Introduction

Mount Rainier is an active volcano of the Cascade Range in Washington State, 50–70 km southeast of the major metropolitan areas of Seattle and Tacoma. With one exception, rivers heading on Mount Rainier flow to the west and northwest into the southern Puget Sound lowland where more than 150,000 people live on deposits from lahars and related floods released from Mount Rainier over the last 5,600 years (Crandell and Waldron 1956; Scott et al. 1995; Vallance and Scott 1997; Sisson et al. 2001). Mount Rainier produced few voluminous tephra-fall deposits in the Holocene, and glaciers cover much of the upper mountain, hindering detailed estimates of its recent eruptive history. This history is important because it has been unclear if major lahars formed mainly during times of eruptive activity (due to dislodgement of edifice flanks and by newly erupted hot materials transiting and melting glaciers), or if large lahars took place during non-eruptive times (by spontaneous gravitational collapse of unstable edifice flanks, or by triggering from tectonic or severe rainfall events (Scott 2004)). Accurate estimates of the number, timing, and style of Holocene eruptions can help to evaluate links between eruptions and lahars, and thereby guide monitoring and event-response planning.

To better determine Mount Rainier’s late-Holocene eruptive history, we undertook a detailed stratigraphic and geochemical study of thin apparent tephra-fall deposits on the volcano’s proximal edifice flanks, as well as of limited exposures of young lava and pyroclastic-flow deposits. Here we present evidence for ten–12 temporally and compositionally distinguishable eruptions over the last 2,600 years. Most eruptions probably involved multiple explosive ejections of ash, and some were accompanied by pyroclastic or (and) lava flows. This estimate supercedes previous published evidence for only three or four eruptions over that time span, and for only 11 or 12 eruptions for all of the Holocene (Mullineaux 1974; Hoblitt et al. 1998). The last 2,600 years were studied in detail because they encompass an interval of frequent eruptions and lahars from ∼2,600 to ∼2,200 cal year BP, referred to here as the Summerland eruptive period, that followed a period of apparent dormancy from ∼4,400 to ∼2,600 cal year BP, and because major lahars are also known at ∼1,500, ∼1,000, and ∼500 cal year BP that might have been caused by eruptions (Crandell 1971; Scott et al. 1995; Zehfuss et al. 2003) (cal year BP refers to radiocarbon ages calibrated to calendar years, 14C year BP refers to radiocarbon ages as measured, year BP refers to interpolated ages, all relative to AD 1950). In the following we first describe limited exposures of late-Holocene lava and pyroclastic flow deposits, beginning with the volcano’s summit, and then describe tephra-fall deposits chiefly from a particularly complete exposure near Summerland on the volcano’s east flank (Fig. 1). Tephra and flowage deposits are then correlated using glass and mineral compositions to arrive at the likely number and style of eruptions over the last 2,600 years.

Fig. 1
figure 1

Map of Mount Rainier volcano and vicinity showing discussed localities, extent of Quaternary volcanic rocks (green), flank vents (yellow stars), glaciers (white), and Tertiary basement (gray). Rectangle shows area of Fig. 2. Solid black lines show roads, some with Washington State Route numbers. TG is Tahoma Glacier, Puy. Clvr. is Puyallup Cleaver. Marginal ticks and internal crosses mark latitude and longitude

Late-Holocene lava and pyroclastic flow deposits

East summit crater lava flows

The present summit cone of Mount Rainier grew subsequent to the major Osceola collapse of 5,600 cal year BP (Crandell and Waldron 1956; Vallance and Scott 1997) that removed the volcano’s summit and east–northeast flank during a period of magmatic eruptions. The Osceola collapse excavated a horseshoe-shaped crater 1.5 km across, open to the east–northeast. The collapse amphitheater is now almost entirely filled by younger lava flows that constructed a new summit cone, but the amphitheater’s upper margin can be recognized as an arc of small peaks flanking the true summit on the north, northwest, and southwest (Figs. 1 and 2). The former outlet of the collapse amphitheater is now replaced by a smooth dip-slope of ice-buried young lava flows that spans more than 1,500 m (4,800 ft) elevation from the summit to Camp Schurman and Steamboat Prow on the volcano’s east–northeast flank. The smooth constructional slope of the young lava cone contrasts with the steep, irregular incised headwalls elsewhere on the upper volcano. The young cone forms the volcano’s summit and is capped by two overlapping craters, each 0.4 km across, with the eastern crater younger and truncating the western crater. The east summit crater consists of one or two conformable, radially outward-dipping lava flows of compositionally uniform porphyritic andesite (plagioclase–hypersthene–augite–Fe-Ti oxides) (Table 1), overlain at the true summit by about 5 m of poorly bedded unconsolidated (explosion) breccia. The east summit crater rim exposes the youngest lavas erupted from the volcano, and paleomagnetic measurements of these are consistent with an eruption age of 2,000–2,200 year BP (Vallance et al., in prep). The flows are beheaded and exposed in cross section along the inner crater wall, indicating that explosions enlarged the crater subsequent to the youngest lava effusion. Rocks forming the western half of the east summit crater, as well as those of the earlier west summit crater, are pervasively replaced by clays and other secondary minerals due to acid-steam alteration (Frank 1995; Zimbelman 1996; Crowley and Zimbelman 1997; Finn et al. 2001; John et al. 2008), though blocks of intensely altered lava can still be distinguished.

Fig. 2
figure 2

Detailed map of Mount Rainier’s summit and northeast slope showing upper perimeter of Osceola collapse amphitheater (hachured line), approximate area of young summit cone (dashed line and shaded), Emmons–Winthrop high-Sr lava flows (orange), east summit crater lava flows (red and dotted line where concealed), west summit crater (mauve), area of summit hydrothermal alteration (cross pattern), Pleistocene lava flows (green), Tertiary basement (gray), and glacial deposits (yellow). Paleomagnetic measurement sites (pm) from Vallance et al., in prep. TG is Tahoma Glacier. Contour interval 500 ft (152 m), index contours every 2,500 ft (762 m). Marginal ticks and internal crosses mark latitude and longitude

Table 1 Representative whole-rock chemical compositions of late-Holocene Mount Rainier lava flows and block-and-ash bombs

Emmons–Winthrop lava flows

Isolated windows through the large Emmons and Winthrop Glaciers expose lava flows that form the surface of the young summit cone in the former breach of the Osceola collapse amphitheater (Figs. 1 and 2). These rock exposures are directly down-slope of the east summit crater, consistent with a late-cone-building age for the exposed lava flows. Outcrops at 2,900–3,000 m (9,500–9,800 ft) elevation consist of minimally incised, variably glassy, polygonal-jointed ice-chilled andesite, and a steeply flow-banded lava levee of the same material. Lava samples are compositionally uniform porphyritic andesite (plagioclase–hypersthene–augite–Fe-Ti oxides, one rounded amphibole found in thin sections from five localities), similar in major element concentrations (Table 1) to andesite samples from the east summit crater. They are distinguished from the summit lavas, however, by their much higher Sr concentrations (∼800 vs. 550 ppm), and to a lesser degree by their higher concentrations of Th (6.8 vs. 5.3 ppm), U (2.1 vs. 1.6 ppm), and light rare earth elements (57 vs. 45 ppm Ce), and lower Cr concentrations (18 vs. 30 ppm). Such high Sr concentrations are uncommon for Mount Rainier eruptive products (McKenna 1994; Stockstill et al. 2002; Sisson, unpublished), facilitating correlations with Holocene tephras.

