Surficial Extent and Conceptual Model of Hydrothermal System at Mount Rainier, Washington

Journal of Volcanology and Geothermal Research 65 ( 1995) 5 l-80

Surficial extent and conceptual model of hydrothermal system at Mount Rainier, Washington

David Frank U.S.Environmental Protection Agency, 1200 Sixth Avenue, ES-098 Seattle, WA 98101, USA

Received 8 January 1992; revised version accepted 10 February 1993*

Abstract

A once massive hydrothermal system was disgorged from the summit of Mount Rainier in a highly destructive manner about 5000 years ago. Today, hydrothermal processes are depositing clayey alteration products that have the potential to reset the stage for similar events in the future. Areas of active hydrothermal alteration occur in three representative settings: ( 1) An extensive area (greater than 12,000 m*) of heated ground and slightly acidic boiling-point fumaroles at 76-82°C at East and West Craters on the volcano’s summit, where alteration products include smectite, halloysite and disordered kaolinite, cristobalite, tridymite, opal, alunite, gibbsite, and calcite. (2) A small area (less than 500 m’) of heated ground and sub-boiling-point fumaroles at 55-60°C on the upper flank at Disappointment Cleaver with smectite alteration and chalcedony, tridymite, and opal-A encrustations. Similar areas probably occur at Willis Wall, Sunset Amphitheater, and the South Tahoma and Kautz headwalls. (3) Sulfate- and carbon dioxide-enriched thermal springs at 9-24°C on the lower flank of the volcano in valley walls beside the Winthrop and Paradise Glaciers, where calcite, opal-A, and gypsum are being deposited. In addition, chloride- and carbon dioxide-enriched thermal springs issue from thin sediments that overlie Tertiary rocks at, or somewhat beyond, the base of the volcanic edifice in valley bottoms of the Nisqually and Ohanapecosh Rivers. Maximum spring temperatures of 19-25°C and 38-50°C respectively, and extensive travertine deposits have developed in these more distant localities. The heat flow, distribution of thermal activity, and nature of alteration minerals and fluids suggest a conceptual model of a narrow, central hydrothermal system within Mount Rainier, with steam-heated snowmelt at the summit craters and localized leakage of steam-heated fluids within 2 km of the summit. The lateral extent of the hydrothermal system is marked by discharge of neutral sulfate-enriched thermal water from the lower flank of the cone. Simulations of geochemical mass transfer suggest that the thermal springs may be derived from an acid sulfate-chloride parent fluid which has been neutralized by reaction with andesite and highly diluted with shallow groundwater. The model may accomodate some of the thermal springs beyond the base of the edifice. Present heat flow from Mount Rainier is substantial relative to other Cascade Range volcanoes and does not appear to have diminished since at least the late 19th century. Evidence of older hydrothermal processes found in Holocene lithic tephra and debris avalanches record activity more extensive but similar in chemical composition to that of today.

* This paper was initially part of the Harry Glicken Memorial special issue. At a later stage, after consultation with the Guest Editors, the Publisher has removed the paper from the special issue to be published in a regular issue.

1. Introduction 2. Geologic setting

Mount Rainier rises to 4392 m from a lOOO-m-high Like many of the other large stratovolcanoes in the base of deeply eroded, Tertiary rock west of the crest Cascade Range of western North America, Mount of the Cascade Range in western Washington (Fig. 1) . Rainier’s hydrothermal system produces a distinctive The Tertiary base consists of a 5000-m-thick, gently folded pile of Oligocene and younger volcanic flows set of mineralogical alteration products from primary and detritus ranging from basalt to rhyolite, though volcanic rocks. Concurrently, the hydrothermal system andesite and rhyodacite predominate (Fiske et al., overlays a chemical imprint on groundwater. Surface 1963). These volcanics were intruded, mainly during leakage of these fluids produces thermal activity in the the Miocene, by numerous dikes and sills and a major form of fumaroles and heated ground on the volcano’s granodioritic complex (Fiske et al., 1963; Mattinson, upper slopes, and thermal springs at lower elevations. 1977; Murphy, 1991). Hydrothermal activity opens a window into subsur- Early Pleistocene volcanic sediments of horne- face processes that can dramatically affect the course blende-hypersthene andesite (Lily Creek Formation) of future volcanic events, be they magmatic or phreatic. west of Mount Rainier may represent the earliest erup- Hydrothermally altered rocks or pressurized hydro- tions from a Rainier vent that has since been buried thermal fluids provide source areas for non-magmatic (Crandell, 1963b, pp. A21-A22; Mattinson, 1977, p. explosions and debris avalanches. With the addition of 15 12). The bulk of the present edifice is built of lava magma intrusion, hydrothermal systems can produce flows and brecciated lava flows, with lesser amounts of preferential pathways for structurally weakening vol- interlayered tephra and other pyroclastic deposits. canic cones. Also, heat stored in hydrothermal systems These products vary little in composition, consisting can help drive eruptions initiated by dike-related frac- mainly of hypersthene-augite andesite, though two turing. Future scenarios or localities of destructive, small satellite vents of olivine andesite occur on the hydrothermally influenced events may be anticipated north flank (Fiske et al., 1963). by constructing conceptual models of hydrothermal Mount Rainier has been one of the more active Cas- systems. cade volcanoes during Holocene time. Evidence of 11 This paper relies primarily on minerals, gases, and eruptions can be found in Holocene tephra deposits water found in surficial deposits to construct a concep- (Mullineaux, 1974)) with major eruptive episodes between 6500 and 4000 years ago and again between tual model for Mount Rainier that considers the follow- 2500 and 2000 years ago. Lava flows, flow breccia, and ing factors: explosion rubble that form a young summit cone at - Locations of hydrothermal leakage at the surface; least 400 m thick are no more than 2200 years old - Structures that provide permeable paths of fluid (Crandell, 197 1, p. 11). The most recent dated eruption egress to the surface; episode produced pumice sometime during the early to - Amount of excess heat discharge; middle 19th century (Mullineaux et al., 1969). - Composition of surficial thermal fluids; In addition to eruptive deposits, at least 55 large - Composition, guided by mineralogy, of subsurface Holocene mass-movement deposits from Mount Rai- thermal fluids. nier, including debris avalanches and lahars as large as Analytical data used as a basis for the model are from 3 X lo9 m3, have been identified (Crandell, 1971; Scott samples collected during field investigations in 1982- et al., 1992). Many of these events apparently accom- 1985 (Frank, 1985), whereas other field and remote panied eruptions. However, some of the larger mass- sensing observations span a longer time period up to movements occurred during time periods for which no the present. In addition, a critical overview of historical eruptive deposits have been recognized (Crandell, information on hydrothermal activity at Mount Rainier 197 1, p. 57)) and may have resulted from non-eruptive is provided as an important context in which to place a phenomena such as hydrothermal explosions, earth- conceptual model. quakes, or oversteepening from long-term erosion. D. Frank/Journal of Volcanology and Geothenn~l Research 65 (1995) 51-80 53

\ \ MT. ST. HELENS ,

480

Fig. 1.Regional map of Mount Rainierlying between the crest of the CascadeRange and the densely populatedPuget Lowland(heavy outline). Black area denotes the extent of Mount Rainier lava flows. Shading shows approximatearea inundatedby the largest Holocenelahar~ 5000 years old or younger. Adapted from Swanson et al. ( 1992, fig. 1).

3. Thermal activity - The lower flank, 1000-3000 m altitude; - Adjacent to the base of the cone.

Historical reports of visual observations, more recent field investigations, and aerial thermal-infrared surveys 3.1. Summit area provide a basis for mapping active thermal areas at Mount Rainier. The complete record of data on thermal Historical observations activity can be conveniently discussed with reference Boiling-point fumaroles and adjacent heated ground to four representative settings (Fig. 2), the first three cover an exposed area of 12,000 m2 along the rims of of which are on the cone: two overlapping craters at the volcano’s summit (Fig. - The summit cone, above 4000 m altitude; 3). Additional fumarolic activity occurs beneath the - The upper flank, 3ooo-4000 m altitude; ice cover that fills the basins of both craters. The first 54 D. Frank/Journal of Volcanology and Geothermal Research 65 (1995) 5140

Fig. 2. Map of localities of thermal activity at Mount Rainier. A large fumarole field occurs at the top of the young summit cone (large dot) ; fumaroles and heated ground occur at Disappointment Cleaver (DC) and, in some cases ephemerally, at 5 other similar but poorly accessible areas on the upper flank (small dots). Thermal springs (squares)occur on the lower flank of the volcano near Paradise (ES) and Winthrop ( WS) Glaciers and beyond the volcanic edifice at Longmim (23) and Ohanapecosh (OS). Dashed ama= present extent of Mount Rainier lava flows; shaded area= glacier cover. recorded visit to the summit in about 1855 and many abundant sulfurous vapor issued from the West Crater more from 1870 through the late 19th century noted while the East Crater produced only steam that lacked the presence of fumaroles, some with sulfurous odor, sulfurous fumes. Shortly thereafter, I.C. Russell within the crater rims (Molenaar, 1979, pp. 33-49). (1898) and others of the U.S. Geological Survey The earliest report is that of the Indian guide, Salt&in, described fumarolic activity in 1896 on the inner slope who maintained that two of his clients in 1855 observed of the East Crater rim, much as can be seen today. a crater lake and steam vents (Haines, 1962, p. 17). Coombs (1936, pp. 202-203) gave similar descrip- Later, Stevens ( 1876) reported the occurrence of sul- tions, and added that no traces of sulfur were detected furous fumes and steam vents along the north part of in the steam emanations from the East Crater. the West Crater rim in 1870. A veteran Rainier climber, Saluskin’s second-hand report of an exposed crater ES. Ingraham ( 1895, p. 21)) noted that as late as 1894 lake in the mid- 19th century appears credible, since the D. Frank/ Journal of Volcanology and Geothermal Research 65 (1995) 51-80 55

