lava dome.  Both crevasses penetrate through the lowermost layers of the permanent ice field. The crevasse on the northwest side was revealed when the roof of an ice cave collapsed, due possibly to thrust fault activity in the crater floor around September 1994.  The ice density at the base of this crevasse was 0.85 g/cc (taken September 1994). The ice density in the lowest cave passage was 0.86 g/cc (taken September 1996; see Fig. 6).  Density measurements were performed using a cylindrical saw, and by weighing and measuring the cut samples in the field.

Progressive Recrystallization

When a winter's snowpack survives the following summer and is buried by the following winter pack, the buried snow layer will be compacted and recrystallized.  New fallen snow has a density of  0.06 to 0.08 g/cc.  As water percolates through the snowpack, and daily temperatures alter from high to low, individual snowflakes metamorphose first to subspherical porous grains, but later to granules of solid ice (corn snow).  By this time, the density of the snowpack has risen from about 0.1 to about 0.3 g/cc.  When the snow has recrystallized so that its density reaches an arbitrary value of 0.55, the snow has become firn.  As long as the firn has air pockets in it, the recrystallization process can increase its density.   After many seasons (25 to 150) the density in a glacier will reach 0.88 g/cc or more. The density of solid ice is about 0.92 g/cc.  The process will not continue if the confining pressure does not increase (Sharp, 1960).

Generally in the ice caves, firn can be distinguished from recrystallized recent snow (corn snow) by stratigraphic relationships.  Glacial ice forms from firn at a density of 0.82 g/cc, at which point the individual crystals become firmly interlocked with one another and the material possesses an inherent hardness (Sharp, 1960).

The fact that winter snowpacks from multiple years were preserved between 1986 through 1997 provided the pressure increase necessary to convert crater snowfall into a permanent firn field. As recrystallization continued in the deepest layers, the individual ice crystals grew together to form a rigid fabric with limited permeability (glacier ice).  From 1986 to 1996, gradual increase in basal ice densities were observed subjectively (though not measured) in cave passages.  In these cases the transition from snow to firn to ice was observed.  An apparently abrupt decrease in percolating water was observed in the final stage of this transition.  We interpret this condition to result from bulk freezing in intergranular pores.  Clearly, after a series of heavy winters and/or mild summers, there can be such a sequence of yearly net accumulations that it would take many years to degrade the body enough to remove them. In this way a "permanent" glacial core is being developed and perpetuated in the Mount St. Helens crater.

Geothermal Activity in the Crater

The Mount St. Helens lava dome is the locus of the active volcanic vent.  It therefore is a source of volcanic gas emissions throughout the crater area.  The caves are a primary result of the concentration of heat in an ice and snow covered terrain.  They are localized at active fumaroles, and form as conduits of escape for the heated gases.  They are further modified by the drainage of heated surface water from the dome directly into the ice body.

Periodic observations and resurveys of cave passages, in which changes in passage dimensions and location are noted, enable the detection of heat-flow changes and of locations of volcanic emanations. Sulfurous fumes occur locally in the caves.  Hundreds of small fumaroles emit considerable quantities of steam that frequently impair visibility in the firn caves and make mapping, photography, and other observations difficult. Some of these fumaroles make audible hissing and gurgling noises.  Although the rising heat and steam cause the ice walls and ceilings to drip constantly, no appreciable quantities of standing or flowing water have been observed in the caves.

Gases from the numerous fumaroles and slowly circulating surface air mix throughout the cave passages.  The degree of such mixing is most obviously recognized by the presence of breathable air throughout the known cave system.  We have not yet observed passages with either stagnant or poisonous air compositions.  It is considered routine safety practice to carry portable hydrogen sulfide and carbon monoxide detectors during new passage exploration.  Many of the larger cave rooms provide a protected environment for monitoring volcanic gas composition.  These rooms are ideal sites for prolonged monitoring of changes in volcanic emanations because they are relatively easy to find and their narrow connections with the up slope cave passages prevent rapid mixing with outside air.