The broad geographic extent of the high-Sr andesites beneath the Emmons and Winthrop Glaciers is evidence that the lavas are multiple flows or flow lobes from a single, widespread effusive episode that predated eruption of the east summit crater andesite lavas. Paleomagnetic orientations from four Emmons–Winthrop high-Sr lava flow localities are indistinguishable, differ from those from the summit lava flows, and match that of a late-Holocene pyroclastic flow deposit in the valley of the South Puyallup River on the volcano’s lower southwest flank (Vallance et al., in prep; Hagstrum and Champion 2002). These geologic, compositional, and paleomagnetic results are consistent with the Emmons–Winthrop high-Sr andesites having erupted late in the growth of the post-Osceola summit cone, but before the eruption of the normal-Sr andesites of the east summit crater. Contacts between the high-Sr and the younger summit crater lava flows are concealed by thick glacial ice.

South Puyallup block-and-ash flow

A non-welded, sandy block-and-ash flow deposit containing prominent dark brown-to-black breadcrust bombs, 0.2–1.25 m diameter, is exposed where the West Side Road of Mount Rainier National Park crosses the valley of the South Puyallup River (Crandell 1971) (Fig. 1). Uniform paleomagnetic orientations of undisturbed breadcrust bombs indicate that the flow was emplaced hot and cooled in the same magnetic field direction as the high-Sr lava flows exposed through the Emmons and Winthrop Glaciers (Crandell 1971; Hagstrum and Champion 2002; Vallance et al., in prep.). Radiocarbon dating of a carbonized log (Crandell 1971; Table 2) indicates eruption of the pyroclastic flow at 2,580 ± 150 cal year BP (calendar ages with uncertainties report the midpoint and range of the most probable calibration-age window from OxCal 4.0 (Bronk Ramsey 1995, 2001); thus, 2,580 ± 150 represents a calibration window of 2,730–2,430 cal year BP). Roadcut incisions show that the pyroclastic deposit is no more than 15 m thick and forms a small (∼0.3 km2) valley-flooring terrace near 1,130 m (3,700 ft) elevation. Erosion has removed the deposit down-valley, but glassy breadcrust bombs are scattered discontinuously up the valley of the South Puyallup River to at least has high as 1,300 m (4,300 ft) elevation, and isolated breadcrust bombs are perched atop narrow Puyallup Cleaver as high as 2,800 m (9,200 ft) elevation; rare breadcrust bombs are also in the upper valley of Tahoma Creek to the south of the South Puyallup locality. This distribution indicates that the pyroclastic flow descended southwest from the summit cone through a low notch in the rear wall of the Osceola collapse amphitheater at the head of the Tahoma Glacier (Figs. 1 and 2), and then traveled into the valley of the South Puyallup River system, with a small amount also spilling south into Tahoma Creek.

Table 2 Calibrated 14C ages establishing timing of Mount Rainier’s Summerland period and younger eruptions

The South Puyallup breadcrust bombs are compositionally uniform (Table 1) porphryitic andesites (plagioclase–hypersthene–augite–Fe-Ti oxides, one resorbed amphibole found in thin sections of 11 bombs) that are slightly less evolved in major element abundances (lower SiO2, higher FeO*, MgO, CaO concentrations) than andesites from the east summit crater or the Emmons–Winthrop Glaciers. The bombs’ trace element abundances are unremarkable for Mount Rainier andesites and are similar to those of samples of the east summit crater (Table 1). The bombs are thus distinct from the high-Sr lava flows exposed through the Emmons and Winthrop Glaciers, despite sharing a common paleomagnetic orientation. This paleomagnetic direction is atypical for the late Holocene (Hagstrum and Champion 2002), which is evidence that the block-and-ash flow and Emmons–Winthrop lava flows resulted from separate eruptions of compositionally distinct magmas over a brief time period (<100 years, D. Champion 2005, personal communication). Relative ages of the Emmons–Winthrop and South Puyallup eruptives are interpreted in the section: Correlation of flow and fall deposits based on glass and mineral compositions.

Crandell (1971) notes an abundance of black andesite bombs in the ∼500 cal year BP Electron Mudflow in the vicinity of the towns of Orting and McMillan, 40–50 km NW of Mount Rainier’s summit. Bombs collected from that area (SMN sample in Table 1) are identical in chemical composition and appearance to breadcrust bombs collected directly from the South Puyallup pyroclastic flow deposit, indicating that those bombs in the Electron Mudflow were entrained from South Puyallup-related deposits and are not a juvenile magmatic component of the Electron Mudflow event.

Late-Holocene Mount Rainier tephra deposits

Mullineaux (1974) identified, dated, and named 11 tephra-fall deposits interpreted as products of explosive Holocene eruptions of Mount Rainier (Fig. 3). These Mount Rainier tephras consist chiefly of juvenile pumiceous-to-scoriaceous lapilli, ash, and bombs, generally deposited as lobes extending east from the volcano, with the exception of two units that are a local directed blast deposit of dense lithic blocks (layer S) and a mixed fall deposit of clay and pumice (layer F), both associated with the Osceola collapse event of 5,600 cal year BP (Vallance and Scott 1997). To these 11, Hoblitt et al. (1998) added a 1,080 ± 250 14C year BP (1,040 ± 410 cal year BP) fine-grained tephra that is restricted to the upper valley of the White River (Fig. 1). Five of the named tephras, in order layers H (∼4,700 year BP), B (∼4,500 year BP), C (∼2,200 cal year BP), White River (∼1,000 cal year BP) and X (AD 1820–1854), were deposited subsequent to the Osceola collapse event and record some of the eruptions that rebuilt Mount Rainier’s summit cone (excepting layer X, for which we present evidence following that this is not a true eruption deposit).

Fig. 3
figure 3

Summary stratigraphic section of thick, prominent Holocene tephra-fall deposits in subalpine meadows near Mount Rainier, modified from Mullineaux (1974). Yellow tinted intervals are deposits of ash-sized tephra erupted from Mount St. Helens (MSH). Orange tinted interval is ash-sized tephra from the paroxysmal eruption of Mount Mazama, Oregon, that created Crater Lake. Pink tinted intervals are deposits of pumiceous-to-scoriaceous ash, lapilli, and bombs from Mount Rainier eruptions. Red tinted interval is pumice-bearing clay-rich ashfall deposit from the Osceola collapse event; not shown from that event is the layer S directed blast block layer. Gray intervals contain both thin, dark, poorly vesicular tephras from Mount Rainier (not distinguished at this scale), and non-eruptive accumulations of ash-sized sediments reworked from earlier tephra deposits or carried by water and wind from nearby till

Mullineaux (1974) noted that additional eruptions might be recorded by numerous thin (mm–cm), dark strata consisting of sparsely vesicular ash grains and crystal fragments that are interlayered with the prominent named tephras, but that the thin dark ashes could not, at that time, be distinguished from local accumulations of ash redeposited by wind and water. Much effort in the present study has been directed at developing a stratigraphic sequence for these dark tephra layers, investigating if the layers are products of eruptions, and correlating some with the previously described young lava and pyroclastic flow deposits. A consistent sequence of dark, fine-grained tephra layers or groups of layers is preserved in sub-alpine meadows, generally 6–12 km from the volcano’s summit, as would result if most of those ash deposits were products of eruptions.

