FIRE and MUD: Eruptions and Lahars of Mount Pinatubo, Philippines


The eruption of Pinatubo carried samples of ash far and wide, but a cooperative effort by field geologists and petrologists spread samples of Pinatubo pumice even farther and almost as fast. A number of fascinating discoveries ensued.

The lava dome that preceded explosive eruptions of 1991 consisted of hybrid andesite, the product of mixing of an unusually hydrous olivine-pyroxene basalt (present as quenched inclusions) and phenocryst-rich dacite (mostly assimilated into the mixture) (Pallister and others; Bernard and others; Hattori). Andesitic pumice of the June 12 eruption, compositionally identical to andesite of the dome, was gradually replaced by dacite in eruptions of June 13-15 (Hoblitt, Wolfe, and others; Pallister and others). Compositions intermediate between andesite and dacite were not found; rather, proportions of the two lithologies changed.

The climactic eruption produced two types of dacitic pumice--one was tan-gray and phenocryst poor, the other was white and phenocryst rich. The white was always dominant (80-90%); the proportion of tan-gray pumice decreased from about 20% to 10% from bottom to top in the tephra-fall deposit (David and others). The differences are mainly textural; hence, Pallister and others make an intriguing suggestion that intense explosions shattered phenocrysts of the once-phenocryst-rich magma into tiny crystal fragments of the phenocryst-poor magma. Luhr and Melson suggest that mechanical fragmentation occurred upon violent mixing of melt-rich dacitic magma into crystal-rich dacitic magma.

The dacitic magma was volatile saturated, exceptionally rich in water (Rutherford and Devine), and had a high oxygen fugacity (previously cited papers, plus Gerlach and others; Hattori; Imai and others). Most authors estimated the preeruption temperature of the dacitic magma to be about 780-800°C, on the basis of Fe-Ti oxide thermometry and the presence of cummingtonite as a stable phase. Luhr and Melson give a higher range, 840-890°C, based on the oxide thermometers of Ghiorso and Sacks (1991) and Anderson and others (1993). They attribute the discrepancy to either a late-stage heating event that affects Fe-Ti oxides but not cummingtonite or to the need for a correction for highly oxidizing conditions (as applied by Rutherford and Devine and other authors). Preeruption pressure is estimated to have been about 2.2+-0.5 kbar (7-11 km depth) from the Al-in-hornblende geobarometer, confirmed and refined to about 2.0 kbar by experimentally determined phase relations for the cummingtonite-bearing dacite (Rutherford and Devine).

Anhydrite was a relatively abundant, primary phenocryst phase. Sulfur isotopic composition of single grains suggested that anhydrite of the June 12 (vent-clearing) eruption was both primary and hydrothermal in origin, while that of the June 15 eruption was mostly primary (McKibben and others); sulfur of anhydrite in basalt was so light (about 1 per mil) that it is probably not the sole source for the bulk sulfur isotopic composition of June 15 dacite of 5 to 7 per mil. The presence of anhydrite inclusions in hornblende and plagioclase phenocrysts suggests to Fournelle and others that anhydrite was present in magma at depth and that prehistoric sulfur-rich eruptions might be found by a search for such inclusions. Pasteris and others note an absence of anhydrite inclusions in their samples and conclude (on that basis in part) that the dacitic magma reached saturation with respect to a CO2-SO2-H2O fluid before anhydrite saturation. The two conclusions are potentially compatible because, as concluded by Gerlach and others, a separate volatile phase probably existed at depth, potentially before anhydrite crystallization and yet still early and deep enough for subsequent anhydrite to be included within plagioclase and hornblende.

The high oxygen fugacity and high sulfur content of Pinatubo magma are closely related; the former makes the latter possible. Indeed, Pinatubo magma may be a prime example of an important class of hydrous, highly oxidizing, sulfur-rich, anhydrite-bearing magmas that produce eruption clouds so rich in SO2 that they can influence global climate (Gerlach and others; Luhr and Melson; Self and others). The high oxygen fugacity might also inhibit early, deep sulfide fractionation and allow abundant sulfur to remain in the magma until that magma is emplaced at shallow depths, where porphyry copper deposits can then form (Hattori; Imai and others).

The immediate source of sulfur of the eruption cloud is still under debate, as discussed in an earlier section by Gerlach and others and in this section by Rutherford and Devine, Pallister and others, Bernard and others, McKibben and others, Fournelle and others, and Pasteris and others. Evidence for a separate volatile phase at depth is strong; some contribution from anhydrite breakdown during magma ascent also seems possible.

Oxygen isotope data (Fournelle and others) and compositions and textures of olivines, pyroxenes, and spinels (Pallister and others) argue against significant contamination from the underlying Zambales Ophiolite Complex; beryllium isotopes show no evidence for large input from marine sediments (Bernard and others). Strontium-rubidium and neodymium-samarium isotope ratios for dacite and andesite pairs are identical; basalt has slightly more radiogenic strontium and slightly less radiogenic neodymium (Bernard and others). Bernard and others considered lead isotopes in andesite and dacite to be identical and to thus reflect a common origin. In contrast, Castillo and Punongbayan found that dacite was slightly less enriched in incompatible trace elements and in 87/86Sr and 206/204Pb and slightly higher in 143/144Nd than the andesite. Castillo and Punongbayan attributed the difference to variable mixtures of depleted basalts (mantle source) and an enriched component from subducted sediments.

In an unusual confluence of geophysical and petrologic studies, seismic monitoring captured an intrusion of basalt into a silicic magma reservoir (White), thereby providing unusual time constraints on magma mixing and disequilibrium reequilibration processes (Pallister and others; Rutherford and Devine). Estimates from this natural laboratory are broadly consistent with those from controlled laboratory experiments, and efforts to resolve minor differences in time estimates may bring yet more insights into these processes.

Subhedral anhydrite

Subhedral anhydrite in phenocryst-rich dacite pumice of the climactic, June 15, 1991, eruption. Crossed nicols; field of view, 0.34 mm maximum dimension. Photograph by John Pallister (USGS).

FIRE and MUD Contents | Next

PHIVOLCS | University of Washington Press | U.S.Geological Survey
This page is <>
Contact: Chris Newhall
Last updated 03.15.99