Fig. 3. View looking northeast at the summit fumarole field. Fumaroles and heated ground keep about 12,000 m‘ of the rims of East and West Craters free of snow during summer months. The craters span about 700 m. Aerial photograph 69Rl-61 by Austin Post, USGS, August 22, 1969. most recent verifiable eruption, represented by a sparse In a detailed compilation of historical reports and tephra deposit, occurred not more than a few decades newspaper accounts, Majors and McCollum ( 198 la,b) earlier between 1820 and 1840 (layer X of Mullineaux, make a case for a tephra eruption from the summit area 1974). The heat released from the cooling products of based on observations of apparent smoke emission by that eruptive episode could have provided the energy residents in the Puget Lowland in November 1894 and to maintain a crater lake for a few decades. Later by an exploratory climbing party on the lower flank of observers suspected the presence of a subglacial lake the volcano the following month. The descriptions, based on audible splashing noises as rocks are rolled however, indicate only small amounts of “smoke” downslope through ice caves, for example in 1911 were seen; this and the small fraction of observers rel- (Flett, 1912) and in 1954 (Molenaar, 1979, p. 186). ative to the population of the time suggest that, more 56 Frank/Journal of Volcanology and Geothermal Research 65 (1995) 51-80

likely, increased minor fumarolic activity may have relic activity and heated ground not only lined the inner occurred in 1894, perhaps accompanied by ejecta rims of the craters but also extended partway down the entrained in the fumarole plumes. Close-range climb- outer west slope of the West Crater. ers’ descriptions of summit fumaroles in subsequent Subsequent aerial infrared surveys show a thermal years do not indicate any persistent change in fumarolic pattern for the crater area that is very similar (Fig. 4) activity. to earlier data acquired by Moxham ( 1972). Small More detailed field studies have been made during variations in 1969, interpreted by Moxham ( 1972, p. the last three decades. Field observations over a period 121) as real changes in geothermal emission, however, of several weeks in 1959 as part of a human-endurance are more easily explained by seasonal variation in snow experiment called Project Crater (Molenaar, 1979, p. cover adjacent to the thermal areas. Much of the warmer 179) provided the first repetitive set of temperature ground is bridged over by snow in the winter and thus measurements. Reported values include maximum masked from infrared sensors. For example, Fig. 4 temperatures of 77°C for West Crater fumaroles (Mil- allows comparison of surveys made during different ler, 1970). Molenaar (1979, p. 176) made a suite of seasons. The early spring survey of April 1973 failed temperature measurements around both crater rims in to detect many of the small thermal areas in the south- August 1970 and found that both fumarole and near- west part of the East Crater rim (Fig. 4) because in surface ground temperatures (8-15 cm deep) ranged April, a time normally near the peak of the snowpack, up to 72°C. The following month, Moxham et al. these areas were covered by snow. Field observations ( 1972) installed a temperature-monitoring platform during this same time period in the mid- 1970’s did not with satellite telemetry on the northern part of the West verify any unusual changes in thermal activity. Crater rim. The instrument operated for five weeks and provided the only continuous thermal measurements at Recent observations and crater structure the summit for an extended period of time. Three Infrared images acquired through joint US. Depart- probes in small fumaroles on the West Crater rim ment of Energy and U.S. Geological Survey efforts recorded maximum near-surface temperatures of 70- (Kieffer et al., 1982) show a representative pattern of 72”C, accompanied by several decreases in temperature heat emission from the summit area (Fig. 5). Figures that were probably caused by meltwater entering the 4-5 indicate that the greatest heat emission is concen- vents (Moxham et al., 1972, p. 197). trated in four clusters of activity: Further Project Crater efforts during 1969-1972 - The inner slope (A and B) of the northwest quadrant (Lokey et al., 1972; Lokey, 1973) led to increased of the East Crater rim; exploration of the geothermally formed ice caves -The inner and outer slope (C) of the northwest quad- within the crater basins (Kiver and Mumma, 197 1). rant of the West Crater rim; and Highlights of these studies were the production of a - An isolated outcrop (D) on the outer southwest slope plane-table map of the crater rims and adjacent glacier of the West Crater. surface (Lokey et al., 1972), and the discovery of a Lesser amounts of activity occur elsewhere along the cold subglacial lake in West Crater and topographic inner slopes of both crater rims and far down the outer evidence (a depression in the snow surface) for an northwest and west slope of the West Crater. additional subglacial lake in East Crater (Kiver and Field investigations in September 1982 and August Steele, 1972). The West Crater finding resolved dec- 1983 (Frank, 1985) recorded typical vapor tempera- ades of speculation over the existence of a crater lake. tures of 72°C for West Crater fumaroles and 80°C for These ground studies yielded a number of spot ther- East Crater fumaroles. The maximum fumarole tem- mal measurements at various localities near Mount peratures were 74 and 81°C respectively. Heated Rainier’s summit. The first synoptic display of heat ground temperatures in unvented areas ranged up to 72 emission in the crater area, however, is the result of a and 76”C, respectively. For comparison, the boiling series of aerial infrared surveys in 1964,1966, and 1969 point of pure water at the crater altitude is about 86°C. by Moxham (Moxham et al., 1965; Moxham, 1970, Two temperature-probe traverses shown in Fig. 6 illus- 1972). His results, coupled with the ground investi- trate representative shallow temperatures of the hottest gations of others, demonstrated that widespread fuma- parts of the summit fumarole field. D. Frank/ Journal of Volcanology and Geothermal Research 65 (1995) 51-80 57

Crater , . l c!Ezf July 21,1981,0429 . i

April 29,1973,0540 .

. . \ .

. . . 0760 (Moxtlain ad b others, 1965) l4r l 4 N . Pig. 4. Diagram of representative thermal-infrared anomaly patterns for the summit fumarole field at East and West Craters during the past 29 years, as shown by aircraft-based line-scan images. The thermal patterns represent areas of fumaroles and heated ground that line the inner slopes of both crater rims and that cover a large area extending down the outer southwest to northwest slope of the West Crater. Scale bars indicate image distortion. 1973 and 1981 infrared images courtesy respectively of J.D. Friedman and H.H. Kieffer, U.S. Geological Survey. 58 D. Frank/Joumal of Volcanology and Geothermal Research 65 (1995) 51-80

121%‘15” 121044’45”

46930’45”

Fig. 5. Thermal area at summit fumarole field. (a) Thermal-infrared image of summit, July 21, 1981, 0429 PST. Polarity: white=warm. black = cool. Scale bars show limited distortion. Digitally recorded image is by EG & G Corporation using a Daedalus DS- 1260 scanner recording in the 8-12 pm spectral band. Image courtesy of H.H. Kieffer, U.S. Geological Survey. (b) Map of 198 1 thermal pattern (solid). Lettersdenote areas of greatest thermal activity including: A = inner north slope of East Crater rim; B = inner northwest slope of East Crater; C= crest and outer northwest slope of West Crater rim; D = outer southwest slope of West Crater; E = thermal area at slump scarp on outer northwest slope of West Crater. Outline - summer D. Frank/ Journal of Volcanology and Geothermal Research 65 (1995) 51-80 59

East Crater Rim

Crater-fill debris

B’

A L I I 1 A’ 0 10 20 30 meters

Fig. 6. Cross-sections A-A’ and B-B’ and temperature-probe traverses acrossthe rim of East Crater. Surface (dashed) and H-cm-deep (solid) ground temperatures and fumarole (triangles) temperatures are shown in relation to two schematic profiles. See Fig. Sb for location of profiles.

Summit fumaroles are typically small, mostly a few sound but are not highly pressured. The warmest fuma- centimeters across, but some are up to 20 cm. The vents roles have sparse patches of moss and liverwort around with the greatest discharge make an audible hissing their orifices. Fumaroles are commonly situated along 60 D. Frank/Journal of Volcanology and Geothermal Research 65 (1995) 51-80 margins and joints of isolated andesite blocks embed- which reported values include - 10°C in 1959 (Miller, ded in the clayey to sandy matrix of the explosion 1970) and -4°C in 1971 (Kiver and Steele, 1972). rubble or in joints at the top of the highly fractured lava flows that crop out at the crest and inner slopes of the 3.2. Upper&nk crater rims. In the East Crater, the largest and most noticeable Several small areas of thermal activity have been fumaroles occur near the top of a single andesite flow observed on the upper flank of Mount Rainier (Fig. 7). that intermittently crops out on the inner slope of the These occurrences contrast with other Cascade Range crater rim (Fig. 6). Typically, these fumaroles are at volcanoes where most activity is confined to sites of or near the contact of the jointed lava with overlying old eruptive vents, such as central craters or margins of clayey rubble which effectively seals off the adjacent domes. Table 1 lists selected references to reported flow path for vapor. This cluster of activity produces areas of flank thermal activity and makes note of the the long thermal lineament noted by Moxham ( 1970, type and number of observations. Close attention to the p. 86). Concentric thermal patterns occur both above source of such data is important for Mount Rainier and below the andesite flow as seen in Fig. 5A. In because of common reports each year of “steam vents” contrast to the fumarole clusters in the lava flow, these that are in fact non-volcanic, such as wind-blown snow patterns result from hot unvented ground associated or dust plumes. with the clayey rubble (Fig. 6). Table 1 includes events with a relatively high weight Similar jointed flow and clayey rubble occurrences of evidence, either visual observation within 1 km or respectively localize fumaroles and areas of unvented observation by thermal-infrared surveys. For localities identified only by instrumental means, Table 1 includes hot ground at the crest and inner slope of West Crater. five that produced an image density (infrared bright- Outward slumping, however, has controlled the ther- ness) equal to or greater than that of the Disappoint- mal pattern to some extent on the outer slope of the ment Cleaver fumarole field discussed below. West Crater (E in Fig. 5B) by forming hot scarps in Disappointment Cleaver serves as a control locality for the sheath of flows that make up the summit cone. interpretation of infrared anomalies because it is the Subglacial clusters of thermal activity also extend only upper flank thermal area having detailed field down the inner slopes of both crater rims and, in com- measurements. bination with rivulets of meltwater and circulating air Localities noted in a number of other references to currents, form caves within the glacier cover of the thermal activity on the flanks of Mount Rainier are crater basins. Mapping by Kiver and Mumma ( 197 1) shown in Fig. 7, but are not included in Table 1 because shows a concentric pattern of ice-cave development they were based on visual observations made at great 50-80 m inside the crater rim. The pattern suggests distance (more than 5 km) and were not instrumentally localization of thermal activity along a concentric zone verified. Of note are common late-summer observa- of relatively high permeability, perhaps inward-dip- tions of dust plumes and perhaps water vapor in ero- ping faults, vertical tension cracks, or jointed flows in sional alcoves at about the 3570 m level of the South the enclosing rim. Tahoma headwall and the 3440 m level of Curtis Ridge Kiver and Mumma ( 1971) measured subglacial (SH and CR in Fig. 7). Both localities are the source ground temperatures as high as 86°C and vapor tem- of frequent small- to moderate-size ( lO,OOO-100,000 peratures at 56°C in East Crater caves, whereas Kiver m3) debris avalanches, the most recent in 1992 from and Steele ( 1972) measured vapor temperatures up to Curtis Ridge (Norris, 1994). Observations of the South 76°C in West Crater caves; these measurements are 4- Tahoma headwall are discussed in detail by Crandell 5°C higher than maximum reported vapor temperatures ( 1971, pp. 60-63) who concluded that small steam at the surface and reach the boiling point of pure water explosions could have occurred there during late sum- at the crater altitude. The existence of a lake in the West mer 1969. Crater, though only 0.5”C (Lokey, 1973) suggests sig- Other sites where short-lived thermal activity or nificant thermal activity beneath the ice cover when steam explosions have been considered as possible compared to the subfreezing temperature of the ice, for causes for rock-fall include the north slope of Little D. Frank/Journal of Volcanology and Geothermal Research 65 (1995) 51-80 61

Fig. 7. Map of reported thermal activity on the flank of Mount Rainier. Continuous hydrothermal emission (squares) occurs in those areas which have been detected more than once by instrumental observations over a span of several years at Disappointment Cleaver (DC) and probably Sunset Amphitheater (SA). Possibly discontinuous or ephemeral activity (circle) occurs in those areas which were detected by isolated instrumental or short-range visual observations at Willis Wall ( WW), South Tahoma headwall (.SH), Kautz headwall (HZ), Gibraltar Rock (CR), and lngraham Glacier (IG) Unverified thermal activity (triangles) has been reported, in several source areas of debris avalanches at Curtis Ridge (CR), Gibraltar Rock, South Tahoma headwall, South Tahoma Glacier (SC) and Little Tahoma Peak (U”). S= summit.