To further investigate if the dark, fine-grained tephras result from eruptions, multiple glassy ash grains were hand picked from samples of the various ash deposits, and their glass compositions were measured by electron-microprobe (see Electronic Supplementary Materials for analytical methods). Most glass-rich grains are blocky and sparsely vesicular, consistent with quenching and explosive fragmentation of largely degassed magma traversing the edifice hydrothermal system, but grains range with increasing vesicularity to scoriaceous with fluidal shapes, or less commonly, pumiceous (Fig. 4), indicative of some syn-eruptive vesiculation. Visibly glassy grains constitute a minority of ash (<30%) in each of the fine-grained dark tephra deposits, with the majority of grains being well-crystallized angular volcanic rock fragments and broken phenocrysts. Phenocrysts and microphenocrysts of plagioclase, pyroxenes, and Fe-Ti oxides are ubiquitous in the glassy grains. Microlites of these mineral phases are also present in varying amounts in the blocky-to-scoriaceous grains, probably due to syn-ascent crystallization driven by degassing, but are less abundant in pumiceous grains. Variable growth of microlites contributed to differences in glass composition between grains in single deposits. Nevertheless, we show that in most cases, field-based stratigraphic boundaries between fine-grained tephra units correspond with shifts in the most common glass composition, in the range of compositions, or in other aspects of composition or mineralogy, as would result if most of the glassy grains were juvenile magmatic ejecta, and the different dark, fine-grained tephra units resulted from distinct eruptions of magma, rather than by reworking of earlier sediments.

Fig. 4
figure 4

Backscattered electron micrographs of representative Summerland eruptive period glassy ash grains. Brightness corresponds to average atomic number, decreasing in the order Fe-Ti oxides (ox), pyroxenes (pxn), plagioclase (pl), glass (gl), and voids. Left grain—blocky, middle grain—scoriaceous-fluidal, right grain—pumiceous

The Summerland eruptive period

Tephra-fall products from the Summerland eruptive period are defined as a succession of eight stratified deposits. These consist of six thin, fine-grained ash units (SL1–SL6) that commence between the upper two of four cream-colored Mount St. Helens P tephras that are widespread at Mount Rainier. Summerland products also include the pumiceous-to-scoriaceous C tephra (= SL7) that immediately overlies the SL6 deposit, and may include an inconspicuous fine-grained ash (SL8) that overlies the C tephra. The SL1 to SL6, C, and several subsequent tephras are well developed in the upper half of a ∼2.5 m thick stream bank exposure near the Summerland campsite on Mount Rainier’s east flank (NAD27 latitude 46.8646° longitude −121.6611°), at a distance of 7.5 km from the vent. This exposure (Figs. 5, 6 and 7) was sampled both with a spatula at spaced intervals and as a continuous series of box cores that were impregnated with epoxy and polished to reveal fine stratigraphic details. In addition, isolated pumice lapilli were discovered by excavating large amounts of the SL1 deposit at Summerland, from the SL2 deposit near Paradise on the volcano’s south flank, and from the SL4 deposit in an exposure near Fan Lake on the volcano’s southeast flank (Fig. 1). The following descriptions mainly from the Summerland locality characterize tephra products of the Summerland eruptive period and sequentially younger tephras, and with glass compositions and radiometric ages, are the primary evidence for the number of Mount Rainier eruptions over the last ∼2,600 years. Representative tephra glass compositions are presented in Table 3 (see Electron Supplementary Materials for all analyses). Lahars spread from Mount Rainier during the Summerland eruptions, including the collapse-generated Round Pass Mudflow, numerous unnamed non-cohesive lahars of the Summerland lahar assemblage, and probably the pumice-rich National Lahar that may have been triggered by the C-tephra eruption (Crandell 1971; Scott et al. 1995; Zehfuss et al. 2003). Key tephra deposit features, correlative eruptive products, and associated lahars are summarized in Table 4.

Fig. 5
figure 5

Annotated photograph of Summerland eruptive period tephra section exposed near Summerland campsite, east–northeast flank of Mount Rainier. Figure 7 shows upper portion of exposure including SL8 deposit

Fig. 6
figure 6

Line drawing from photograph of tephras and non-eruptive sediments exposed near Summerland campsite, spanning from deposition of Mount St. Helens (MSH) Y to Mount Rainier C. Subdivisions of MSH-P tephras (MSH-Pu, etc.) are inferred from thickness and stratigraphic position (Mullineaux 1996) but have not been verified geochemically. Illustrated upper portion of MSH-Y is reworked

Fig. 7
figure 7

Annotated photograph of post-C tephra ash deposits exposed near Summerland campsite, east–northeast flank of Mount Rainier. MSH-W and MSH-X are thin but widespread Mount St. Helens tephras. CIgW1, 2, 3, and fine ash between the Mount Rainier C and SL8 tephras are probably non-eruptive deposits of fine-grained reworked ash

Table 3 Representative electron-microprobe major-oxide analyses of glasses from late-Holocene Mount Rainier tephras
Table 4 Summerland eruptive period and subsequent tephra deposits at Mount Rainier

Pre-summerland deposits

The Summerland eruptive period was preceded by an interval of apparent dormancy at Mount Rainier (Fig. 3) starting at about the onset of the Mount St. Helens Y tephra eruptions (≤4,400 cal year BP) and ending late in the time of eruption of the Mount St. Helens P tephras (∼2,700 cal year BP) (Mullineaux 1974, 1996; Crandell et al. 1981; M. Clynne 2007, unpublished ages). This dormant interval was preceded at Mount Rainier by the Osceola collapse and an early phase of summit regrowth, recorded at the Summerland locality by the collapse-associated clay-rich F tephra, the shortly following scoriaceous-to-pumiceous H and B tephras, and numerous unnamed thin, dark, fine-grained tephras (combined thickness of 30–35 cm). These juvenile Mount Rainier tephras are capped by strata of yellowish pumiceous Mount St. Helens Y ash (MSH-Y, 4–6 cm), mainly of the voluminous Yn event of ∼3,700 cal year BP (Crandell et al. 1981; Mullineaux 1996). Sediment from the dormant interval begins above the MSH-Y deposits as 9–10 cm of brownish silty ash with common horizontally oriented plant needles, probably from heather, and scattered grains of reworked Y ash. The lower pair (3–4 cm) of four cream-colored Mount St. Helens P pumiceous ashes interrupts the brown, non-eruptive sediments. These lower two light-colored ash deposits are probably units MSH-Pm and MSH-Ps (Mullineaux 1996). This lower P doublet is overlain by an additional 14–16 cm of brownish, plant needle-rich, silty sediment that includes at least two lenses of reworked Mount Rainier D pumice lapilli in its upper half at the Summerland locality. A radiocarbon sample from beneath these pumice lenses gives an age of 2,870 ± 90 cal year BP. Such pumice lenses have not been found at this exposure level elsewhere around Mount Rainier, indicating they are a local depositional feature. The organic-rich, non-eruptive sediment and reworked pumice lenses are capped by the upper doublet of Mount St. Helens P ash deposits, probably units MSH-Pu and MSH-Py (Mullineaux 1996), which are separated at the Summerland locality by unit SL1, the first Summerland eruptive period tephra (combined MSH-Pu, SL1, and MSH-Py thickness of 5–7 cm).

SL1

The SL1 tephras are the earliest recognized fall deposit of the Summerland eruptive period and lie between the upper two of four widespread Mount St. Helens P ash deposits (Figs. 5 and 6). At the Summerland locality unit SL1 consists of 1–4 cm of fine-to-medium-grained, poorly sorted, brownish-gray ash bounded above and below by the white-to-cream colored, amphibole-bearing, pumiceous MSH-P ashes. Contacts with the MSH-P tephras are generally sharp, but with minor interfingering, and up to 1 cm of SL1-like ash locally overlies the upper P deposit, possibly due to minor post-depositional reworking. The SL1 deposit is considerably thicker (8–9 cm) near Glacier Basin, on the volcano’s northeast flank (Fig. 1), where it can be divided into three subunits that probably record multiple explosive events, though these have not been studied in detail.