If Mount St. Helens were to begin another eruptive phase, the first indications expected would be changes in the firn cave morphology (triggered by increased heat flow), together with increases in volcanic gas concentrations and microseismic activity.  The passages would be enlarged due to increases in the volume and temperature of steam and gas emissions from the fumaroles.  Such observations would signal danger, and curtail further in-crater human activity.  As volcanic activity increases, the heat release would be apparent from increased steam emissions and changes in the crater ice topography visible to remote sensors.  Large steam emissions would also be visible from distant viewpoints in the greater Mount St. Helens area.  Earthquake intensity could increase to a level observable casually by people nearby. Seismographs,  tilt meters, and visual observations would be the most useful means of monitoring volcanic activity at this stage.

Crater Firn Caves of Mount St. Helens

The crater firn caves are located in firn ice behind the lava dome of Mount St. Helens.  The firn ice field is elongated east-to-west with a steep crater headwall rising up to 2550 m (8,365 feet) on the south margin (Fig. 5).  The firn ice field proper is below the headwall on the southeast wall of the crater, rising to a maximum elevation of 1990 m (6,520 feet) on the south side of the lava dome, and sloping downward to the northeast. Further to the northeast, the firn ice field rises gently to a saddle (elevation 1890 m or 6,200 feet) adjacent to the crater wall.

The caves are called firn caves, because ice density ranges from 0.55 to 0.82 g/cc.   Sub-ice fumaroles and warm air currents form and maintain the cave passage beneath the ice field behind the lava dome.  Heat and steam from the crater fumaroles have melted over 2,415 m (7,920 feet) of cave passage in the crater ice mass (Fig. 7).  The caves are interpreted to be approximately in balance with the present geothermal heat release, because they have reached an apparently stable morphology.  Future changes in the thermal activity will influence the dimensions, location, ceiling, wall, and wall ablation features of these caves.  Rapid enlargement forms "steam cups" (Kiver and Steel, 1975).  Air circulation converts these to the typical scalloped ceiling and wall form seen in ice caves ubiquitously (Anderson, et al., 1994).

Entrances to 15 firn caves have been found around the perimeter of the lava dome.   These caves were mapped near the lava dome in 1996 (Fig. 7).   Some have spectacular large rooms.  Most have small rooms and crawlways.  Cave features include scalloped surfaces of ceilings and walls, moulins in the ceiling, multiple domes connected by crawlways, skylights, and toward winter, helectites, stalactites, and stalagmites of ice ( Figs. 8 and 9).  Room sizes varied from 12 by 24 by 6 m high to 4.6 by 4.6 by 2.4 m high in 1996. The caves generally occur in the presence of fumaroles.  Other caves form on the crater and dome walls where melt water undermines the firn and glacial ice.

Six main entrances and numerous smaller ones behind the lava dome lead down the 40 degree sloping crater floor immediately adjacent to the dome (Fig. 10).  The perimeter passage is surprisingly horizontal; the horizontality may be controlled by localized thermal activity along an arcuate fault or fracture zone within the lava dome.  If structural control were lacking, passage patterns should be more dendritic and follow the crater slope.  An arcuate distribution of thermal anomalies also suggests that volcanic emanations are escaping around a plug-like lava body (the dome itself) in the vent, and suggests a circumferential trend for further cave exploration.  Descending passages have vertical sides and ceilings that are convex upward.  Passages paralleling the slope contours are often shaped like right triangles with the 90 degree angle located at the junction of the down slope ice wall and the ice ceiling.  The floor slopes are about 30 degrees where mud to boulder-size volcanic rubble occur, and occasionally over 40 degrees where bedrock is exposed (Fig. 11).