Tahoma Peak (LT, Fig. 7) at 3000 m in December Weak anomalies recorded by ground-based thermal- 1963 (Crandell and Fahnestock, 1965, pp. A25-A29; infrared surveys (Lange and Avent, 1973, 1975) are Fahnestock, 1978), the icefall at 1980-2130 m on the not included in Table 1 because of difficulties in inter- lower South Tahoma Glacier (SC) in February 1969 pretation that arise from the small radiant temperature (Crandell, 197 1, p. 62)) and cliffs in the middle of the differiences (less than 3K) and great distances (6-12 Emmons Glacier (EG) at 3,000 m in 1969. Additional km) involved in these surveys. Perhaps significantly, data are required to determine if any of these sites have however, one of Lange and Avent’s (1973, Fig. 2) had either continuing or intermittent thermal activity. infrared anomalies coincides with a prominent ava- 62 D. Frank/Journal of Volcanology and Geothermal Research 65 (I 995) 51-80

Table I Thermal areas on the upper flank of Mount Rainier

Locality Altitude Date Type of observation Ref.’ (m/ft)’

Disappointment 36301 I 1,900 I l/15/72 Aerial IR survey 2 Cleaver, DC 4129173 Aerial IR survey 12113175 Aerial IR survey 19741982 Field observation

lngraham 36301 I.900 -l/23/.56 Ranger report of strong sulfur fume from Glacier, IG ice pit, 9 m across, 30 m deep, no noise

Gibraltar 34101 1,200 S/61 Guide report of steam fissure at NE end of 3 Rock. CR Gibraltar shooting 20-30 m for several days

Kautz 35701 1 1,700 317165 Climber report of steam vent I Headwall. KH 3570-3600 4129173 Aerial IR survey, 2 sites 2 /11,700-l 1,800

South Tahoma 3350-3410 4/29/73 Aerial IR survey, 2 sites 2 Headwall. SH /I l,OOo-11,200

Sunset 3930112,900 4/29/73 Aerial IR survey 2 Amphitheater, SA 3900/ 12,700 7121181 Aerial IR survey s

Willis Wall. 3780/ 12,400 4129173 Aerial IR survey, 2 WW 12113175 Aerial IR survey 4

‘Approximate altitudes are also provided in feet, the unit used in most the original reports. ‘References: I =Danes (1965); 2=Frank and Friedman (1974): 3=Moxhamet al. (1965); 4=Rosenfeld (unpubl.); S-Frank (1985). lanche-source area in the South Tahoma headwall (SH, time. Maximum temperatures show a decreasing trend Fig. 7). of I-2°C per month from July to November which is probably a seasonal effect from cooler weather and Disappointment Cleaver perhaps changes in groundwater recharge from snow- Two clusters of small, weak fumaroles occur along melt. Superimposed, short-termdownward fluctuations the ridge of Disappointment Cleaver at about 3630 m may result from passing storms. Additional field meas- altitude; 320 m* of heated ground surrounds the fuma- urements in August 198 1 and June 1982 yielded similar roles and extends 30 m along the ridge crest and 15 m temperatures of 58°C for the vapor and 5760°C for down the south facing slope of the cleaver The thermal nearby heated ground, verifying a lack of significant activity occurs at and below the contact between a grey long-term change. andesite flow and an overlying zed ftowabyeccia. The contact zone contains a mixture o$poorIyansolidated Other headwall thermal areas block and cinder rubble which is variabIy cemented by Thermal activity elsewhere on the upper flank of opal and calcite. Mount Rainier may be ephemeral or highly variable. A spring-driven Foxboro recorder with a vapor-liq- For example, observations of apparent fumarole emis- uid temperature probe was installed at the largest, hot- sion in the Ingraham Glacier in 1959, at Gibraltar Rock test fumarole on Disappointmat Cleaver in July 1974 in 1963, and on the Kautz headwall in 1965 (Table 1) and run intermittently for 159 days duriirg the following have not been repeated even though all of these areas two years. Recorder data for a X)-cm depth (Fig. 8) occur within 1 km of popular climbing routes. Subse- show a range of 3%62°C. Thermal peaks of 5442°C quent infrared surveys detected an anomaly at only one occur in several episodes of a few days duration at a of these sites, the Kautz headwall in 1973. 63

70 r

June July August September October November

Fig. 8. Graph of temperature variation at 50-cm depth in Disappointment Cleaver thermal area. Daily maximum and minimum temperatures taken from a continuous record produced by a Foxboro recorder show thermal peaks at 54-62”C. Downward fluctuations probably result from cooling by shallow infiltration ofsnowmelt.

No known visual observations of thermal activity Structural imjdicptions have been made at the more poorly accessible sites in The upper-flank thermal areas in Table 1 all lie Table 1, the South Tahoma headwall, Sunset Amphi- within a remarkably restricted range in altitude of about theater, and Willis Wall. All have been detected by 3400-3900m. Such a relationship suggests localization infrared surveys, although only Willis Wall and Sunset of fluid emission in these areas by common structural Amphitheater more then once (Table 1) . On the basis or hydraulic controls. All of the upper-flank thermal of multiple detection across a span of several years, areas OCCURwhere outward-dipping flows, or flow brec- Sunset Amphitheater is considered to have probable cias, have been truncated by headwall erosion. Per- continuing activity and is so noted in Fig. 7. meable contacts betiveen these layers, such as that observed at Disappointment Cleaver, would provide paths for fluid transport. Other possible permeable 64 D. Frank/Journal of Volcanology and Geothermal Research 65 (1995) 51-80 structures such as flank eruptive vents, dome margins, and meadow below, and eventually discharge into the dikes, cone sheets, or inward-dipping faults are absent West Fork White River along the west margin of the at most of the flank thermal areas, except for Sunset Winthrop Glacier terminus. A significant amount of Amphitheater. additional seepage also occurs through the talus. Cross sections (Fig. 9) through the flank thermal areas indicate that, with the exception of Sunset Amphi- Paradise Springs theater, the loci of activity occupy comparable posi- Numerous thermal springs and areas of warm seep- tions relative to the central vent of the volcano. These age ranging in temperature from 11 to 25°C occur along relations are consistent with migration of upper-flank a 100-m segment of the west valley wall of the Paradise thermal fluids downward and outward from the central River between 1950 and 1980 m altitude. The main vent along permeable contacts between flows. As dis- springs issue from joints in thin hypersthene andesite cussed below, helium concentrations in fumarole vapor flows that are exposed in the valley wall. The zone of at Disappointment Cleaver support this conclusion. seepage, marked by abundant moss and algae, occurs Sunset Amphitheater differs from other upper-flank just above a layer of well-packed interflow rubble near localities in that it contains a number of vertical struc- the base of the hillslope and from rock-slide debris and tures adjacent to the thermal areas, including dikes, lateral moraine halfway up the slope. Spring discharge buried crater-wall contacts, and faults. Thermal fluids of over 5 L/s flows down the valley wall and enters the at Sunset Amphitheater may be channelled through the Paradise River. During the early part of this century the vertical structures and therefore rise directly from depth Paradise Glacier extended downvalley from the (Fig. 9). springs; their warm effluent may have been a factor in forming tunnels through the glacier (Evans, 1930) 3.3. Lowerjlank which then developed into the once-spectacular Para- dise Ice Caves. The warmest spring temperatures found Two clusters of low-temperature thermal springs today are comparable to earlier measurements of 24°C occur on the lower flank of Mount Rainier at about by Evans ( 1930). 2000 m altitude. References to their activity may be found extending back many decades. Available evi- 3.4. Adjacent to base of cone dence indicates little variation in spring activity during this century. Two clusters of thermal springs occur within Mount Winthrop Springs Rainier National Park, but beyond the volcanic edifice. A cluster of at least seven thermal springs ranging in Both areas are easily accessible by road and at one time temperature from 9 to 17°C issues from a 900-m stretch supported tourist spas. Comparison of present spring of hillslope at 2040-2070 m altitude along the west temperatures with data in older reports (Stearns et al., valley wall of the Winthrop Glacier. A compilation of 1937; Majors, 1962; Korosec, 1980) indicate little historic newspaper accounts (Majors and McCollum, change in activity of these springs. 1981a, pp. 442443) revealed that these springs had been observed as long ago as 1894, when a winter Longmire Springs expedition led by E.S. Ingraham camped at the spring Longmire Meadow is a shallow, poorly drained basin site and reported a maximum temperature of 16°C. of 22,000 m2 that is bisected by a small cold stream Most of the springs flow from a moss-covered boul- and perforated by numerous small springs with flows der and cobble talus deposit at the base of a lava cliff and temperatures which range up to 0.1 L/s and 28°C of olivine andesite. The largest and easternmost spring respectively. Low travertine and tufa mounds built up issues from algae-coated joints in the lava cliff. Total by the mineralized water are scattered around the spring flow is about 3 L/s on the basis of timed bucket meadow, with the largest mound in the west-central measurements for the largest springs and extrapolation sector. Water is ponded locally by travertine and at to the smaller springs. Streams fed by the springs com- times in the northern (upstream) part of the meadow, bine with larger snowmelt streams, flow through talus by beaver dams. D. Frank/ Journal of Volcanology and Geothermal Research 65 (1995) 51-80 6.5

VW SA -3 km

I WN KH SH DCSA 1 2500 ’ I

1 I I I 0 1 2 3 km

Fig. 9. Overlaid cross-sections through the upper part of Mount Rainier showing spatial relationship of thermal activity (triangles) near the summit (S) and on the upper flank at Disappointment Cleaver (DC), Willis Wall ( WW) , Sunset Amphitheater (SA), South Tahoma headwall (SH) , and Kautz headwall ( KH) Upper flank thermal areas lie within a restricted region between 3350 and 3950 m altitude and 1.4-Z. 1 km from the summit. Recharge occurs from percolating snowmelt. Arrows show possible discharge paths of subsurface thermal fluids, with most upper flank thermal areas fed by downward and outward transport from beneath the summit fumarole field along permeable zones in or between outward-dipping flows and breccias. Only Sunset Amphitheater, clearly anomalous in position, contains numerous vertical structures capable of providing significant upward transport of fluid. Inset shows position of sections.