The SL1 tephra contains common dark, microlite-bearing glassy grains, 0.2–1.5 mm, ranging from poorly vesicular and blocky to scoriaceous. Angular gray lithics to 1.5 mm predominate. Reworked white pumiceous MSH-P ash grains, some with fresh brown amphibole phenocrysts, are scattered in the SL1 deposit, indicating minor post-depositional reworking. Isolated lapilli of brownish, microlite-poor pumice (plagioclase–hypersthene–augite–Fe-Ti oxides) and similar brownish pumiceous ash grains are also present in the SL1 deposit but are uncommon (tephra-fall deposits originating from Mount St. Helens consist exclusively of ash-sized grains at the distance of Mount Rainier (Mullineaux 1974), so these and other pumiceous and scoriaceous lapilli are Mount Rainier eruptive products). Glasses in dense-to-scoriaceous SL1 grains (Table 3) are mainly rhyolitic and differ in composition between grains (70–78 wt.% SiO2), whereas the pumice-lapilli glasses are dacitic (66.8 wt.% SiO2) and uniform from grain to grain (Table 3, Fig. 8). A radiocarbon sample (twig) collected from the SL1 unit yields 2,450 ± 100 cal year BP (Table 2).

Fig. 8
figure 8

Stratigraphic succession of glass SiO2 concentrations for late-Holocene Mount Rainier deposits, subdivided as ash grains (small blue diamonds), pumice lapilli (large circles), and rinds on South Puyallup breadcrust bombs (yellow squares). Glasses from the subplinian Mount Rainier C tephra are subdivided into matrix glass from dominant brown pumice (tan), matrix glass in poorly inflated dacite lapilli (gray), and white blebs and streaks of dacitic pumice (white) dispersed within andesitic pyroclasts. Thin dotted lines show periods of repose marked by non-eruptive accumulations of silty, organic-rich reworked ash. Small white circles show glass compositions of ash grains from Mount St. Helens, probably reworked from P tephra units

SL2

The SL2 tephras form a relatively thick (6–8 cm at the Summerland exposure) fine-to-medium-grained ash deposit overlying the uppermost Mount St. Helens P ash. The SL2 deposit is distinctive in that it darkens upward gradually from a brownish base to uniform, dark brownish-gray ash in its upper half (Figs. 5 and 6). The lowest part of the SL2 unit (to 1 cm at Summerland) is brown, fine-grained silty ash with abundant horizontally oriented plant needles, scattered angular lithics (to 2.5 mm), and few dark glassy grains. This brownish interval probably records a hiatus between the SL1 and SL2 eruptions, with the brownish silty ash being windblown sediment trapped by plant communities that dropped the abundant needles. Above this basal zone the SL2 tephras are increasingly dominated by blocky-to-scoriaceous glassy grains (commonly 0.1–1 mm, rarely to 3.5 mm), accompanied by more abundant angular gray lithics (0.1–1.5 mm) and phenocryst fragments. The SL2 tephras are also faintly internally stratified on a scale of 0.25–1.5 cm. Stratification in the unit’s lower half is mainly defined by alternations between apparently non-eruptive sediments rich in brown, silty ash with abundant plant needles, similar to the basal layer, and coarser zones rich in lithics, crystal fragments, and glassy grains. Zones of brownish silty ash are absent in the upper half of the SL2 deposit where stratification is defined by slight variations in overall grain size. This upward increase in juvenile eruptive components typifies the SL2 unit and probably results from shortened repose periods, and perhaps eruption of greater amounts of magma, for successive SL2 explosive events. Isolated brown pumice lapilli (to 1 cm) are present but rare in the upper half of the SL2 tephra deposit in exposures near Paradise on the south flank of the volcano (Fig. 1), but have not been found at the Summerland locality. These pumice lapilli differ from those in the underlying SL1 unit in that they contain phenocrysts of fresh, non-resorbed brown amphibole in addition to plagioclase, hypersthene, augite, and Fe-Ti oxides. The weighted mean age of three radiocarbon samples (twigs) from the SL2 deposit is 2,610 ± 90 cal year BP (Table 2; for averaging, calibrated ages are weighted by (±value)−2 and reported weighted uncertainties are the larger of the weighted sample variance or (Σ weights)−0.5). The dark SL2 tephra set is overlain abruptly by another interval of non-eruptive silty brown sediment containing scattered, horizontally oriented twigs and plant needles that marks the base of the SL3 deposit.

Because of its thickness and vertical zonation, the SL2 tephra was sampled at five evenly spaced levels (Table 3). Glasses in the common, blocky-to-scoriaceous microlite-bearing ash grains have variable but mainly rhyolitic compositions that do not change systematically with height in the deposit (Fig. 8). Their compositions are similar to those of glasses in blocky-to-scoriaceous grains from the underlying SL1 deposit, but glasses in the SL2 amphibole-bearing pumice lapilli are dacitic and are distinctly less evolved (64.2 wt.% SiO2) than glasses in the amphibole-free pumice lapilli of the underlying SL1 tephra (66.8 wt.% SiO2). Glass compositions in the SL2 hornblende pumice lapilli are also uniform from grain to grain.

SL3–SL4

The SL3 and SL4 tephras are sequential deposits that can be distinguished by superposition but that probably represent the initial and main stages of a single eruptive episode. A non-eruptive zone of silty fine-grained brown ash with abundant horizontally oriented plant needles and twigs (1–1.5 cm thick at the Summerland locality) caps the underlying SL2 tephra and marks the base of the combined SL3–SL4 tephra set. The SL3 unit is the lowermost and relatively continuous stratum (1–1.5 cm thick) of fine-to-medium-grained brownish-gray ash immediately above the twig-bearing zone, whereas the SL4 tephras consist of a thicker overlying interval (4–6 cm) of similar but discontinuously stratified and burrowed ash. The absence of a well-developed silty organic-rich interval between the SL3 and SL4 deposits is evidence for little or no hiatus between deposition of those tephras. The SL4 tephra consists of multiple strata (five or six at Summerland, individually 0.75–2 cm thick) of fine-to-medium-grained brownish gray ash with scattered outsized lithics (to 4 mm) interlayered with laterally discontinuous zones up to 1 cm thick of finer-grained and better-sorted ash, commonly with faint internal laminations. The internal complexity of the SL4 unit is suggestive of slight reworking, perhaps due to falling on snow, followed by melting into place. The SL3 and SL4 tephras are distinctive in that their glassy ash population contains common brown pumiceous grains (0.5–3 mm) that are particularly abundant in a 1–2 cm thick zone two-thirds height above the base of the combined units. Isolated brown pumice lapilli (plagioclase–hypersthene–augite–Fe-Ti oxides) are present in the SL4 deposit near Fan Lake on Mount Rainier’s southeast flank (Fig. 1), but only ash-sized brown pumice have been found at the Summerland exposure. The SL3–SL4 unit is bounded above by a thin (3–4 mm), fine-grained, well-sorted, light grayish brown ash with faint internal laminations and minor, outsized lithics and scoriaceous glassy grains that marks the base of the SL5 tephra set.