Ridge-like accumulations of rock debris from the lava dome occur in many places on the floor against or near the ice wall of the passage (Fig. 12).   They are composed of unsorted, unstratified mud and rock debris derived from the up slope portion of the cave floor.  In places they occur toward the center of the floor and in others closer to or in contact with the down slope ice wall.  They probably represent talus formed against a down slope ice wall. This wall appears to retreat in response to temperature fluctuations.  Fluctuation could be due to normal seasonal changes or to changes in volcanic thermal activity.

The Ablation Process

Ablation is melting at the ice surface due to heat conduction, direct absorption of solar radiation, or evaporation/sublimation, which causes reduction of the volume of ice over time.  Exposure to geothermal heat from fumaroles or a generally elevated ground temperature can cause a similar volume loss from beneath. Glaciers and snowfields undergo ablation at all times when snow is not being added to them. Even when air is below freezing and after dark, ice will sublimate to some degree and degrade the ice body.  Summer is clearly the time of most rapid ablation.  All the ablative mechanisms are in their most active state then.  Temperatures are highest, air is usually drier than winter or spring, days are also longer with less obstruction from clouds.

Within the cave, evaporation/sublimation and heat conduction are the major active processes (Anderson, et al., 1994). Since caves are sheltered from sunlight, radiation from the sun has no effect, but radiant heat from the heated ground and fumaroles may have an appreciable effect.  The main control of cave ablation is the amount of air flow against the walls of the cave.   Since the crater cave passage networks extend over vertical distances, circulation within the ice cave system is affected by convection, especially since volcanic heat sources typically occur in the cave passages.  The flow rate is greatest in the least restrictive passage morphology. Since the ablation rate is faster where air flow is greatest, trunk passages will be initiated and become dominant.

External Air Communication

As cave ablation and surface ablation continue through a summer season, it is normal for the cave ceiling to approach and intersect the ice surface progressively over time.  If the ice is fractured, or perhaps after winter snow adds weight to the ceiling, a cave passage may experience ceiling failure. In either case the cave system suddenly gains a vent to outside air.  The effect of venting in summer time is to allow cold cave air out and warm outside air in.  The effect in winter is reversed.  The importance of ablation vents is exaggerated when there is any superimposed restriction in the system, such as winter snow or rock fall blocking other entrances.  In these cases, the vent entrance becomes a major means of communication with outside air.  When all vents to the surface are closed, the ordinary glacier cave becomes dormant.  In crater firn caves that contain internal heat sources, the ablation process can continue by convection even when all external openings are blocked. The system is therefore less seasonally dependent, and may evolve much faster than an ordinary glacier cave.

Crater Access at Mount St. Helens

Crater access in the Mount St. Helens Administrative Closure Area is through a permit issued to the International Glaciospeleological Survey by the National Forest Service.  This permit is for scientific research, and contains specific conditions that must be met by the group as an whole, and by each participating individual. Specific application is required by the Forest Service for each crater visit.  The crater is closed to public access, and anyone found without a crater permit within closed areas can be arrested and fined.

Hazards of Crater Cave Study

Since Mount St. Helens is active, there is an ever-present danger of volcanic eruption occurring while exploring the crater area (USGS, 1994).  Since November 1986, Mount St. Helens has been relatively quiet except for occasional steam explosions and ash plumes reaching as high as 5.6 km (3.5 miles) above sea level (Swanson et al., 1987; Tilling et al., 1990).  Even a small such eruption would be life threatening to anyone in the crater, if for no other reason than the poisonous nature of volcanic gases.  Field observations indicate that these explosions have thrown rocks more than 1.0 km from the lava dome, and have generated small pyroclastic flows in the crater.  Though these explosions have generated widespread public interest, they have been confined to the crater.  In the recent geologic past, when pyroclastic flows encountered an abundant water supply (perhaps snow and ice), they generated volcanic debris flows (lahars) that have been traced more than 16 km (10 miles) from the crater down Mount St. Helens' north flank and connecting valleys (Wolfe and Pierson, 1995).