The meadow lies on a low terrace at 835 m on the ante of inflowing and outflowing water in the main floor of the Nisqually River valley (Fig. 2). Volcanic stream traversing the meadow, was about 21 L/s in late mudflows from Mount Rainier built the terrace within summer 1982. By comparison, Mariner et al. (1990) the last few hundred years (Crandell, 1971, p. 43). used the chloride inventory method to estimate 12 L/s Tertiary volcaniclastic rocks of the Ohanapecosh For- of thermal water discharge downstream to the Nis- mation (Fiske et al., 1963) are adjacent to and probably qually River. underlie the terrace. Springs in the southeast part of the meadow have Ohanapecosh Springs moderate gas discharge but very little water flow, Thermal springs occur on the east side of the Ohan- whereas those in the western part, particularly around apecosh River from river level at 560 m altitude to the largest travertine mound, have larger flow and about 30 m above the river. The springs have formed a higher temperature. The center of upwelling thermal lobate fan of travertine and tufa terraces that extend water is therefore probably beneath the large travertine down to the river bed over an area of about 40,000 m2, mound. Total spring discharge, estimated from a bal- although over half of the fan is now forested and no 66 D. Frank/ Journal of Volcanology and Geothermal Research 65 (1995) 51-80

Table 2 Heat-flux at Mount Rainier compared to estimates for other Cascade Range volcanoes. Values are rounded to one or two significant figures

Volcano Hydrothermal Area of thermal Average excess Ref’ energy yield features’ heat flow (MW) (m*) (W/m*)

Rainier Summit craters 8.4’ 12,000’ 700 7 Upper flank fumaroles I .O’ 3,000 300 I Lower flank springs 0.9 7 Baker Sherman Crater, 1972 11 8,800 I.300 3 Sherman Crater, 1975 81 35,200 2,300 3 Dorr Field, 1972 1.2 3,000 400 3

Hood, 1972, 1977 Crater Rock/Steele Cliff 9.8 9,700 1.ooo 5

Mt. St.Helens Old dome,1972 0.2 1,000 200 1 New crater and bulge, May 16, 1980 30-75 30,000 1,OOO-2,500 637 New crater and dome, August 13, 1980 4,00&l 0,000 800,000 5,000-13,000 4.7

Lassen Volcanic National Park Lassen Peak, 1972 0.3 900 300 7 Bumpass Hell, 1972 55 45,000 1,200 2 Devils Kitchen, 1972 41 41,000 1,100 2 Boiling Springs Lake, 1972 31 16,000 2,000 2

‘Not counting areas of subglacial activity. Addition of these areas raises estimates to 8.6 MW at 300 W/m* over 28,000 rn’. ‘References: 1 = Friedman and Frank ( 1977) ; 2 = Friedman and Frank ( 1978) ; 3 = Friedman and Frank ( 1980) ; 4 = Friedman et al. ( 1981) ; 5=Friedmanetal. (1982);6=Kiefferetal.~(1981);7=Frank(1985). 3Probable maximum value. Range 0.2-l .OMW. longer has any apparent discharge of thermal water. chloride inventory method, suggesting that a substan- The fan covers talus and river deposits which lie over tial fraction of thermal discharge occurs through the volcaniclastic rocks of the Ohanapecosh Formation. river bed. The nearest lava flows from Mount Rainier cap ridge- tops 7 km to the northwest of Ohanapecosh Springs. 3.5. Heatflow Springs with the highest flow and temperature, up to about 0.5 L/s and 50°C occur near the fan’s apex, which Thermal discharge from the upper part of the hydro- is probably the center of upwelling thermal water. thermal system was quantitatively assessed in Frank Other springs are scattered about the fan, particularly ( 1985). Using the concept of heat balance at the earth’s in old excavations for former resort buildings and an surface (Budyko, 1956) in a technique described by unfinished pool as well as along the river bank where Friedman and Frank ( 1980), excess geothermal flux erosion has eaten into the fan’s toe. A few additional for the summit fumarole field, not counting subglacial thermal springs with lower temperatures occur in mar- activity, was estimated to be 8.4 MW with an average ginal areas within 400 m of the travertine fan, and flow of 700 W/m2 from the thermal area of 12,000 m* scattered travertine remnants occur on the west bank of (crater rims). Counting subglacial activity, flux is 8.6 the river. MW at 300 W/m* over 28,000 m2, or 16 W/m* aver- Total thermal water discharge above river elevation aged over the whole summit crater area of 520,000 m*. is estimated at 4 L/s. However, Mariner et al. ( 1990) Comparison with similar flux estimates at other Cas- reported a discharge of 14 L/s to the river based on the cade Range volcanoes (Table 2) suggests that Mount D. FrankIJoumal of Volcanology and GeothermalResearch 65 (1995) 51-80 67

Rainier has a modest repose-period discharge through 1OAvariety in a few samples, but is generally observed its central vent, comparable to that at Mount Hood but as 7A-halloysite. Both di- and trioctahedral smectite spread over three times the vent area of Hood. The occur, with the dioctahedral form more common. Trace large area of heat emission relative to that at other amounts of possible mixed-layer clays are evident. Cascade volcanoes suggests the absence of an imper- A silica phase accompanies clay minerals in the clay meable cap on the hydrothermal system. fraction of many samples. This is generally opal-CT Heat flux estimates for other areas of the volcano (Jones and Segnit, 1971) which shows poorly defined yielded much lower values in comparison to the summit cristobalite and tridymite XRD peaks, in some samples (Table 2). Upper flank fumarole areas (based on a superimposed on a broad opal-A XRD hump. Silt-sized total of five localities) yield about 1 MW, and the two fractions of many of the clay-rich samples contain well- lower flank thermal spring areas release a similar ordered cr-cristobalite, or in fewer cases, tridymite. amount of heat. Undiscovered thermal activity or pos- The most conspicuous difference in mineralogy sible activity in active avalanche-source areas (trian- between East and West Craters is the presence of abun- gles in Fig. 7) could significantly alter the heat-flux dant alunite and gibbsite at West Crater thermal areas, estimate. particularly in the silt fractions. Trace amounts of boehmite are found at West Crater and probably result from minor alteration of gibbsite. Alunite and gibbsite 4. Alteration mineralogy are commonly associated with one of the kaolin minerals, but all three phases were never found together in 4.1. Swjicial alteration the same sample. Gibbsite and alunite were not found in the East Crater. Secondary minerals at the summit craters, identified At Disappointment Cleaver, calcite and opal-A occur by thin-section and X-ray diffraction (XRD) analysis throughout the active part of the thermal area as white (Frank, 1985), occur primarily in the core areas of to beige encrustations and as vesicle and fracture fill- present thermal activity. In adjacent areas of explosion ings. Chalecedony and tridymite, perhaps as a result of rubble, not now affected by high temperatures, the only aging of opal-A, occur as encrustations in areas no secondary minerals found were coatings and vesicle longer bathed by heated fluids. Rubble in the hottest and fracture fillings of white fibrous or saccharoidal part of the thermal area contains pockets of clayey calcite, and colorless to white earthy or saccharoidal debris whose main constituent is dioctahedral smectite. opal. No alteration of primary minerals was identified at Toward the centers of thermal activity, calcite and any of the thermal spring deposits, though precipitates opal coatings are more abundant, andesite fragments occur at all four spring localities. At Winthrop Springs, are progressively more bleached along alteration rinds, white powdery to saccharoidal crusts of calcite, opal- and pink, yellow, red, green, and purple clay-rich A, and gypsum coat cobbles downstream from the patches of altered rubble increase in number and extent. spring outlets. Similar but more abundant crusts of cal- Deposits richest in clay coincide with the hotter areas cite, aragonite and opal-A occur at Paradise Springs, shown on Fig. 5. In the East Crater, these include the with opal making up a greater proportion of the precip- north and northwest inner slope of the rim, both above itate. and especially below the prominent lava flow. At the Deposits of travertine at Longmire and Ohanapecosh West Crater, clay-rich alteration occurs mainly at the Springs range from dark gray to light tan and are com- northwest crest and outer slope of the rim. posed mainly of fibrous, concretionary, and platy or Clay minerals typically occur in the once glassy granular calcite. Aragonite and minor rhodochrosite groundmass of flow rock and lithic rubble, and as an also occur in Longmire travertine. Warm, algae-rich alteration product of the fine-grained matrix of uncon- pools in and adjacent to the main travertine mound at solidated surficial debris. Smectite and kaolin are the Longmire contain a black aragonitic mud with pyrite most common clay-mineral groups in the fumarole in the fine-grained fraction. Stream channels that drain areas. Kaolin occurs either as disordered kaolinite or Longmire meadow are coated with bright orange amor- halloysite. The more widespread halloysite is of the phous iron oxyhydroxides. 68 D. Frank/ Journal of Volcanology and Geothermal Research 6.5 (199.5) 51-80

