Paradise Ice Cave

                         Mount Rainier National Park

                               Charles H. Anderson Jr.

                                       Mark Vining

The inside of the Paradise and Stevens Glacier is more beautiful and more impressive by far than any its surface sculptures. Every year in the past sub-glacial passage are produced in the frontal glacier margin by one of the many streams, which flow within the glacier. Wonders are often expressed at the height of the archway, which seems out of proportion to the streamlet a few feet deep flowing among the rocks under foot.  The stream has done little more than initiate the course of the cave’s main tunnel. Once the opening is made, circulating currents of balmy summer air gradually enlarge it. A constant struggle goes on between forces enlarging the cave upward into the body of the glacier and those shortening it as the glacier flows slowly downward and melts at the lower end. Sculpturing continues until winter sets in and freezing temperatures call a halt or a section of passage collapses.

in these “Natural Deep-Freezes” transparent speleothems in a great variety of forms can develop. Since even the most stagnant cave has some circulation, stalagmites and pillars are much more prominent. Draperies sometimes persist on slantings, frost crystals of ten reach enormous heights.

A conversation in the Jackson Visitor Center--- “Boy, I hear the ice caves are beautiful!”  “Yeah, when I was there years ago the blue light inside was amazing! Lets go up there and explore them!” - end of conversation and maybe your life. The ice caves today are not the same experience as prior to the late 1970’s. The story is ever changing at the ice caves, and directly involves visitors to the caves.

Early human records of the ice caves are somewhat murky. The Paradise and Stevens Glacier Cave System was the most celebrated example of glacier caves in the world.  While several other glacier caves are well known in the world (some of which are in North America), most are difficult to reach and are very dangerous to enter. 

Early history of the Paradise Glacier Caves System, (the old name erroneously being “Paradise Ice Caves” is very fragmentary and obscure.  After 33 years of research and studies, little more is known of the earliest explorers today than the beginning of the study.  All that can be said is that near the turn of the century the cave system was know to the be quite sizable, (although maps were probably not made.) N.L. Toles, in 1908, described the caves in the Paradise and Stevens Glacier as well as the snowfield above Reese Camp in Pictures of Mount Rainier. Photographs from John H. William’s  1911 edition of The Mountain That Was God show the Stevens Glacier Caves and the Paradise Glacier Caves, the former being photographed by A.H. Barnes in 1906.  The text describes the two caves as being quite separate. Today we know that the caves above Reeds Camp are the upper Paradise and the lower Stevens ice caves. The Caves around Reeds Camp are the lower Paradise Ice Caves near Slurskin Falls.

At the time of William’s photographs, the Paradise Glacier extended down the rocky cliff West of the caves and covered the entire basin floor below.  Keen-eyed observers of today can see that an inconspicuous low ridge divides this basin, which is important in understanding the history of the caves and the now greatly diminished glacier.  The low dividing ridge was evident in the form of a “slight elevation” in the ice, which was considered to divide the ice mass between the Stevens and Paradise Glaciers. 

Other writers have described the same ice bodies as the Paradise Lobe and the Stevens Lobe of the Paradise Glacier.  It has always been assumed that the headwaters of the Paradise River (and the caves) were restricted to the Paradise Lobe, and likewise, the   headwaters of Stevens Creek (and the Stevens Cave) to the Stevens Lobe. Today it is known that both the Paradise River and Stevens Creek have their headwaters in the Paradise Glacier, those streams fed by streamlets tumbling down the step-like cliff overlooking the Stevens Glacier Basin. Many of these streamlets interconnect and some of them flow into both the Paradise River and Stevens Creek simultaneously. The Paradise Glacier Cave System thus straddles the low dividing ridge. 

Beginning in the early 1920”s the Guide Service offered regular tours in the the Paradise Glacier Caves in the Paradise Lobe. They created artificial entrances with explosives I years when winter snow covered the natural entrance. Recently, an old steel danger sign and a half-rotten wooden ladder were found in the cave. National Park Service officials and the Guide Service as being the type used until 1935, when steel was replaced by wood, identified the sign.  The ladder was identified as being the type used by the Guide Service to enter their artificial entrances.  According to long-term members of the staff of Paradise Inn, the Stevens Glacier Caves were found and first visited by tourists on a large scale about 1946. Emphasis was switched from the Paradise to Stevens Glacier Caves because the glacier was receding; by 1950 the Paradise Lobe had completely disappeared. Though the new tourist cave was really the previously known Stevens Glacier Cave, the name “Paradise Ice Caves” (erroneous from the beginning) was attached to it creating a double misnomer.  After 1950 the course of the Paradise Glacier Cave was only intermit-tenthly covered by snow banks in which remnants of the once famous glacier cave count be seen.