Glasses from the SL3–SL4 tephra set are mainly dacitic (66–70 wt.% SiO2) and are distinctly less evolved than glasses from the SL1 and SL2 deposits (Fig. 8; Table 3). Rhyolitic glasses are nearly absent. There is a possible temporal trend of slightly increasing melt evolution in SL3–SL4 ashes, with values of 66–69 wt.% SiO2 in the SL3 deposit, rising to 68–71 wt.% SiO2 in the upper SL4 deposit. Glasses from SL4 pumice lapilli have uniform dacitic compositions (68 wt.% SiO2), similar to those of glasses in many of the ash-sized grains; the pumiceous lapilli and ash appear to be the same magmatic component differing only in grain size. Notably, glasses forming the dense rinds of breadcrust bombs from the South Puyallup block-and-ash-flow deposit (68.4–68.7 wt.% SiO2) are a close compositional match to SL4 ash and pumice-lapilli glasses (Table 5, Figs. 8 and 9), which is evidence that the South Puyallup block-and-ash flow was emplaced during the SL4 eruptions. A radiocarbon sample (twig) from the middle of the SL4 unit gives an age of 2,540 ± 180 cal year BP, and the South Puyallup pyroclastic flow deposit has an age of 2,580 ± 150 cal year BP (outer portion of carbonized log); combining these yields a weighted mean age of 2,560 ± 120 cal year BP for the SL3–SL4–South Puyallup episode (Table 2).

Fig. 9
figure 9

Plot of glass SiO2 versus Al2O3 concentrations showing similarity of glasses between amphibole-bearing SL2 and C pumices, and between SL4 pumice and South Puyallup bomb-rinds glasses. Symbols as in Fig. 8

Table 5 Electron-microprobe major-oxide analyses of glasses from South Puyallup breadcrust bomb rinds

SL5

The SL5 tephras consist of faintly but regularly stratified fine-to-medium-grained brown and brownish gray ashes (5–9 cm aggregate thickness at Summerland) that overlie the thin, light-colored ash stratum capping the SL3–SL4 tephra deposits. Internal stratification (6–11 strata, individually 0.2–2 cm thick at Summerland) is defined chiefly by slight variations in grain size and color between fine-grained, well-sorted brown ash, and fine-to-medium grained, moderately sorted grayish brown ash, both with scattered outsized angular lithics (commonly to 2 mm, rarely to 6.5 mm). Many of the finer-grained brown strata also contain scattered horizontally oriented plant needles. At the Summerland locality two of the finer-grained strata are closely spaced and appear as a prominent light-brownish gray doublet near the top of the SL5 deposit that aids its field recognition. As with other Summerland period tephras, lithic grains predominate in the SL5 unit, followed in abundance by phenocryst fragments, but the glassy ash in the SL5 tephra is distinctive in consisting mainly of relatively coarse (to 2 mm) dark brown-to-black scoriaceous grains that stand out against the overall brownish-gray deposit.

Glassy grains were sampled at four levels spaced through the SL5 deposit. The SL5 glasses differ from the underlying SL3–SL4 glasses in that rhyolitic compositions reappear in abundance, similar to SL1–SL2 compositions, accompanied by dacitic glasses similar to those of the SL3–SL4 units (Table 3, Fig. 8). It is unclear if those SL5 dacitic glasses were reworked from earlier tephras, but the abrupt reappearance of rhyolitic glasses, coupled with the distinctive scoriaceous habit of many grains, are evidence that the SL5 tephras were produced by an eruptive episode following that of the SL3–SL4 deposits. This inference is supported by the compositions of plagioclase phenocrysts, as discussed in the section: Correlations of flow and fall deposits based on glass and mineral compositions. A single radiocarbon sample (twig) from the SL5 deposit has an age of 2,240 ± 250 cal year BP (Table 2).

SL6

The SL6 tephras form a fine-to-medium grained ash deposit (5–9 cm at the Summerland locality) that overlies the SL5 tephras and directly underlies the widespread Mount Rainier C pumice-and-scoria-fall deposit. The SL6 unit is distinguished from earlier Summerland period tephras by a markedly greater abundance of horizontally oriented plant needles that impart a dark brownish hue and a laminated appearance to the deposit, as well as by thin (to 0.5 cm), discontinuous strata of gray-to-light-gray gritty lithic ash (mostly 0.1–0.5 mm, rarely to 1 mm). Whitish hydrothermally altered (silicified?) lithic grains are present in these gritty laminae and increase in abundance in the upper quarter of the deposit. Minor post-depositional silicification of plant debris as hard white coatings and infillings is also developed in uppermost SL6, apparently precipitated by components leached from the overlying C pumice-fall deposit.

Glassy grains are rare in the SL6 unit, but include both poorly vesicular blocky, and fewer scoriaceous-to-pumiceous types. Glassy ash grains were sampled from the lower, middle, and upper parts of the SL6 deposit (Table 3, Fig. 8). Lower SL6 glassy grains are wholly blocky and microlitic with rhyolitic glass compositions (72.0–74.8 wt.% SiO2). Middle SL6 glassy grains are similar, including some with even more evolved glass compositions (72.2–76.5 wt.% SiO2), but two anomalous pumiceous ash grains were found that have dacitic glasses (69.3 wt.% SiO2), similar to pumiceous grains from the earlier SL3–SL4 deposits. Upper SL6 glassy grains are diverse, including blocky microlitic ash with rhyolitic glass (72.9–78.2 wt.% SiO2), pumiceous ash with evolved dacitic glass (69.1–69.5 wt.% SiO2) similar to some SL3–SL4 ejecta, and a single scoriaceous grain with less evolved dacitic glass (64.7 wt.% SiO2) that has abnormally low Al2O3 (14.5 wt.%) compared with other Mount Rainier glasses at similar SiO2 values. The origin of this dacitic grain is uncertain, although some andesitic glasses from Mount St. Helens have similarly low Al2O3 concentrations (T. Sisson, unpublished analyses).

The abundant plant material in the SL6 deposit, the scarcity of glassy grains, and the similarities between SL6 glasses and those of earlier Summerland tephras, suggest that SL6 time was a mainly non-eruptive interval when plants grew vigorously in sub-alpine meadows, trapping wind-blown sediments. However, the conspicuous hydrothermally altered ash grains, scarce-to-absent in earlier Summerland tephras, is evidence that eruptions took place and disrupted altered portions of the edifice. Possibly these were phreatic or weakly phreatomagmatic eruptions precursory to the subplinian C tephra event. A single radiocarbon sample (twig) from the SL6 unit gives an age of 2,420 ± 80 cal year BP (Table 2).

C tephra (SL7)

The ∼2,200 cal year BP C tephra (Table 2) is the largest-volume single Holocene fall deposit from Mount Rainier (Mullineaux 1974; Swanson 1993) and forms a broad lobe extending to the east–northeast of the volcano (0.2 km3 within 20 km of the volcano, or roughly 0.1–0.15 km3 as dense rock). The deposit attains a thickness slightly in excess of 30 cm along its main depositional axis, including at the Summerland locality where it is measured at 30–35 cm (7.5 km from the vent). The C deposit consists mainly of brown porphyritic andesite pumice lapilli and ash (plagioclase–hypersthene–augite–amphibole–Fe-Ti oxides), but includes common darker scoriaceous lapilli and bombs of slightly more mafic andesite (Table 6). Both the pumice and the scoria contain small (to 1 cm) but conspicuous white blebs and streaks of pumiceous, phenocryst-poor high-Sr dacite that contains rhyolitic glass (Venezky and Rutherford 1997). The fluidal shapes of these white pumiceous blebs and streaks are evidence that the crystal-poor dacite was molten when incorporated into the andesitic hosts. Some scoria pyroclasts also contain minor angular lithics and (or) fragments of hydrothermal clay, probably entrained from shallow conduit walls. Additional components (<20%) of the C deposit are lapilli of a cement-gray, crystal-rich, poorly inflated low-SiO2 dacite with moderately elevated Sr concentrations (Table 6), and dense prismatic lithics, many of which are high-Sr andesite similar in composition to the high-Sr lava flows exposed through the Emmons and Winthrop Glaciers. These high-Sr components are probably shallow conduit linings from that earlier magmatic event, exhumed by the subplinian C eruption.