4.2. Subsugace alteration atively small amounts of highly altered deposits less than 2200 years old that do contain the smectite, kaolin, Lithic eruptive products and mass-movement depos- alunite assemblage represent new exposures of vent its during the Holocene have incorporated large masses material older than 5000 years in Sunset Amphitheater of hydrothermally altered rocks from Mount Rainier and the South Tahoma headwall. The bulk of this mate- (Crandell, 197 1) . Hydrothermal minerals in these rial probably remains in place in the upper part of the deposits reflect both surface and subsurface alteration cone. in the cone. Hydrothermal minerals identified in the larger deposits should represent deeper samples of the 4.3. Hydrothermal environment at depth volcano’s interior allowing characterization of the hydrothermal system during earlier time periods. Secondary minerals reflect chemical and thermal Extensive hydrothermal alteration in the upper part conditions at the time of their formation. Laboratory of the cone about 5000 years ago is represented by the experiments (Hemley et al., 1969) and alteration prod- alteration assemblage in the largest Holocene deposit, ucts found in drillholes at a variety of geothermal sys- the Osceola Mudflow, estimated by Scott et al. ( 1992) tems around the world (Browne, 1978) show that the at 3 km” with 2-l%% clay. The fine fraction of the alunite and kaolinite assemblage occurs under acid sul- Osceola contains dioctahedral smectite, disordered to fate conditions (typically pH O-5; sulfate greater than well-ordered kaolinite, talc, a trace of illite, and several 100 mg/L) at temperaturesup to a few hundred degrees silica phases dominated by chalcedony. Tephra Layer Celsius. In general, higher temperatures within the F (0.025 km”, Crandell, 1971)) erupted at about the range of alunite stability favor less acidic conditions. same time that the Osceola Mudflow occurred (Mul- Higher temperatures also favor more structurally lineaux, 1974, pp. 66-67)) contains a lithic component ordered kaolinite. Thus the disordered kaolinite, hal- with the same suite of secondary minerals. The coeval loysite, and alunite found at Mount Rainier’s summit Paradise Lahar (0.1 km3, l-6% clay, Scott et al., 1992) indicate a relatively low temperature, acidic environ- from the south part of the summit, is generally similar ment which is appropriate for mineral formation in in clay composition to the Osceola, but also contains boiling-point, acid sulfate fluids. an abundance of distinctive white clots of alunite. In older mass-movement deposits such as the Osce- The youngest mass-movement deposit containing a ola Mudflow and Van Trump Lahar, kaolinite is well large fraction of hydrothermal minerals is a debris ava- ordered and suggests a temperature of formation above lanche (less than 0.001 km3) that traveled just beyond that of surface boiling. Although aging of disordered the Tahoma Glacier terminus sometime during 19lO-- kaolin may produce ordered kaolinite, the presence of 1930 (Crandell, 1971, p. 17) from a hydrothermally illite, pyrophyllite, and talc in these same deposits also altered plug now exposed in Sunset Amphitheater. This suggest a higher temperature and hence deeper envi- bright yellow to orange deposit contains abundant ronment of formation. The large fraction of clayey trioctahedral smectite, kaolin, jarosite, alunite, and sil- hydrothermal products in the Osceola Mudflow is con- ica phases, with lesser illite, gypsum, and pyrite. sistent with alteration at extensive depth, perhaps as The mineralogic record based by Frank ( 1985) on much as 900 m, which is the estimated thickness (Cran- samples from 11 Holocene tephra and mass-movement dell, 1963a, pp. B 138-B 139) of the source area for the events indicates that an extensive area of alteration Osceola. existed 5000 years ago at a central vent. Subsequently, but mostly prior to construction of the young summit cone 2200 years ago, lesser amounts of altered material 5. Fumarole and thermal spring composition were incorporated into avalanches and mudflows. If extensive areas of alteration were present 2200-5000 5.1. Fumaroles years ago, they may have remained sheathed by relatively unaltered rock or they may have been enclosed A few literature references to sulfide odor exist for within a caldera or deep landslide trough such as must fumaroles in the West Crater, for example that by Kiver have existed following the Osceola Mudflow. The rel- and Mumma (1971) in 1970, and suggest that H$ D. Frank / Joumal of Volcanology and Geothermal Research 65 (I 995) 51-80 69 may currently be an ephemeral component of fumarole 5.2. Them1 springs vapor. No reports of sulfide gas from the East Crater, however, have been found. Water samples of the four sets of thermal springs Two East Crater fumaroles were sampled in dupli- were collected during summer 1982 and 1983 (Frank, cate for bulk gas analysis in 1982 using flow-through 1985) and analyzed using standard field and laboratory IOO-ml glass flasks (Frank, 1985). Gas methods. Data from samples at Winthrop and Paradise chromatography analysis showed dry-gas composi- springs represent the only compositional information tions in these samples to be similar to air, but with on thermal waters that issue from andesite flows of somewhat elevated levels of CO, at 1.5-4.2% and Mount Rainier. Older analyses for Longmire and Ohan- helium as high as 9 ppm. Sulfur species were below apecosh Springs, on the other hand, can also be found detection level (0.5 ppm) in all samples. The very weak in the following reports: Korosec (1980) for 1978- fumarole discharge and the opportunity for shallow 1979 data, Mariner et al. (1982) and Nehring et al. entrainment of air into the fumarole system through ( 1979) for 1977 data, and Campbell et al. ( 1970) for fractures in exposed lava flows (see Fig. 6) suggests 1969 data. that these samples, though highly air-contaminated, are For each locality, two or three sites with the highest representative of fumarole vapor. temperature and discharge were selected for sampling. Eight fumaroles from East and West Craters in 1982- In addition to thermal waters, adjacent cold waters from 1983 and one from Disappointment Cleaver in 1981 springs or streams were also sampled to provide rep- were sampled for helium analysis using lo-ml-syringe resentative shallow diluent water that could potentially injection into evacuated test tubes (Frank, 1985). Mass mix with and alter the composition of the thermal water spectrometry analysis showed somewhat enriched val- prior to discharge. A total of 20 thermal samples from ues for the summit craters of 7.5-8.9 ppm and depleted 10 sites and 8 cold samples from 6 sites were collected values of 3.3-3.4 for Disappointment Cleaver. during the two-year period. Table 3 lists paired 1982 Measurements of condensate pH were also made of data for the warmest spring sampled at each locality several fumaroles in 1983 at both summit craters by along with an adjacent cold water. More complete anal- inserting pH paper in the vent openings. Values were yses are listed in Frank ( 1985). typically between pH 5 and 6, which at 70-80°C is slightly acidic (neutral pH is 6.3 at 8O’C). Major constituents These gas results suggest that fumarole vapor is All of the spring waters are near-neutral or slightly largely recycled air plus plus vapor from snow and ice acidic with pH ranging from 5.6 to 7.3. Most dissolved meltwater which has been slightly charged with CO, constituents increase in concentration with tempera- and He. Except for the latter two gases, the fluids have ture. Total dissolved solids increase from moderate val- probably had little chance to undergo extensive reac- ues of about 270 and 850 mg/L, respectively for tions with rock and consequently have probably not Winthrop and Paradise springs up to 3300 and 5200 traveled to any great depth. If magmatic gases are pres- mg/L for Ohanapecosh and Longmire Springs. Cold ent in the Mount Rainier hydrothermal system, they water adjacent to the thermal springs contains less than must be largely going into solution or reacting with 70 mg/L, except for the cold stream at Ohanapecosh minerals at depth. In contrast to the helium-enriched which apparently contains a small component of ther- fumaroles at the summit, helium-depleted samples mal water. from Disappointment Cleaver are consistent with a Major constituents, as summarized in a Piper (1944) mechanism by which heated fluids beneath the summit diagram for all thermal analyses (Fig. lo), show a thermal field boil, lose helium, and travel downward distinct grouping for each of the fourclusters of springs. and outward toward the flank of the cone as suggested The thermal-water types classified according to species by the distribution of upper flank thermal areas in Fig. greater than 50% abundance include Na-SO., water at 9. Winthrop Springs and Na-Cl water at Ohanapecosh Springs. More complex water types according to species greater than 30% abundance include Na-SO,- HCOs water at Paradise Springs and Na-Ca-HCO,-Cl 70 D. Frank/ Journal of Volcanology and Geothermal Research 65 (1995) 51-80

Table 3 Selected analyses for thermal springs and adjacent cold water during 1982. Winthrop springs: WS 1 = southeastern main spring; WS3 = cold spring southwestof WS-1; Paradisesprings: PS-5 = lower main spring; PS-3 = south cold spring; Longmire Springs: LON = main southwestern spring; LOC = main inflow stream; Ohanapecosh Springs: OHA = main spring; OHC = eastern cold stream.

Site: ws-1 ws-3 PS-5 PS-3 LON LOC OHA OHC Date: 820906 820906 820904 820904 820904 820904 820905 820905 Time: 1100 1200 1540 1630 0930 0830 1130 1030

K 5.1 0.2 16 0.9 46 0.80 48 1.6 Na 28 0.8 100 2.8 640 2.9 950 40 Li 0.027 0.005 0.15 0.011 2.2 0.011 3.1 0.13 Ca 9.4 0.40 49 1.7 550 4.6 64 18 Mg 8.6 0.20 36 0.81 170 0.79 5.1 1.8 Mn

PH 7.1 5.3 6.2 5.1 6.1 5.6 6.5 5.5 Eh (0.44) (0.56) (0.35) (0.49) (0.17) (0.39) (0.11) (0.49) SP CD 310 9.2 980 33 5700 43 4400 330 T 14.0 9.0 24.0 2.0 19.5 8.0 48.0 6.5

IS 0.0030 0.00018 0.014 0.00052 0.084 0.00069 0.049 0.0035 IB (%) 1.9 -45 -0.7 -64 1.8 -7.2 0.1 - 1.8

Dissolved constituents in mg/L; Eh in volts, estimated from 1983 field values; SP CD in microsiemens; T in degrees Celsius. IS = ionic strength. IB = ionic balance. HCO,-titration alkalinity as HC03. ( ) estimated value. water at Longmire Springs. In contrast, cold water adja- the Mean Meteoric Water Line of Craig ( 1961)) show- cent to the thermal springs is a Ca-Na-HC03 type at ing no appreciable oxygen shift as a result of interaction Longmire and Na-HCO, at the other areas (Table 3). with rock, or hydrogen shift as a result of evaporation. The chloride content of the thermal waters provides Where rock-interactive or evaporative effects are evidence of near-surface mixing with cold groundwa- absent, stable isotopes may provide information on ter. Large decreases in chloride with temperature sug- recharge altitudes. Stable isotopes in atmospheric water gests substantial mixing with low-chloride cold water vapor undergo temperature-dependant fractionation at Winthrop, Paradise, and Longmire Springs. Ohana- when water is removed by precipitation. The amount pecosh Springs, on the other hand, has minor chloride of fractionation varies with changes in air temperature variability with temperature among the warmer springs, and, thus, with altitude. indicating cooling primarily by conduction rather than The isotopic data from Mount Rainier may be used by mixing. to estimate an altitude-composition relationship for cold recharge waters, and from that derive a recharge Stable isotopes altitude for the thermal waters. Since cold waters sam- Hydrogen and oxygen isotopic composition was also pled near the thermal springs represent a mixture of examined in 1982 samples (Table 4) to provide infor- local snowmelt runoff and groundwater discharge, their mation on the extent of rock-water interaction and isotopic composition represents recharge values from evaporation, and on the recharge altitude of spring a range of slightly higher altitudes that are here esti- water. All of the thermal and cold waters closely follow D. Frank/Journal of Volcanologyand GeothermalResearch 65 (1995) 51-80 71

variation remarkably close to the average variation of - 3 %&m 6l*O determined by Burk ( 1979, pp. 69- 71) from analyses of precipitation at 300 to 1,200 m altitude at Pack Forest, Nisqually Entrance, Longmire, and Nisqually Bridge. Higher-altitude samples of firn show a variation of about - 9 %olkm al80 and - 65 %o/km 6D. The latter compares favorably with - 50 %o/km 6D for snow and firn from 1400-2200 m at Blue Glacier, Washington (Sharp et al., 1960, p. 4047), and a range of -20 to -60 %olkm for fresh snow from 800-3300 m in the Mont Blanc Alps (Moser and Stich- ler, 1970, p. 46). Keeping in mind the possible differences between available data and true mean annual isotopic composition, Fig. 11 suggests that all thermal springs have been recharged at higher altitudes than nearby cold water. None of the thermal springs, however, are as isotopically light as the summit firn.

Fig. 10. Piper diagram of percent equivalents of major ions in thermal 6. Conceptual hydrothermal model springs (Table 3 and Frank, 1985). Winthrop ( W) and Paradise (P) springs contain a large fraction of SO,, whereas Ohanapecosh (0) and Longmire (L) Springs are primarily Cl and HCO,-Cl waters, Surface data point to a hydrothermal system having respectively. a central region of heated upflow which discharges at the summit and produces sparse amounts of lateral leak- mated by the altitude range of the enclosing local drainage at lower elevations. A geochemically-based simu- age basin (Table 4). lation that could operate within this physical At altitudes above the firn line where springs and framework was constructed by Frank ( 1985) and is perennial surface waters are absent, samples of snow described briefly here for a reaction path that would and fim were collected to estimate the isotopic com- produce measured thermal spring compositions. Major position of recharge. These samples (Table 4) were controls in the simulation are threefold: collected by excavation at three localities which include ( 1) Thermal waters are cooled by cold-water dilu- a snow pit at Muir Peak, the wall of a freshly formed tion, as suggested by the spring characteristics. crevasse at Ingraham Flats, and the wall of an ice cave (2) A hot parent fluid is acidic, as suggested by the at East Crater. The samples were melted by heating in alteration mineralogy. closed containers to prevent vapor exchange with the (3) Equilibrium thermodynamics governs reactions. atmosphere and then bottled in the same manner as the spring samples. 6.1. Possible parentjuid The rationale for using first-year fim to determine recharge altitude is that it has not moved an appreciable Deuterium-temperature relationships in thermal and distance downslope, but it has benefitted from a homog- cold waters allow estimation of the extent of mixing. enizing effect of one season of freezing and thawing. A hot-water end member, prior to mixing, should have Furthermore, first-year fim is accessible. Thus, samples a deuterium concentration no lighter than that found in from the four thermal spring areas and the three firn the lightest recharge water. From Fig. 11, the lightest localities provide a seven-point distribution of data for recharge is found at the summit and could have a SD recharge altitudes from 610 to 4330 m. as low as - 190%0.Mixing lines based on data for the Figure 11 shows the variations in isotopic compo- highest-temperature springs and adjacent cold waters sition with altitude for the two groups of samples. The at Paradise and Winthrop (Tables 3, 4)) intersect at a low-altitude samples of the cold surface waters have a temperature of 126°C and SD of - 188%0 (Fig. 12). 12 D. Frank/Journal of Volcanology and Geothermal Research 65 (1995) 51-80

Table 4 Stable-isotope compositions of snow (S), tim (F), thermal springs and adjacent coldwater

Locality Altitude Depth to fim VO 6D Date (m) (cm) (%o) (%o)

Muir Peak 2,960 200 8/19/82 Snow - 13.3 - 102 Fim - 13.2 - 102 lngraham Flats 3,410 500 8120182 Snow - 19.6 - 144 Fim - 18.8 - 138