Until about 1967, most observers considered the glacier caves to be merely small cavities at the glacier snout, formed by warm air melting back a short distance into the glacier as the stream rushed out.  Some held an incorrect view that the caves formed anew each summer and were filled by glacier flow each winter. Recent systematic studies at first by the Cascade Grotto of the National Speleological Society, and later by the International Glaciospeleological Survey, have shown not only that the cave system is at least as large and complex as the “original “ caves, but that the development and enlargement of these caves are major factors in the glacier’s retreat and disappearance. An entire new subdivision of speleology has developed as new concepts of glacier caves have evolved during these studies.

When these studies were begun in 1967, the system consisted of a single cave, with all passages funneling into the main stream passage, Steven’s Creek.  Subsequently the section close to the glacier snout had changed greatly. Two successive collapses in 1967 isolated several of the lower passages, and most of the once magnificent Big Room collapsed in 1969 and 1970. During all of this time the cave system was easily entered, and sometimes it was even open in the winter. This was the period during which the cave was most eagerly explored.  By the winter of 1970 the total surveyed length was more than two miles, with at least an additional mile yet unsurveyed.

All of these studies took place during a period of rapid retreat of northwestern glaciers and comparable studies had no been possible during a period of glacial advance. However, during the winter of 1970 and 1971 the all time record of annual snowfall was broken at Paradise.  The cave system never opened that year, and surprisingly, the winter 1971 and 1972 broke the record a second time. Not until late July 1972 was another serious attempt make to enter the system (and that attempt was unsuccessful).  It was at this time that thoughts of glacial advance were foremost in the minds of the explorers.

In 1972 exploration turned to the firn caves in the vicinity of Stevens Glacier. At first they were used as a substitute for the glacier-hungry cavers who couldn’t find their way into the Stevens Glacier Cave. The 1972 National Speleological Society Convention, held in White Salmon, Washington, advertised the “Paradise Ice Caves” as a special attraction, but only the firn caves were open. The end of 1972 had mapped about two miles of firn cave and more was known to exist. Continued exploration and mapping through the summer of 1973, and the first expedition devoted entirely to glaciospeleology, conducted in September, discovered the relationship of the firn cave system to the glacier cave system. After the expedition total of 7.3 miles had been mapped in the combined firn and glacier cave system.  During the winter of 1973 and 1974 almost record snowfall was again noted at Paradise.  The glacier caves were finally entered on July first, 1974. The end of 1974 had mapped a total of 8 miles, and more was known to exist in the combined cave system. 

Several types of orifices open into the cave system; only a few are suitable for human entry. The lower entrances at the glacier snout are the only ones normally open to visitors. Several entrances exist along the western headwall and firn margin of the glacier. Whenever streams are seen coming off the headwall, one may well find entrances in those areas.  The upper entrances are far more dangerous than the entrances at the snout and should be out of bounds during periods of extreme melt age of the glacier.  This usually occurs near the end of summer, and such entrances tend to be deep, jagged, and vertical; they are also intakes for snow avalanches and rockslides as well as unpredictable stream flow.

The general visitor should not walk on top of the firn margin in the Paradise Rover or Stevens Creek area. The ceilings of cave passages are very thin towards the end of summer and the unwary visitor may break through and fall into the cave with fatal results.

When the snow cover has melted off the surface of the glacier, many crevasses can be seen. A few of them intersect the cave system. A few rivulets of melt-water on top of the glacier enter the caves through small crevasses, but rarely is it possible to look up a crevasse from inside. Other rivulets enter the caves through Moulin’s (round, vertical pits much like “dome pits “ of limestone caverns), and also through small tortuous channels winding downward through the ice.  Some of these channels eventually enlarge into Moulin’s, and a few Moulins’ elongate up-glacier forming canyon-like vertical slots, which are sometimes mistaken for crevasses.

Two types of foliation or discreet layering within the glacier are seen, one resulting from recystallization and glacier flow, and the other due to accumulation and compaction of snow. The former can be seen in the ceiling at the lower end of Stevens Creek Passage, the Big Room, and the Pillar Passage as streaks of varying density (and thus varying shades of blue) related to the stress system within the glacier. A particularly prominent example can be traced along Stevens Creek, thence up the confluent passage leading down from the Rockslide room. The latter is exemplified in the annual accumulations of snow since 1970.  Each winter’s snow not completely melting off in the summer, adds an identify-able layer, which will be recorded in the glacier until summer melting can penetrate to that level again.  This becomes increasingly difficult, for the past few winters have been adding more to the pile.  Successively younger layers are compacted to a lesser degree because they have a lighter load and have had less time to recrystallize.  The remnant of each year’s snowfall is identified by the rock fall and wind-blown dust that accumulate on the glacier surface. Ablation concentrates the debris into a layer at the level to which the melting penetrated prior to the accumulation of the next winter’s snow.