Inside the crater there are many rock falls from the crater walls and the lava dome.  These rock falls pose a significant hazard to explorers entering the crater between August and November.  In winter, snow avalanches off the crater walls and the lava dome have been large enough to flow out of the crater.

The ice caves themselves present a hazard in the snow field areas.  These caves are changing each day and explorers must expect the entrances to collapse as cave passages grow internally by melting.  At any time during the hottest time of the day from June through September, roof collapse can occur spontaneously.  At any time, day or night, traverse over potentially thin roof ice is dangerous, and should be met with the same caution and preparation as for glacier traverse in the presence of hidden crevasses.  Ice caves can also trap poisonous gas emitted by the volcanic fumaroles occurring everywhere in the lava dome area. Oxygen is displaced by these gases, and people have died entering ice caves formed in these conditions on Mount Hood in Oregon (Kiver and Mumma, 1975).

Conclusions

The results of this ongoing study will lead to a more thorough understanding of crater firn cave evolution at Mount St. Helens or any locale in which ice accumulation interacts with geothermal energy.   Cave formation was initiated above fumaroles located along fractures in and adjacent to the lava dome. Cave passages gradually enlarge by ablation, caused by outside air circulation and by geothermal sources beneath the ice.  Passages grow laterally and vertically toward the surface, spawning ceiling collapse. The network of fumaroles have produced a ring of relatively horizontal passage connected to the surface by a number of ascending entrance passages.

An ice body is forming and expanding in the Mount St. Helens crater.  Its mean density is increasing with each passing year, and the transition from snow to firn to glacier ice (with active crevasses) has been observed.  Net budget balances have been positive in the crater since 1986, when the snowpack was first subjectively recognized to be growing.

Trends and changes in geothermal activity in the crater of Mount St. Helens have become noticeable through cave passage observation and remapping.  Our detailed mapping and investigations of the crater cave system should furnish a more sensitive indicator of geothermal activity than is furnished by remote surveys.

References Cited

Anderson, C. H. Jr., Vining, M. R., & Nichols, C. M. (1994). Evolution of The Paradise/Stevens Glacier Ice Caves.  Journal of Cave and Karst Studies 56, no. 2: 70-81.

Kiver, E. P. & Mumma, M. D. (1975).  Mount Baker Firn Caves, Washington.  The Explorers Journal, 84-87.

Kiver, E. P. & Steel, W. K. (1975).  Firn Caves in the Volcanic Craters of Mount Rainier, Washington. Journal of Cave and Karst Studies 37: 45-55.

Pringle, P. T. (1993).  Roadside Geology of Mount St. Helens National Monument and  Vicinity.  Washington Department of Natural Resources Division of Geology and Earth Resources Information Circular 88: 24-29.

Sharp, R. P. (1960).  Glaciers. Eugene, Oregon: University of Oregon Books.

Swanson, D. A., Dzurisin, D., Holcomb, R. T., Iwatsubo, E. Y., Chadwick, W. W. Jr., Casadevall, T. J., Ewert, J. W., & Heliker, C. C. (1987). Growth of the lava dome at Mount St. Helens, Washington (USA), 1981-1983.  In Emplacement of Silicic Domes and Lava Flows.  Geological Society of America Special Paper 212.

Swanson, D. A. & Holcomb, R. T. (1989).  Regularities in Growth of the Mount St. Helens Dacite Dome, 1980-1986.  In Lava Flows and Domes.  Proceedings in Volcanology, vol. 12.  Springer-Verlag.

Tilling, R. I., Topinka, L. & Swanson, D. A. (1990).  The Eruptions of Mount St. Helens: Past, Present, and Future.  U. S. Geological Survey.

USGS. (1994).  Preparing for the Next Eruption in the Cascades.  USGS Open-File Report 94-585.

Wolfe, E. W. & Pierson, T. C. (1995).  Volcanic-Hazards Zonation for Mount St. Helens, Washington, 1995. USGS Open File Report 95-497.