East Crater 4,330 71 9/02/82 Snow - 14.3 -96 Fim - 24.6 - 190

Locality Altitude Altitude range of ?Y80 6D Temp (m) drainage basin (%o) (%o) (“C) (m)

Winthrop Springs WSl 2040 - 16.8 - 121 14.0 ws3 2070 2070-2210 - 16.3 -118 9.0

Paradise Springs PS5 1950 - 15.4 - 114 24.0 PS3 1910 191&2010 - 13.8 -98 2.0

Longmire Springs LON 840 - 13.6 -96 19.5 LOC 850 850-l 180 - 12.3 -86 8.0

Ohanapecosh Springs OHA 590 - 15.4 - 116 48.0 OHC 610 610-1370 - 13.0 -90 6.5

This deuterium is close to that estimated for summit each greater than 10% of the major constituents to be recharge indicating a parent water in a region expected consistent with proportions found in the thermal to have acidic fluids. Cold-water mixing fractions based springs at Mount Rainier. on deuterium are 83% for Paradise and 96% for Win- A close example to this type of water found in the throp. Cascade Range is at Mount Baker where increased The moderate sulfate and chloride content of Para- fumarolic discharge in 1975 was accompanied by a dise springs suggests that an acid sulfate-chloride type pronounced increase in the chloride content of an acid water (White, 1957) might be a reasonable parent sulfate crater lake. To construct a parent water for sim- water prior to mixing. Natural examples of acid sulfate- ulation purposes, the 1975 crater lake composition at chloride waters reveal a wide range of compositions. Mount Baker (Table 6, A) was modified by addition Table 5 lists some characteristics for a variety of these of enough chloride to yield the Paradise thermal spring waters found in either springs or crater lakes at andesitic concentration upon 83% dilution with cold water volcanoes. These examples are limited to those whose (Table 6, B). Cations were balanced primarily by magnesium, sulfate, and chloride equivalents were sodium and a small amount of magnesium. Carbon D. Frank/ Journal of Volcanology and Geothermal Research 65 (1995) 51-80 73

The reaction path simulates equilibration of an acid sulfate-chloride water with andesite in an aquifer in the upper part of the cone. No cooling occurs until the neutralized water reaches a region of mixing with cold groundwater. At this point, cooling and mixing take place relatively quickly. During cooling the fluid phys- ically moves away from the initial alteration assemblage and reequilibrates with a lower temperature assemblage. The reaction path assumes potential effects of cooling by conduction are not significant. The simulation employs an ion-association geochemical model to determine the distribution of aqueous species and the saturation state of the fluid with respect to a variety of minerals, and a mass-transfer

01 I I 10 model to find the most stable assemblages of minerals -30 -20 -10 -260 -100 which result from heterogeneous equilibrium (Helge- 6” 0 6D son, 1979). Calculations were made with the EQ3/ Fig. Il. Graph of variation of stable-isotopecomposition with alti- EQ6 computer code and thermodynamic data base of tude for thermal (open symbol) and nearby cold (solid symbol) Wolery ( 1979). water at Winthrop (W), Paradise (P), Longmire (L), and Ohana- With reference to the right side of Fig. 14, upon p-ecosh (0) Springs, and for fim. Bars on the top of cold-water symbols indicaterange of attitudefrom which wateroriginated based initial equilibration of the 130°C acid sulfate-chloride on respective drainage basins. Burk’s (1979) trend for oxygen water, the solution precipitates an acid-alteration (dashed) is based on precipitationon south side of Mount Rainier. assemblage of alunite, chalcedony, and pyrite. Reaction Trendfor hydrogenis discussed in text. with andesite in a closed system in which precipitated minerals are allowed to redissolve is simulated by titra- dioxide, like chloride, was set to reproduce the thermal tion of the equilibrated solution with a glass composi- spring concentration after 83% dilution. tion equivalent to summit andesite (Fiske et al., 1963, Through simulation calculations discussed below, p. 87). Reaction progress proceeds from right to left the resulting modified water (Table 6, B) was equilibrated with a stable mineral assemblage to produce a parent water (Table 6, C) at 130°C. The parent water 250 , I I 1000 lies within the range of the natural acid sulfate-chloride waters found at andesitic volcanoes elsewhere (Table 5, Fig. 13).

6.2. Reaction-path simulation

A reasonable scenario which might produce the lower-flank thermal springs and be consistent with available data would entail neutralization of the acid sulfate-chloride parent water with subsequent cooling Winthrop Paradise through dilution. This scenario was tested in a series of 1 ) Lib reaction-path calculations (Figs. 13, 14) by which the -&JO -200 -100 O0 parent water (Table 6, C) was allowed to react with bD andesite in a closed system. Upon reaching an appro- Fig. 12. Graphof deuterium-enthalpymixing relations for Paradise priate point in reaction progress, guided by alteration and Winthropsprings. Mixing between cold springs (solid symbols) and a hot end-memberyields projectedmixing lines which intersect mineralogy, the solution was withdrawn from the ande- at 126°C and - 188 per mil. Calculated mixing fractions of cold site and diluted with a water having the composition of water are 83%and %%, respectively, to produce observed thermal a Paradise cold spring. springsat Paradiseand Winthrop(open symbols). 14 D. Frank/Journal of Volcanology and Geothermal Research 65 (1995) 51-80

Table 5 Localities of high-magnesium, acid sulfate-chloride waters shown in Fig. 13

Locality Temp PH SO,/CI Ref’

Arayu spring, Onikobe caldera, Honshu, Japan 70 2.0 1.1 4 Copahue volcano, crater lake, Argentina 20 4.8 3 Kawah Ijen volcano, crater lake, Java 2.5 3 Kipyashcheye crater lake, Golovnin volcano, Kunashir 35 2.6 0.53 5 Lower Mendeleev volcano, spring, Kunashir 1.7 2.1 3 Mendeleev volcano, spring, Kunashir 82 1.8 0.92 2 Ogama-Megama spring, Onikobe caldera, Honshu, 97 2.4 0.59 4 Japan Osoreyama spring, Japan 12 2.2 0.55 2 Ruapehu volcano, crater lake, New Zealand 1.2 1.2 I Tamagawa spring, Japan 98 1.2 0.41 2 Yakeyama volcano, spring, Japan 88 0.4 0.42 3 Yan Ming Shan spring, Taiwan 81 1.6 2.5 3

‘References: 1 - Ellis and Mahon (1977); 2 - Weissberg et al. (1979); 3 -White et al. (1963); 4 - Yamada (1976); 5 - &otov (1967).

(Fig. 14) as indicated by decreasing water/rock ratio. a 130°C aquifer, and that the aquifer fluid can be derived As the aqueous solution is neutralized by reaction with from reaction between acid sulfate-chloride water and andesite (right half of Fig. 14)) the alteration-mineral andesite. The simulation not only predicts in a general assemblage goes through a major change from a region way the thermal-spring composition but it also pro- of acid clay dominated by kaolinite and alunite to near duces equilibrium assemblages of minerals that are neutral clay dominated by smectite. observed in Holocene deposits, primarily the acid clay Reaction with andesite is halted upon reaching the assemblage of alunite, chalcedony, kaolinite, and neutral smectite assemblage at a W/R of 182 or reac- pyrite, and the neutral clay assemblage of smectites, tion-progress value (ZI) of .056; that is after .056 chalcedony, and pyrite. Because of the basic assump- moles of andesite have reacted with 1 kg of solution. At tion of cooling by dilution, the results indicate a pos- this point, magnesium concentration in the fluid peaks sible, though not necessarily unique, reaction path. and pH reaches a plateau of 5.5 (Table 6, D). The 130°C mineral assemblage consists of chalcedony, Mg- 6.3. Flow path beidellite, minor Mg-nontronite and anhydrite, and trace pyrite. The geochemical simulation, and physical and For the remainder of the simulation (left half of Fig. chemical data suggest a conceptual flow-path model as 14)) the fluid is withdrawn from the neutral clay assem- shown in Fig. 15. A narrow, central zone of upflowing blage and diluted with Paradise cold-spring water along thermal fluids, mainly hot gases (vertical arrow, Fig. a linear mixing line from 130 to 24°C. The final mineral 15), boils shallow groundwater in the upper part of the assemblage is dominated by chalcedony with minor cone. Condensation of steam, oxidation and hydrolosis kaolinite and Mg-nontronite. The composition of the of H$, and mixing with shallow recharge above the final simulated solution (Table 6, E) compares favor- boiling zone forms a perched acid-sulfate aquifer which ably with that actually observed at the main Paradise feeds fumaroles at the summit. Similar processes may Spring, PS5 (left end of Fig. 14). Final model com- also feed thermal activity at Sunset Amphitheater. Lat- positions are within 30% of observed concentrations eral percolation of heated water (lateral arrows, Fig. except for aluminum, iron, and silica, species largely 15), accompanied by secondary boiling upon nearing affected by solubility of low-temperature amorphous the ground surface in headwalls, feeds most other upper phases not adequately accounted for by the geochemi- flank thermal areas including that at Disappointment cal model. Cleaver. Lateral transport of heated water at deeper The simulation demonstrates that high-magnesium, levels feeds thermal springs on the lower flank (Fig. low-temperature thermal springs can be produced from 15). Those at Paradise have a moderate SO,/Cl ratio D. Frank/ Journal of Volcanology and Geothermal Research 65 (1995) 51-80 15

Table 6 Fluid compositions for mass-transfer simulation. Concentrations in mg/L, T in degrees Celsius, Eh in volts, IS-ionic strength

Al.2 B C3 D E Natural acid Modified acid Parent water-B C reacted with D diluted with sulfate-chloride sulfate-chloride equilibrated with andesite to shallow ground water water alteration minerals ZI= 0.056 water, PS3, to 24C

K 24 24 5 100 18 Na 88 155 155 324 58 Ca 130 130 130 180 32 Mg 3-l 60 60 112 20 Fe 38 38 18 0.0019 5 x 10-12 Al 76 76 3-I 0.012 5 x 1o-4 Cl 28 450 450 450 79 F 4.8 4.8 4.8 4.8 0.83 SC, 1300 1300 1140 812 144 HCO, 0 0 0.19 286 352 CD3TOT ND 2700 2700 2690 508 SiO* 220 220 132 132 10 N ND 0.062 0.062 0.062 0.024

PH 2.5 2.5 2.3 5.5 5.4 Eh ND 0.20 0.16 -0.15 0.48 T 34 130 130 130 24

IS 0.053 0.033 0.031 0.032 0.008

‘Sherman Crater lake, Mount Baker, June 6.1975 (Frank, 1983, p. E21). ‘CO3ToT is the componenet species used for carbon in the geochemical model. 31nitialalteration minerals-kaolinite, alunite, chalcedony, pyrite.

and probably are derived from condensation near the If one uses an average heat content and density of lower part of the boiling zone. Springs at Winthrop andesite of 1500 J/g and 2400 kg/m3, the cooling of have high SO,+/Cl and likely originate near the upper the summit cone (estimated to be 1.2 km3) over a 12OO- part of the boiling zone. year period would produce 114 MW of steady-state excess heat. Discharged over the area of the summit 6.4. General discussion craters, this represents a heat flow of 220 W/m’, about a third of the present excess heat flow through the The present hydrothermal system at Mount Rainier summit fumarole field. Present heat flow clearly is characterized by moderate discharge at the summit requires a source other than the young summit cone. of CO,-enriched, slightly acidic fluid from a shallow, Lateral discharge on the upper flank, though not steam-heated boiling aquifer of snowmelt, by lesser abundant in comparison to the summit activity, is more lateral leakage of steam-heated fluids on the upper widespread than analogous upper-flank activity on any flank, and by sparse lateral leakage of neutral S04- other Cascade Range volcano. Transport paths of enriched, low-temperature thermal waters on the lower steam-heated fluids within the top of Mount Rainier’s flank. Since some 80% of the excess heat discharge hydrothermal system are not confined to a central vent, occurs through the central vent while historic activity but spread laterally about 2 km. The high degree of has remained at a relatively constant level, no effective headwall erosion into the upper cone may facilitate sealing of vertical permeability seems to have taken leakage of these fluids. place. The lateral extent of the hydrothermal system below Present heat flow of 700 W/m* may be compared the upper flank is limited as suggested by the minor with the thermal equivalent of the young summit cone. discharge of SO,-enriched fluids. Geochemical simu- 76 D. Frank/ Journal of Volcanology and Geothermal Research 65 (1995) 5140

values for Ohanapecosh are relatively light and suggest recharge at high altitude. In any case, both Longmire and Ohanapecosh Springs are fed by fluids which, unlike Winthrop and Paradise springs, have undergone transport through Tertiary rocks beneath Mount Rai- nier.