Bedrock is visible at only few points in the cave system.  Instead, the floor and lower walls of the caverns are composed of large and small boulders, cobbles, gravel, and silt brought into the system by rock fall, snow avalanches, and running water. Once inside, the cavern streams as much as in any terrestrial river rework this material.  Ice blocks or stream sediments, however, can build up into dams behind which may form quite sizable pools.  When these dams rupture, the sudden increase in the volume of stream-flow can cause a dangerous flash flood, which may trap the unwary visitor for hours without aid. The main effects of subglacial stream flow thus are aggregation and degradation of sediments.  Only minor niches in the ice show that the streams also cause some melting of the glacier.

The airflow responsible for most of the ablative enlargement of the Paradise Ice Caves follows well-recognized principles identical to those in well-studied limestone caverns. The inner portion of the cave system maintains a near constant temperature at all seasons due to the presence of a large volume of ice in contact with water. Near the entrances the air temperature rises and falls considerably above and below freezing, but the fluctuation becomes less and less at increasing distances inside. Even in winter, water and floor temperatures (necessarily read when the lower entrances are open), of the innermost parts of the cave system have all been 0oC.  Since the glacier is honeycombed with water passages, its entire body still must be 0oC.  In order for it to be stably maintained.

In the winter this often becomes a different case. Since warm air may not enter the cave for long periods of time, the consistency of temperature then depends upon the “thermal inertia” or heat capacity of the large of ice at 0oC.  If the present liquid water then freezes, and there ensues no more dependence of 0oC temperature on other conditions it is then free to drop.  Therefore, when the cave is plugged or the outside air remains below 0oC, the interior may freeze up and the inside temperature then fall below 0oC.

Winter blizzards pile up snow banks well into each lower entrance to the caves, sometimes to the point that the caves cannot be entered until late in the season.  When the lower entrances become plugged with snow, the active flow of air halts or is greatly reduced.  Instead, cold air must enter through upper entrances in order to replace the relatively warm cave air, which continues to rise out of them.  This is a much less active process, but important because the caves then become “natural deep-freezes”, trapping super cold air as long as the lower entrances are plugged.  At such times, portions of the system may freeze temporarily despite the internal heat of the earth.

When the lower entrances are not plugged by snow, cold winter air flows into them to replace comparatively light, warmer, cave air. In summer, this process is reversed. The air in the cave is colder than that outside and thus heavier. It flows out of the lower entrances, drawing in summer air through the upper openings.  As a result of these processes, visitors sometimes encounter surprisingly strong underground winds in narrow sections of the cave.  Most of the caves are so large, however, that these air movements are barely detectable.

The volume of air moving through the caves each summer, however, is sufficient to cause rapid ablation.  Ice screws driven into the glacier wall to hold guard ropes must be replaced every few days as the ice melts. Ablation is greatest in the summer due to the higher temperatures and thus lowers relative humidity of air flowing through the cave. The process is self-perpetuating; as passages enlarge, the airflow becomes greater and thus ablation quickens.

Formation of flakes greatly increases the surface exposed to ablation and thus the rate of enlargement and destruction of the glacier and cave. In its final stage, a glacier cave consists of little more than a short-lived three-dimensional complex of collapsing, splintering flakes, changing almost from hour to hour.       

Even in comparison with the surface of the glacier and nearby snowfields, the cave system is almost devoid of life.  Visitors accidentally bring some; bacteria boot soles, mice that ride into the caves in knapsacks attracted into them by the odor of food, and many other small forms of life.  Aside from the ice worm observed by one of us, larval and adult forms of insects, beetles, stoneflies, and a grylloblattid have been the only qualitative life noted to date.

The glacier flea, a small, wingless insect belongs to the order Collembolan, and is known to live exclusively on or in glacier ice and firn, is so abundant in the vicinity of glaciers (especially Mount Rainier) that at times ice surfaces are blackened with them. During the night they freeze fast to the firn and ice. Their food consists of conifer pollen blown on and in the ice channels.

Stoneflies belong to the order of insects Ptecoptera.  They begin life as aquatic nymphs and metamorphose directly into the winged adult by one molting of the skin.  As it so happens, the Paradise River stonefly has not yet been identified or categorized entomologically. Stoneflies are not known from any other Washington cave. They should be watched for in caves with running streams.