Table 6 Whole-rock chemical compositions of ejecta from the 2,200 cal year BP C-tephra eruption of Mount Rainier

Heterogeneity between and within the juvenile C ejecta results from juxtaposition and mingling of magmas shortly before and during the eruption (Mullineaux 1974; Swanson 1993; Venezky and Rutherford 1997). The dominant brown pumice lapilli have abundant strongly sieve-textured plagioclase and pyroxene phenocrysts containing wormy melt inclusions and embayments that are variably overgrown by non-embayed rinds to as much as a few tens of microns thick. These textures record an event of phenocryst resorption due to heating, mixing, or fluxing by introduced fluids, followed by a period of minor crystal regrowth. Small, irregularly shaped amphibole grains in the glassy matrix of the brown pumice appear to be stable post-resorption minerals, but they might also be partly resorbed relicts of earlier amphibole phenocrysts, although they lack reaction-rim overgrowths common on unstable volcanic amphiboles. Microlites and microphenocrysts in the matrix of the brown andesitic pumice are also post-resorption-event crystalline phases, though their abundance is low compared with the non-pumiceous glassy ash grains that dominate the earlier Summerland period tephras. Late magma mingling is represented by co-eruption of pumice and scoria with distinct compositions, and by the late-stage entrainment of the crystal-poor white dacite magma into both andesite types.

Despite textural evidence for a major phenocryst resorption event, analyzed C pumice lapilli have nearly homogeneous dacitic glasses (64.3–64.7 wt.% SiO2) (Table 3). Venezky and Rutherford (1997) present a wider compositional range for glass in brown C pumice, attributed mainly to local, syn-eruptive blending with rhyolitic melts from the entrained white dacite blebs. The homogeneous glasses in the brown C pumice are indistinguishable in composition from glasses in the hornblende-bearing pumice that erupted during the earlier SL2 event (Table 3, Figs. 8 and 9). This, and the presence of amphibole in both deposits (generally minor at Mount Rainier), raises a speculative possibility that the dominant C andesite might have been remobilized SL2 magma stored beneath the volcano. The sharply bounded white dacitic pumiceous blebs scattered in the C andesitic pumice and scoriae have microlite-free rhyolitic glasses (Table 3; Venezky and Rutherford 1997), as do the cement-gray, crystal-rich, poorly inflated dacite lapilli. Pervasive quench crystallites precluded glass analyses in C andesitic scoriae bombs.

SL8

The coarse, porous character of the C tephra makes it a poor substrate for subsequent deposits, with the result that younger tephras are disturbed or absent in many areas around Mount Rainier; however, a thin (1–2 cm), fine-grained, gray ash deposit (SL8) overlies the C tephra at some localities on the volcano’s east flank, commonly separated from the underlying C tephra by up to 1 cm of brown, probably non-eruptive silty ash (Fig. 7). The coarser fraction (to 0.5 mm) of the SL8 unit consists chiefly of plagioclase and orthopyroxene phenocrysts and phenocryst fragments coated with clear, dense-to-moderately vesicular microlite-bearing glass. The dominant finer fraction consists mainly of gray, angular-to-fluidal shaped grains of similar microlite-bearing clear glass. Brown blocky-to-scoriaceous glassy grains similar to dark-colored glassy grains in SL1 through SL6 deposits are also present, along with minor fine lithics, but are subordinate to the grayish grains with clear glass. Glasses analyzed from the SL8 unit are rhyolitic with a restricted compositional range (72.7–73.4 wt.% SiO2) (Table 3, Fig. 8). A distinguishing feature of these glasses is that they have higher Al2O3 concentrations (13.3–13.5 wt.%) than most Holocene Mount Rainier glasses with similar SiO2 values (commonly 12–13 wt.% Al2O3 at 73 wt.% SiO2), consistent with fast ascent and limited degassing-driven growth of plagioclase microlites. The abundance, consistent texture, and appearance of the grayish glass-bearing grains, along with the limited and distinctive compositional range of their glasses, are evidence that the SL8 deposit resulted from a magmatic eruption following the large C event. No ages have yet been obtained from the SL8 deposit to determine if it belongs with the Summerland tephras, or if it erupted at a much later time.

∼1,500 year BP tephras and Twin Creek lahar episode

Two fine-grained Mount Rainier tephras (TC1, TC2) overlie the SL8 deposit at Summerland and elsewhere and underlie the widespread Mount St. Helens W pumiceous ash (Fig. 7). Bracketing radiocarbon samples (twigs) establish that these two post-SL8, pre-MSH-W ashes were deposited at ∼1,500 cal year BP (Table 2 and Vallance et al., in prep). These ashes were deposited approximately concurrent with the emplacement of clay-poor lahars of Mount Rainier’s Twin Creek (TC) lahar episode (weighted mean age of 1,510 ± 110 cal year BP from 11 Twin Creek 14C samples, Zehfuss et al. 2003) whose runout sands reached as far as the (present) Port of Seattle, 130 km of river distance from the volcano.

TC1

The TC1 tephra consists of a single stratum of fine-to-medium-grained, poorly sorted, brownish gray ash, 2–3 cm thick at the Summerland locality (Fig. 7). The most abundant components of the deposit are angular gray lithic grains (to 2.5 mm), and plagioclase and pyroxene phenocryst fragments, but dark-brown-to-black glassy grains are common. These glassy grains range from angular and blocky to fluidal and scoriaceous, and are rich in microlites with distinctively acicular habits. Minor TC1 components are hydrothermally altered ash grains, and brown pumiceous ash that may be reworked from the voluminous C tephra deposit. Glasses in the dark, blocky-to-scoriaceous grains are wholly rhyolitic (73.9–77.2 wt.% SiO2), and are more evolved than glasses from the preceding SL8 deposit (Table 3; Fig. 8). The absence of dense grains with dacitic glasses, like those in many pre-C Summerland tephras, is evidence against a reworking origin for the TC1 glassy grains. Instead, the glassy grains are probably a minor juvenile magmatic component to what may have been a mainly phreatomagmatic eruption.

TC2

The TC2 tephra is a fine-grained, moderately sorted, dark gray ash stratum, ∼1 cm thick at the Summerland locality, that overlies the coarser and browner TC1 tephra (Fig. 7). The darker color of the TC2 deposit results from a much higher abundance (20–30 vol.%) of dark-brown-to-black glassy ash grains, many with fluidal-scoriaceous shapes. The remainder of the TC2 deposit is made up of diverse angular lithics and phenocryst fragments. The dark glassy grains are microlite-rich with rhyolitic glasses (73.0–75.1 wt.% SiO2) that are generally less evolved than those of the underlying TC1 deposit (Table 3, Fig. 8). The contact between the TC1 and TC2 deposits is sharp without intergradation, consistent with two eruptions, but the contact lacks incision or accumulations of organic-rich non-eruptive sediment, consistent with only brief repose between the TC1 and TC2 events.