6.5. Hazard implications

The hydrothermal system within Mount Rainier formed clayey alteration products at depth 5000 years ago. The hydrothermally formed minerals, particularly the smectite, disordered kaolinite/halloysite, alunite assemblage in early to mid-Holocene deposits, are typical acid alteration products commonly found within the upper kilometer of active hydrothermal systems. Trace amounts of clay minerals more common to somewhat deeper, hotter parts, such as well-ordered kaolin- Percent Equivalents ite, illite, and pyrophyllite, further indicate deep, acid Fig. 13. Piperdiagram of percent equivalents of major ions in acid alteration during the mid-Holocene. Large masses of sulfate-chloride waters, in thermal springs at Mount Rainier, and in similar material are now forming at the summit to an model fluids. Winthrop ( and Paradise (P) symbols are as in Fig. W) unknown depth, and exist in the headwall of Sunset 10. A variety of acid sulfate-chloride waters from localities listed in Table 5 are shown by small dots. Tie lines connect points (large Amphitheater. Thermal springs leaking today from the dots) along the reaction path of the mass transfer simulation. lower flank of Mount Rainier suggest that acidic fluids underlie these deposits and continue to produce deep, lation suggests that the few waters of this type that do clayey alteration products. occur may be derived by lateral transport from an acid Mount Rainier has a history of frequent, large sulfate-chloride aquifer in the upper part of the cone. destructive avalanches and mudflows. Typically, those All thermal-fluid discharges appear to be greatly which contain hydrothermally altered material have affected by contamination with shallow cold ground- been the largest. The summit and Sunset Amphitheater water. areas are therefore potential sources for particularly The Cl-enriched thermal water that discharges hazardous avalanches, provided that appropriate slip beyond the base of the volcano (Fig. 15) at Longmire surfaces and triggering mechanisms are also available. and Ohanapecosh could conceivably originate in Potential slip surfaces for debris avalanches are deeper parts of Mount Rainier’s hydrothermal system, already in place at Mount Rainier. On the upper flank below the zone of boiling. Examples of similar scenar- of the cone, rubbly zones within or between outward- ios involving long lateral transport of groundwater and dipping lava flows provide not only structurally weak heat over distances of 10-20 km have been quantita- surfaces but also permeable paths for fluid transport. If tively evaluated elsewhere in the Cascade Range (Inge- hydrothermal alteration produces clay minerals along britsen and Sorey, 1985; Ingebritsen et al., 1992). The these paths, rock stability can be further diminished. composition of Longmire Springs bears similarities to An extreme condition might occur in the upper, central the unmixed end member for Paradise springs, espe- part of the cone in a crater basin where surficial clay cially in cations. If Longmire Springs has a source formation may develop and then be buried by subse- similar in temperature and composition to that for Par- quent eruptive deposits. For example, extensive surfi- adise, appreciably less cold-water mixing must take cial alteration likely occurred in a basin at the summit place. The more distant Ohanapecosh Springs, how- during the 3000-year period between the time of the ever, has a significantly different composition from the Osceola Mudflow and the formation of the young sum- other three spring localities. Nonetheless, deuterium mit cone. If so, the summit cone now sits in a basin D. Frank/ Journal of Volcanology and Geothermal Research 65 (1995) 51-80 II

Az$IJ;S 4

LOG mg/L -5

MODAL PERCENT

MINZALS

0% k.5 MS Eh 2 -.5 % MLlmoN ‘Y 5p ? m 192 wfS&cK 10000

0 I 50 100 130 130 TEMPERATURE

+ DILUTION - ANOESITE REACTION Fig. 14. Graph of variation in modal mineral and fluid composition during mass transfer simulation. Reaction between acid sulfate-chloride water and andesite proceeds from right to left and passes through an acid-clay assemblage of kaolinite, alunite and minor smectite to a near- neutral clay assemblage dominated smectite. Reaction with andesite is stopped upon reaching the smectite assemblage and the solution is cooled with shallow ground water mixing until a temperature of 24°C is reached. Large dots represent composition of Paradise thermal spring, PS5, for comparison. Mineral code: Anh = anhydrite, Kaol= kaolinite, Mg-B = mg-beidellite, Q-N= mg-nontronite. 78 D. Frank/Journal of Volcanology and Geothermal Research 65 (I 995) 51-80

-a Map View of Sections r- lOkIll

N Summit Fumaroles s shanow steam-heated aquifer pper Flank Fumaroles

1 1 1 1 1 * 1 I I I I I 1 I I 1 I I I I I 1 1 I I I I I I 8 , 8 6 4 2 0 2 4 6 8 10 12 14 16 18 20 22 Kilometers From The Summit

Fig. 15. Schematic diagram of flow paths that feed flank thermal areas. Heated gas-rich fluids rise through a narrow central region of the cone and form a shallow steam-heated aquifer containing condensed gases within the upper kilometer. Secondary boiling of this water feeds fumaroles at summit. Lateral flow of gas-depleted fluids and secondary boiling feeds fumaroles on the upper flank. Lateral flow and mixing with cold shallow groundwater feeds springs on the lower flank (Paradise and Winthrop). Chloride-rich springs at Longmire resemble the unmixed end member for Paradise and may have a similar origin but deeper flow path which allows it to surface with less shallow mixing. Section stratigraphy from Fiske et al. (1963). analogous to a greased bowl. Furthermore, the bowl is guidance from Mark S. Ghiorso. I thank J.D. Friedman likely open to the north, judging from the northward and H.H. Kieffer for providing thermal-infrared distribution of the Osceola and the northerly position images, and T. Casadevall and I. Friedman for fuma- of the summit cone. role-gas sampling equipment and analytical support. Potential triggering mechanisms for avalanches Field efforts and discussion with numerous colleagues include earthquakes, eruptions, or steam explosions. was especially helpful, in particular Austin Post for Steam explosions are common in hydrothermal sys- introducing me to volcano/hydrologic interactive tems when transport paths are temporarily plugged by processes and Harry X. Glicken for shedding light on mineral deposition, or when overburden pressures are the resultant muck, blocky and otherwise. National suddenly reduced. In the case of Mount Rainier, any of Park Service personnel provided administrative assis- the active fumarole areas could produce steam explo- tance in working within Mount Rainier National Park. sions. Thermal areas on the upper flank are particularly Review comments by S.E. Ingebritsen, R.H. Mariner, susceptible to sudden reduction of lithostatic pressures D.A. Swanson, and B. Voight are greatly appreciated. as a result of normal erosive processes that allow small, Financial support during the dissertation work was piece-meal rock avalanches. A small avalanche fol- received from the Geothermal Program and the Water lowed by a steam explosion could conceivably yield a Resources Division of the U.S. Geological Survey. The much larger avalanche and produce deposits like those contents of this paper do not necessarily reflect the found in the Holocene record. The several localities of views and policies of the U.S. Environmental Protec- flank thermal activity make such events possible on tion Agency. Mention of trade names or commercial every side of Mount Rainier. products does not constitute endorsement or recom- mendation by the U.S. Environmental Protection Agency. Acknowledgements References This paper is based largely on dissertation work at Browne, P.R.L.. 1978. Hydrothermal alteration in active geothermal the University of Washington, where I received helpful fields. Annu. Rev. Earth Planet. Sci., 6: 229-250. D. Frank/Journal of Volcanology and Geothermal Research 65 (1995) 51-80 79