The nymphs usually possess gills of varying conspicuousness and live in cool running water. They are predators, and may leave the water briefly in pursuit of prey. They also leave the water when fully mature, upon which they fly away t mate. Two nymphs, one dark brown, and the other red brown, and one adult were collected in the Paradise River firn cave in 1972.

The family Grylloblattidae belong to the insect order Orthoptera, and is thus related to grasshoppers, cockroaches, and crickets.  There is only one known genus, Gryloblatta, in the western world.  All stages (which look much alike) thrive in damp environments with low, nearly constant temperatures.  This being the case, it is surprising that they have not been found in more caves.  So far as can be determined, they are scavenging insects, eating other insects that have been disabled or killed by the cold, or insect parts and other organic debris, possibly including some moss and algae in caves containing lava slime. Their metabolism is extremely slow and an individual may take several years to mature. Two species are known from Washington state: Grylloblatta Campodeiformis occurs throughout the Canadian and Montana Rockies and in the North Cascades as far south as Mt. Rainier.  It is best adapted to a temperature of about 4oC and is not known from any caves.  The second type, Grylloblatta Chirurgica, has been found in lava tubes in Washington, and in Stevens Glacier Cave.  Ashley Gurney, in his 1961 description of G.Chirurgica says, “The specific name is adapted from two Latin words meaning ‘pertaining to the surgeon’.”

In 1979 Charles H. Anderson, JR. and other members of the Inter-national Glaciospeleological Survey had finished all of the exploration of the 8.23 miles of presently mapped cave interior, which consisted of a complex maze of passages.  During this active period of exploration, frequent mention was made of the changing nature of the cave system and falling flakes from the walls and ceiling. Flakes are long and often massive slivers or chunks of ice, which separate from the walls and ceiling because of changes in interglacial pressures.

In the late 1960’s many visitors flocked to the cave area atop experience the unusual feeling of walking inside the blue, cold worked of the caves.  Then in the early 1970’s record amounts of snowfall occurred and the openings of the caves remained covered with snow a great portion of the time.  In the late 1970’s and into the 1980’s the openings became somewhat more accessible. In recent years, however, there has been an acceleration of falling ice chunks and flakes breaking loose from the cave’s interior, some being the size of a small car. As a result, which also is due to the marked decrease in annual snowfall, more and more of the ice melted away. This may in some part be attributable to global warming and occurred especially in summer.  The overall structure of the ice melted away and the cave area was increasingly weakened which in turn caused even more collapsing. In fact, the 8.23 miles of ice cave mapped in 1978 is now in 1993 probably down to less than perhaps one hundred feet.

Since 1986 winters at the ice caves have produced less than 700 inches of snow yearly on the average, and as a result, the Glacial Core has become more and more exposed during the summers of these years.  The snow pack, which covered the remnant of the Glacial Core in 1971, was completely removed from the upper Stevens Basin by 1987.  During this continuing recession period, the Trunk Passage of the Glacial Core was almost fully degraded to the Rockslide Room by 1991.  This passage having been increasingly vented and blocked of making further progression has caused the cave morphology to change radically in this area during this period.  In 1988, the Trunk Passage on the Paradise River was gone and on Stevens Creek it was greatly reduced and by the end of the summer the glacier had almost fully degraded to the upper portion of the Suicide Passage and the lower portion of the Waterfall Passage. Since 1990 and 1991 sections of the Glacier have been almost fully degraded to the Rockslide Room and Paradise Lost leaving the Glacier Core fully exposed during the summer’s end. In 1991 most of these passages around the Waterfall Room and the Rockslide Room had been reduced to open or partially filled trenches in the ice from one to four feet thick.

During 1993 the Paradise and Stevens Glacier Caves were in a radical recession. In September 1992 the last big room had collapsed and most of the ice cave was gone. Because of this radical recession period the Paradise and Stevens Glacier Caves by the year 1993 might only be located in the upper Paradise Glacier above the main Head Wall above Suicide Passage. In September 1991, we did find glacier caves in the upper Paradise Glacier above the Head Wall and these newly discovered ice caves might some day in the future become known as the new Paradise Ice Caves. The upper Paradise Glacier and Cowlitz Glacier might have nice Glacier Caves in the future. NOTE.  At any time in the future the Paradise and Stevens Glacier Caves might return if the Paradise Glacier begins advancement again over a period of years. This year, however, marked the end of a beautiful ice cave and at one time the world’s longest recorded Glacier Cave.

We are indebted to the Mt. Rainier National Park Administration and the Northwest Region of the National Park Service for financial assistance in our project for 1974.  Special thanks are in order for Mr. John Wilcox, Paradise Area Manager, for his services as project supervisor and for Mr. Rod Crawford, and Mr. Stanley G. Jewel, JR. for identification of specimens and information on animal life of the glacial environment.