∼1,000 year BP ash in the upper White River valley

Hoblitt et al. (1998) report a thin, fine-grained ash deposit that is exposed near the top of a terrace on the south side of the White River in Mount Rainier National Park, shortly west of the road bridge to Sunrise, and date this ash to 1,080 ± 250 14C year BP (1,040 ± 410 cal year BP) from fir needles separated from the deposit (R. Hoblitt, 2008, written communication). This deposit (1–2 cm) underlies the 471 cal year BP (AD 1479) Mount St. Helens W tephra (Yamaguchi 1983; Fiacco et al. 1993) and overlies a thin lahar deposit (35 cm) that caps the Mount Rainier C pumice-fall deposit in that area. The fine-grained ash deposit is poorly sorted, containing common brown pumice in the ash fraction coarser than 1 mm, probably reworked from the Mount Rainier C and SL3–SL4 deposits, but it also contains abundant dark glassy ash grains (<1 mm). Most of these glassy grains are dense, angular, and blocky, with few vesicles, but fluidal-scoriaceous glassy grains are also present. All of the dark glassy grains are rich in microlites and have uniformly rhyolitic glass compositions (72.0–73.8 wt.% SiO2). Glasses are distinguished by their lower SiO2 concentrations from the earlier TC1 and TC2 tephra glasses (Fig. 8), and by their lower Al2O3 concentrations from the similar-SiO2 SL8 tephra glasses (Table 3). A potentially correlative thin (∼1 cm) fine-grained ashy deposit is present on the opposite (north) side of the White River and yields a calibrated radiocarbon age of 980 ± 80 cal year BP (Vallance et al., in prep.), but glasses in that sediment have not been analyzed.

The common dark glassy grains with their restricted glass compositions are consistent with a magmatic eruptive origin for the ∼1,000 cal year BP tephra deposit. The fine grain-size and localization to a valley floor setting could result by deposition as fallout from the dilute ash cloud from a pyroclastic flow (Hoblitt et al. 1998). No contemporaneous pyroclastic flow deposits have been found in the area, but far-traveled lahars and runout sands of the Fryingpan Creek lahar episode were emplaced down the White River system at about this time (weighted mean age of 1,120 ± 70 cal year BP from four 14C samples from Fryingpan Creek deposits, Zehfuss et al. 2003). Possibly, pyroclastic flow(s) transformed directly to lahars during transit down the large Emmons Glacier, leaving no primary deposits other than the fine-grained tephra.

∼500 year BP deposits

An enduring issue has been if an eruption triggered the voluminous, clay-rich Electron Mudflow of ∼500 cal year BP, or if the mudflow started by non-magmatic processes such as spontaneous gravitational failure of an altered edifice flank, dislodgement by shaking from a tectonic earthquake, or perhaps by a heavy rainfall event (Scott 2004). To investigate this we examined several ashy deposits immediately beneath and above the 471 cal year BP MSH-W tephra, searching for evidence of contemporaneous eruptive activity from Mount Rainier. None of the ashy deposits yielded compelling evidence for an eruptive origin, but a lahar deposit exposed at the confluence of the Main and West Forks of the White River, potentially age-correlative with the Electron Mudflow (Crandell 1971, Scott et al. 1995), contains a texturally and compositionally consistent clast type that could represent juvenile eruptive material. Without a corroborating tephra deposit, the evidence is weak for an eruptive trigger for the Electron Mudflow. See Electronic Supplementary Materials for descriptions of the lahar and of the non-eruptive ashy sediments.

Mount Rainier X-tephra

Mullineaux (1974) interprets lapilli of brown pumice and scoria scattered on some neoglacial moraines as a deposit from a minor pumice eruption of Mount Rainier, dated by tree ring counts to between AD 1820 and AD 1854. This deposit was named the X tephra, not to be confused with the similarly young X tephras erupted from Mount St. Helens and Glacier Peak, Washington (Mullineaux 1996, Beget 1982). At localities below the Emmons, Inter, Winthrop, and Ohanapecosh Glaciers (Fig. 1) we confirmed that brown pumice lapilli are scattered locally on neoglacial moraines, but in each case these lapilli are indistinguishable from Mount Rainier C-tephra pyroclasts based on presence of resorbed and overgrown phenocrysts, minor amphibole, and distinctive dispersed blebs and streaks of white pumiceous dacite. Below the Emmons Glacier the brown pumice lapilli are most abundant on young moraines below a large avalanche chute descending from Burroughs Mountain, which is carpeted with C tephra. Other localities where Mount Rainier X tephra has been reported are also surrounded by highlands where Mount Rainier C tephra is abundant, and we interpret the brown pumice and scoria lapilli scattered on most of the young moraines as Mount Rainier C tephra that was redeposited by snow avalanches, not by a mid nineteenth century pumice eruption. The situation below the Ohanapecosh Glacier is somewhat different. There, Sigafoos and Hendricks (1972) describe a concentration of presumed Mount Rainier X pumice within the area of neoglacial moraines. Reexamination of that locality shows the presence of a low ridge of disturbed Holocene tephras that survived erosion by thin neoglacial ice advance. The pumiceous lapilli are Mount Rainier C tephra, underlain by disturbed Mount St. Helens Y pumiceous ash, underlain by locally exposed Mount Rainier F tephra.

Other historic eruptions

Hopson et al. (1962) briefly summarize and generally discount reports of Mount Rainier eruptions on 14 occasions between AD 1820 and AD 1894. No physical evidence has been found in this study for these possible eruptions, but accounts of the November–December 1894 event, edited by Majors and McCollum (1981a,b), and dispatches by WM Sheffield and ES Ingraham (some by homing pigeon) while on a winter expedition to Mount Rainier in December 1894 to investigate the event, include credible descriptions, one at close range on 24 December 1894, of small, dark eruption plumes rising from the summit. Many other claims for the 1894 eruption, such as that the profile of the mountain was changed or that there were large rock avalanches, did not survive later scrutiny. Independent confirmation for an eruption is, however, lacking and a clear photograph of the volcano from Tacoma, dated 29 December 1894, shows neither eruptive plumes nor dark tephra on the summit snowfields (University of Washington Special Collections image WAT042). We consider it possible that phreatic eruptions took place in late 1894, but were too small to leave preserved deposits. Likewise, some of the earlier discounted events may have been of similar scale, but the last magmatic eruption that can be amply documented was the ∼1,000 cal year BP event that deposited ash in the White River and generated the Fryingpan Creek lahar assemblage (Hoblitt et al. 1998; Zehfuss et al. 2003).

Correlations of flow and fall deposits based on glass and mineral compositions

Compositions of glass and plagioclase phenocrysts can be used to refine the inferred sequence of events during the Summerland eruptive period. Previously it was noted that glass compositions in the SL3–SL4 tephra units match those of glass in breadcrust bomb rinds from the South Puyallup pyroclastic flow deposit. This and the presence of brown pumiceous ash and rare brown pumice lapilli in the SL4 unit are evidence that the South Puyallup block-and-ash flow erupted as part of the SL3–SL4 episode, dated in combination at 2,560 ± 120 cal year BP.