Budyko, M.I., 1956. Teplovoi BaIans Zemnoi Poverkhnosti [The sen volcanic region. National Technical Information Services, Heat Balance of the Earth’s Surface]. Hydrometeorological NTIS N78 23499/LL, 46 pp. Press, Leningrad, 255 pp. Translated by N.A. Stepanova, 1958. Friedman, J.D. and Frank, D., 1980. Infrared surveys, radiant flux U.S. Department of Commerce, Washington D.C., 259 pp. and total heat discharge at Mount Baker volcano, Washington, Burk, R.L., 1979.Factors affecting “O/160 ratios in cellulose. Ph.D. between 1970 and 1975. U.S. Geol. Surv., Prof. Pap., 1022-D. dissertation, University of Washington, Seattle, WA, 127 pp. 33 PP. Campbell, K.V.. Miers, J.H., Nichols, B.M., Oliphant, J., Pytlek, S., Friedman, J.D., Frank, D. and Kieffer, H.H., 1981. ThermaI infrared Race, R.W., Shaw, G.H. and Gresens, R.L., 1970. A survey of surveys of the May 18 crater. subsequent lava domes, and asso- thermal springs in Washington State. Northwest Sci., 44: I-1 1. ciated volcanic deposits. In: P.W. Lipman and D.R. Mullineaux Coombs, H.A., 1936. The geology of Mount Rainier National Park. (Editors), The 1980Eruptions of Mount St. Helens, Washington. Univ. Washington Publ. Geol., 3: 131-212. U.S. Geol. Surv., Prof. Pap., 1250: 279-293. Craig, H., 1961. Isotopic variations in meteoric waters. Science, 133: Friedman, J.D., Williams, D.L. and Frank, D.. 1982. Structural and 1702-1703. heat-flow implications of infrared anomalies at Mount Hood, Crandell, D.R., 1963a. Paradise debris flow at Mount Rainier, Wash- Oregon, 1972-1977. J. Geophys. Res., 87: 2793-2803. ington. U.S. Geol. Surv., Prof. Pap., 475-B: B 135-B 139. Haines, A.L.. 1962. Mountain Fever, Historic Conquests of Rainier. Crandell, D.R., 1963b. SurficiaI geology and geomorphology of the Oregon Historical Society, Portland, OR, 255 pp. Lake Tapps quadrangle, Washington. U.S. Geol. Surv., Prof. Pap. Helgeson. H.C., 1979. Mass transfer among minerals and hydro- 388-A. 84 pp. thermal solutions. In: H.L. Barnes (Editor), Geochemistry of Crandell, D.R., 1969. Surf&I geology of Mount Rainier National Hydrothermal Ore Deposits. John Wiley and Sons, New York, Park, Washington. U.S. Geol. Surv., Bull. 1288.41 pp. NY, pp. 568610. Crandell, D.R., 1971. PostglaciaJ lahars from Mount Rainier volcano, Hemley, J.J., Hostetler, P.B., Gude, A.J. and Mountjoy, W.T., 1969. Washington. U.S. Geol. Surv., Prof. Pap. 677.73 pp. Some stability relations of alunite. Econ. Geol., 64z 599-612. Crandell, D.R. and Fahnestock, R.K., 1965. Rockfalls and ava- lngebritsen, S.E. and Sorey, M.L., 1985. A quantitative analysis of lanches from Little Tahoma Peak on Mount Rainier, Washington. the Lassen hydrothermal system, north central California. Water U.S. Geol. Surv., Bull. 1221-A. 30 pp. Resour. Res., 21: 853-868. Danes, Z.F., 1965. A new steam vent on Mt. Rainier, Washington. Ingebritsen, S.E., Sherrod, D.R. and Mariner, R.H., 1992. Rates and J. Geophys. Res., 70: 2003. patterns of groundwater flow in the Cascade Range volcanic arc, Ellis, A.J. and Mahon, W.A.J., 1977. Chemistry and Geothermal and effect on subsurface temperatures. J. Geophys. Res., 97: Systems. Academic Press, New York, NY, 392 pp. 4599-4627. Evans, L., 1930. Preliminary studies of the NisquaIly, Paradise and Ingraham, E.S., 1895. The Pacific Forest Reserve and Mt. Rainier. Stevens Glaciers, Mount Tacoma-Rainier National Park. City The. Calvert Company, Seattle, WA, 48 pp. Light of Tacoma, Tacoma. Jones, J.B. and Segnit, E.R., 1971. The nature of opal, I. Nomencla- Fahnestock, R.K., 1978. Little Tahoma Peak rockfalls and ava- ture and constituent phases. Geol. Sot. Aust. J., 18: 57-67. lanches, Mount Rainier, Washington, U.S.A. In: B. Voight (Edi- Kieffer, H.H., Frank, D. andFriedman, J.D., 1981. Thermal infrared tor) , Rockslides and Avalanches, 1, Natural Phenomena, Chapter surveys at Mount St. Helens -observations prior to the eruption 5. Elsevier, Amsterdam-New York, pp. 181-196. of May 18. In: P.W. Lipman and D.R. Mullineaux (Editors), Fiske, R.S., Hopson, C.A. and Waters, AC., 1963.Geology of Mount The 1980Eruptionsof Mount St. Helens, Washington. U.S. Geol. Rainier National Park, Washington. U.S. Geol. Surv., Prof. Pap. SW., Prof. Pap., 1250: 257-277. 444 93 pp. Kieffer, H.H., Frank, D. and Friedman, J.D., 1982. Calibrated air- Flett, J.B., 1912. The thermal caves. The Mountaineer, 5: 58-60. borne thermal infrared survey of the Cascade Range volcanoes Frank, D., 1983. Origin, distribution and rapid removal of hydro- (abs.). In: Proc. International Association of Volcanology and thermally formed thy at Mount Baker, Washington. U.S. Geol. Chemistry of the Earth’s Interior (IAVCEI), Reykjavik, August Surv., Prof. Pap. 1022-E. 3 1 pp. 1982. Frank, D., 1985. Hydrothermal Processes at Mount Rainier, Wash- Kiver, E.P. and Mumma, M.D., 1971. Summit fim caves, Mount ington. Ph.D. dissertation, University of Washington, Seattle, Rainier, Washington. Science, 173: 320-322. WA, 195 pp. Kiver, E.P. and Steele, W.K., 1972. Summit fim cave research, Frank, D. and Friedman, J.D., 1974. Newly detected infrared anom- August 1971, Mount Rainier Nat. Pk., Wash. (abs.). Geol. Sot. alies at Mount Rainier, Washington (abs.). In: Symposium on Am., Abstr. Programs, 4: 385. Applications of Satellite and Airplane Remote Sensing of Natural Korosec, M.A., 1980. IV. Thermal and mineral spring investigations Resources in the Pacific Northwest, Tumwater, Washington, May (surveys and analyses, 1978-1979). In: M.A. Korosec and J.E. 2, 1974. Washington Department of Natural Resources, p. 22. Schuster (Editors), The 1979-1980 Geothermal Resource Friedman, J.D. and Frank, D., 1977. Heat discharge from Mount St. Assessment Program in Washington. Wash. Dep. Natural Helens, Washington. U.S. Geol. Surv., Open-File Rep. 77-541, Resour. Open File Rep., 81-3: 42-92. 29 PP. Lange, I.M. and Avent, J.C.. 1973. Ground-based thermal infrared Friedman, J.D. and Frank, D., 1978. ThermaI surveillance of active surveys as an aid in predicting volcanic eruptions in the Cascade volcanoes using Landsat-l Data Collection System: Part 4, Las- Range. Science, 182: 279-281. 80 D. Frtrnk / Journal of Volumology and Grothermul Research 65 (I 995) 51-80

Lange, I.M. and Avent, J.C., 1975. Ground-based thermal infrared Survey Research 1969. U.S. Geol. Surv., Prof. Pap.. 650-B: B IS- surveys of Mount Rainier volcano, Washington. Bull. Volcanol., 818. 38: 929-943. Murphy, M.T., 1991. A model for the evolution of the Tatoosh Lokey, W.M., 1973. Crater studies at a sleeping volcano. Explorers magmatic complex, southern Washington Cascades (abs.). EOS. J., 51: 167-170. Trans., Am. Geophys. Union, 72: 602. Lokey, W.M., Mack, R., Miller, M-M., Prather, B.W. and Kiver, Nehring, N.L., Mariner, R.H., White, L.D., Heubner, M.A., Roberts. E.P., 1972. Project Crater: Mt. Rainier glacio-volcanological E.D., Harmon, K., Bowen, P.A. and Tanner, L., 1979. Sulfate research, 1970-72 (abs.) Arctic and Mountain Environments geothermometry of thermal waters in the western United States. Symposium, Michigan State University, April 22-28, 1972. U.S. Geol. Surv., Open-File Rep. 79-l 135. I 1 pp. Majors, H.M., 1962. The springs of Lassen Volcanic National Park, Norris, R.D.. 1994. Seismicity of rockfalls and avalanches at three Mount Rainier National Park, and Olympic National Park. Seat- Cascade Range volcanoes: implications for seismic detection of tle, March, 1962. 88 pp (unpubl.). hazardous mass movements. Bull. Seismol. Sot. Am., 84: l925- Majors, H.M. and McCollum, R.C. (Editors), 198la. First winter 1939. expedition to Mount Rainier, 1894. Northwest Discovery, 2: Piper, A.M., 1944. A graphic procedure in the geochemical interpre- 406-465. tation of water analyses. Trans. Am. Geophys. Union, 25: 914 Majors, H.M.and McCollum, R.C. (Editors), 1981 b. Mount Rainier. 923. the tephra eruption of 1894. Northwest Discovery, 2: 334-38 1. Russell, l.c., 1898. Glaciers ofMount Rainier. U.S. Geol. Surv.. 18th Mariner, R.H., Presser, TX and Evans, W.C., 1982. Chemical and Annu. Rep., 189697, pt. 2, pp. 349415. isotopic composition of water from thermal and mineral springs Scott, K.M., Pringle, P.T. and Valiance. J.W., 1992. Sedimentology. of Washington. U.S. Geol. Surv., Open-File Rep. 82-0098, 20 behaviour, and hazards of debris flows at Mount Rainier. Wash- PP. ington. U.S. Geol. Surv., Open-File Rep. 90-385, 106 pp. Mariner, R.H., Presser, T.S., Evans, W.C. and Pringle, M.K.W., Sharp, R.P., Epstein, S. and Vidiziunas, I., 1960. Oxygen isotope 1990. Discharge rates of fluid and heat by thermal springs of the ratios in the Blue Glacier, Olympic Mountains. Washington. Cascade Range, Washington, Oregon. and Northern California. U.S.A. J. Geophys. Res., 65: 40434059. J. Geophys,Res.,95: 19,517-19,531. Stearns, N.D.. Stearns, H.T. and Waring, G.A., 1937. Thermal Mattinson. J.M., 1977. Emplacement history ofthe Tatoosh volcanic- springs in the United States. U.S. Geol. Surv, Water-Supply Pap. plutonic complex, Washington: ages of zircons. Geol. Sot. Am. 679-B. 206 pp. Bull., 88: 1509-1514. Stevens, H., 1876. The ascent of Mount Tahoma. Atlantic Monthly. Miller, M.M.. 1970. Wind, sky, and ice. Explorers J., 48: 192-197. Nov., pp. 51 l-530. Molenaar, D., 1979. The Challenge of Rainier. The Mountaineers, Swanson, D.A., Malone, SD. and Samora. B.A., 1992. Mount Rai- Seattle, WA, 364 pp. nier: a decade volcano. EOS, Trans. Am. Geophys. Union, 73. Moser, H. and Stichler, W.. 1970. Deuterium measurements on snow 177-186. samples from the Alps. In: Isotope Hydrology. IAEA (Intema- Weissberg, B.G., Browne, P.R.L. and Seward, T.M., 1979. Ore met- tional Atomic Energy Agency), Vienna, pp. 43-57. als in active geothermal systems. In: H.L. Barnes (Editor), Geo- Moxham, R.M., 1970. Thermal features at volcanoes in the Cascade chemistry of Hydrothermal Ore Deposits. John Wiley and Sons, Range, as observed by aerial infrared surveys. Bull. Volcanol.. New York, NY, pp. 738-780. 34: 77-106. White, D.E., 1957. Thermal waters of volcanic origin. Geol. Sot. Moxham, R.M., 1972. Thermal surveillance of volcanoes. In: The Am. Bull., 68: 1637-1657. Surveillance and Prediction of Volcanic Activity. UNESCO. White, D.E. and Waring, G.A., 1963. Volcanic emanations. In: M. Paris, pp. 103-124. Fleischer (Editor), Dataof Geochemistry, Chapter K. U.S. Geol. Moxham, R.M., Crandell, D.R. and Marlatt, W.E.. 1965. Thermal Surv., Prof. Pap. 440-K. 29 pp. features at Mount Rainier, Washington, as revealed by infrared Wolery, T.J., 1979. Calculation of chemical equilibrium between surveys. U.S. Geol. Surv., Prof. Pap., 525-D: D93-DlOO. aqueous solution and minerals: the EQ3/6 software package. Moxham, R.M., Boynton, G.R. and Cote; CE., 1972. Satellitetelem- Lawrence Livermore Laboratory report UCRL52658.4 1 pp. etry of fumarole temperatures, Mount Rainier, Washington. Bull. Yamada, E., 1976. Geological development of the Onikobe caldera Volcanol.. 36: 191-199. and its hydrothermal system. In: Proceedings, Second Sympo- Mullineaux, D.R., 1974. Pumice and other pyr@stic deposits in sium on the Development and Use of Geothermal Resources, San Mount Rainier National Park. Washington. U.S. Geol. Surv., Francisco, May 20-29, 1975. I : 665672. Bull. 1326, 83 pp. Zotov, A.V., 1967. Recent formation of alunite in the Kipyashcheye Mullineaux, D.R., Sigafoos, R.S. and Hendti&s, E.L., 1969. A his- (“Boiling” Crater Lake. Golovnin volcano, Kunashir Island. toric eruption of Mount Rain&r, Washington. In: Geological AN Armyanskoy SSR Dokl., 174: 124- 127