Paleomagnetic measurements establish that the South Puyallup block-and-ash flow and the high-Sr Emmons–Winthrop lava flows were erupted at close to the same time (Hagstrum and Champion 2002; Vallance et al., in prep), but cannot determine the order of those events. Strontium is compatible-to-strongly compatible in plagioclase, relative to melt, so we use electron-microprobe analyses of Sr in plagioclase phenocrysts (see Electronic Supplementary Materials for analytical methods) to search for tephras correlative to the high-Sr Emmons–Winthrop lava flows (Fig. 10, Table 7). Plagioclase phenocrysts from SL1 through SL4 juvenile tephra grains and from the South Puyallup block-and-ash flow have similar ranges in Sr and anorthite concentration, with the exception of the distinctly higher-anorthite plagioclase phenocrysts in the SL2 amphibole-bearing pumice lapilli that are consistent with plagioclase grown from melt with higher dissolved H2O concentrations (Yoder et al. 1957; Sisson and Grove 1993). High-Sr plagioclase phenocrysts appear abruptly in the SL5 tephras and persist in the SL6 tephras. These high Sr concentrations overlap those of phenocrysts from the Emmons–Winthrop high-Sr lava flows. These results lead to the interpretation that the SL5, and perhaps SL6, tephras were deposited concurrent with effusion of the high-Sr Emmons–Winthrop lavas, and therefore that the high-Sr lavas effused shortly after eruption of the South Puyallup block-and-ash flow and SL3–SL4 tephras.

Fig. 10
figure 10

Stratigraphic succession of average plagioclase phenocryst compositions for late-Holocene Mount Rainier deposits, subdivided as Summerland eruptive period ash phenocrysts (blue diamonds), pumice lapilli (circles), South Puyallup breadcrust bombs (yellow square), C-tephra resorbed and overgrown phenocrysts (tan circle), non-resorbed phenocrysts (light blue circle), and phenocrysts in white dacitic blebs (white circle) within andesitic pyroclasts. Phenocryst compositions from lava flows are shown at inferred stratigraphic levels as blue and red bars. Brackets show one standard deviation

Table 7 Summary of electron-microprobe analyses of plagioclase phenocryst compositions from late-Holocene Mount Rainier eruptive products

Relative timing of the Mount Rainier C tephra and the east summit crater lava flows was also not resolved by paleomagnetic and radiocarbon methods. The east summit crater lava flows have normal Sr concentrations for Mount Rainier eruptives, as do their relatively simple-textured plagioclase phenocrysts (Fig. 10, Table 7). In contrast, the gray dacite, white dacite blebs, and prismatic lithic components in the C tephra have elevated whole-rock Sr concentrations. Plagioclase phenocrysts in the dominant brown andesitic C pumice can be categorized into three types. The highest Sr grains are water-clear idiomorphic phenocrysts within the blebs and streaks of white, pumiceous dacite dispersed in the dominant andesitic pumice and scoriae. Phenocrysts directly in the groundmass of the brown pumice can be subdivided into simple and strongly resorbed-overgrown textural types. Neither type has notably elevated Sr concentrations, but both have high anorthite concentrations that, along with the presence of amphibole, are consistent with higher magmatic H2O concentrations than for typical amphibole-free or -poor Mount Rainier eruptives. Fluidal shapes of the white dacite blebs and streaks show that the crystal-poor dacite was molten when entrained into the andesite, and Venezky and Rutherford (1997) interpret based on phase equilibrium results that the dacite was stored at a very shallow level prior to eruption.

Together, these results are consistent with the high-Sr components in the C tephra (prismatic lithics, crystal-rich gray dacite lapilli, white dacite blebs) originating as variably solidified conduit linings and residual melt segregations left from the Emmons–Winthrop high-Sr effusive episode. Such high-Sr conduit fillings likely would have been removed if the east summit crater magmas had erupted between the Emmons–Winthrop and C episodes. High-Sr plagioclase phenocrysts have also not been found in east summit crater lava samples, consistent with high-Sr material having been removed by the time the east summit crater magmas passed through the conduit system. These factors lead to the interpretation that the C eruption followed the Emmons–Winthrop–SL5–SL6 event(s) and exhumed residual high-Sr conduit fillings. The C eruption was then followed by effusions of the normal-Sr andesite magmas now preserved as the rim of the east summit crater. Effusion of the east summit crater andesites may correspond with deposition of the fine-grained SL8 tephra.

Summary eruptive history of Mount Rainier over the last 2,600 years

Stratigraphic, compositional, and geochronologic results document ten–12 distinguishable eruptions of Mount Rainier over the last ∼2,600 years BP (Fig. 11, Table 4). Each eruption probably consisted of multiple explosive events spanning months to possibly years. Additional smaller eruptions may have taken place, but did not leave deposits that have been recognized. Seven or eight eruptions took place over the period 2,600–2,200 cal year BP and compose the Summerland eruptive period. Deposits from these eruptions are (1) SL1—magmatic and phreatomagmatic ash plus minor pumice lapilli (plagioclase–pyroxene), (2) SL2—magmatic and phreatomagmatic ash plus minor pumice lapilli (plagioclase–pyroxene–amphibole), (3) SL3–SL4—magmatic and phreatomagmatic ash, minor pumiceous ash and lapilli (plagioclase–pyroxene), contemporaneous with the South Puyallup block-and-ash flow, (4) SL5—high-Sr magmatic and phreatomagmatic ash contemporaneous with the high-Sr Emmons–Winthrop andesite lava flows, (5) SL6—infrequent phreatomagmatic ash eruptions precursory to the C tephra, (6) the subplinian C pumice and scoria deposit, (7 or 8) east summit crater andesitic lava flows accompanied, preceded, or followed by the SL8 tephra. Lavas may also have effused during the SL1, SL2, and SL3–SL4 events, but any flows are concealed by subsequent lava flows and by ice. Small-volume block-and-ash flows may also have accompanied those eruptions, but no primary deposits survive other than in the South Puyallup drainage. Lahars of the Summerland lahar assemblage (Zehfuss et al. 2003) accompanied the Summerland eruptions.

Fig. 11
figure 11

Summary of late-Holocene eruptive events at Mount Rainier. Red circles designate eruptions with direct physical evidence for ejection of juvenile magma, orange circle represents possible eruption documented only by uniform clast type in White River confluence lahar, light blue circle represents reported 1894 eruption for which physical evidence is lacking. Other historically reported small eruptions might have taken place but also lack corroborating physical evidence. See Table 4 for approximate eruption magnitudes and correlations with lahars

Three or four eruptions are documented following the Summerland eruptive period. Deposits from these eruptions are (1, 2) the fine-grained TC1 and TC2 tephras that erupted near 1,500 cal year BP and correlate with the Twin Creek lahar assemblage (Zehfuss et al. 2003), (3) the ∼1,000 cal year BP fine-grained ash in the White River valley floor, and speculatively, (4) the ∼500 year BP lahar at the confluence of the main and west forks of the White River. The fine-grained character of the post-Summerland ashes, as well as the contemporaneous formation of far-traveled lahars at ∼1,500 and ∼1,000 cal year BP (Zehfuss et al. 2003) are consistent with eruption of small-volume pyroclastic flows or surges that transformed to lahars while transiting ice, or with deposition of hot fallout tephras onto the upper mountain ice cap, leading to melting and lahar generation.

The results presented here greatly increase the number of physically documented late-Holocene Mount Rainier eruptions. Most of these eruptions were small-to-moderate (Table 4), and some left no known primary volcanic deposits other than thin, fine-grained, dominantly lithic tephras. Nevertheless, the eruptions were penecontemporaneous with far-traveled lahars, showing that even modest magmatic or phreatomagmatic eruptions of Mount Rainier can produce hazardous lahars, irrespective of lava effusion or dome growth. Despite considerable effort, however, evidence is weak for a magmatic eruption associated with the ∼500 cal year BP Electron Mudflow, and we cannot rule out that sizeable lahars may also form at Mount Rainier without a direct eruptive trigger.