No abstract

The onset of anomalous seismic activity in 1989 beneath Mammoth Mountain on the southwestern rim of the Long Valley caldera, California, was followed within -•4 months by a large increase in 3He/4He in vapor discharged from a fumarole on the north side of the mountain. The helium isotopic ratio at this vent rose to a maximum of 6.7 RA in July 1990 and subsequently declined to values near 3 5 RA. Potential sources of the He-rich vapors include degassing of fresh magma, degassing from fresh surfaces generated in newly fractured igneous rocks, and volatile release from a 3He-rich gas chamber situated above previously emplaced intrusives. The magnitude of the increase in helium isotopic composition (from 3.8 to 6.7 R•), the persistence of relatively high values (>5 R•) over a period of 3 years, the increase in the flux of total He relative to gases in air-saturated water, and the increases in the rates of discharge of steam and gas from this fumarole indicate that magmatic intrusion did in fact begin in 1989 beneath Mammoth Mountain. Seismic activity and limited measurements of extensional deformation at the surface suggest that the depth of intrusion may be as shallow as 2 km, consistent with the prompt appearance of increased 3He/4He ratios in the fumarolic gas, and that the intrusive process may have persisted for -• 1 year. In contrast, a similar combination of magmatic intrusion and anomalous seismic activity beneath the resurgent dome-south moat region during the 1989-1991 period resulted in at most relatively small changes in 3He/4He in fumarolic discharge at the southern edge of the resurgent dome. The more subdued response may result from a combination of greater intrusive depths and greater dilution of 3He-rich inputs to thermal fluid reservoirs in the shallow hydrothermal system in this area compared with Mammoth Mountain.

In 1985, Unocal drilled the first deep test well in the western moat of Long Valley caldera, ∼1 km east of the Inyo Craters. Well IDFU 44‐16 penetrated a thinner section of Bishop Tuff than that found in earlier deep exploration wells on the resurgent dome. Precaldera volcanics were encountered at 1183 m and metamorphic basement at 1615 m. The precaldera volcanics yielded a radiometric age of 1.98±0.1 Ma and are tentatively correlated with Tertiary andesites and dacites on the caldera's western wall. These stratigraphic relations suggest that the caldera's structural margin is at least 4 km east of the current topographic margin. Temperatures reach a maximum of 218°C at 1100 m in IDFU 44‐16; however, temperature reversals occur below this depth. Temperature and permeability measurements indicate a lateral outflow of hot water over cold recharge near the caldera's structural margin. While these discoveries support the existence of hightemperature fluids in the western moat of Long Valley caldera, the ultimate source of the geothermal system remains to be discovered.

Detailed surface mapping, subsurface drill hole data, and geophysical modeling are the basis of a structural and hydrothermal model for the western part of Long Valley caldera. Six fault zones are recognized in the western caldera with dominant orientations of north, northwest, and northeast, sub‐parallel to regional fault trends in the surrounding Sierran basement. The internal structural geometry of the cores of 12 exogenous domes inside and outside the caldera suggests that the domes erupted at the intersections of these principal fault trends rather than along the axis of a single dike. Gravity modeling and subsurface data from deep geothermal wells indicate that the floor of the caldera is segmented into a number of discrete fault blocks with varying offsets. One of the northeast trending fault zones, designated Discovery fault in this paper, appears to be part of the original Sierran embayment that existed before caldera collapse. Recent hydrothermal alteration occurs along Discovery fault strands and composite vertical offset of intracaldera volcanic units across the entire fault zone may be as much as 400 m. Field relationships, geophysical interpretations, and interpretaions of drill hole data suggest that this fault is a fundamental flaw in the western caldera. The preexisting tectonic framework of the basement rocks controlled the configuration of the western caldera floor and, through it, the location of postcollapse eruptive centers. These deep basement structures may also provide the high fracture density which controls circulation in the present geothermal system of Long Valley.

Andesite inclusions are found within the vent areas of two 600‐year‐old rhyolite domes in the Inyo Craters volcanic chain, eastern California. Both domes lie within the present northwestern sector of 0.73 Ma Long Valley caldera. Inclusion morphology indicates that they were partially molten and ductile during incorporation within coarsely porphyritic rhyolite. The xenocryst assemblage within the inclusions (plagioclase + hornblende + biotite + quartz ± sanidine) is identical to the phenocryst assemblage within the host rhyolite. Microprobe analyses of xenocrysts within the andesite inclusions show that they have compositions similar to phenocrysts in the host rhyolite. These data suggest that the andesite inclusions are the result of mixing of the coarsely porphyritic rhyolite with a more mafic end member. Whole rock geochemistry of the inclusions generally supports a simple mixing model between the coarsely porphyritic rhyolite of the Inyo domes and typical postcaldera mafic lavas found throughout the western caldera moat. The inclusions provide the first evidence for involvement of basaltic magma in the 600‐year‐old rhyolite eruptions of the Inyo Craters volcanic chain and suggest the possibility that the rhyolitic eruptions were initiated by injection of basalt into the lower part of the silicic magma system. Periodic thermal replenishment of the magmatic system through basaltic injection may explain the relatively high temperatures encountered in drill holes within the western caldera moat. The inclusions further suggest that although rhyolite eruptions are statistically more likely from the Inyo Craters volcanic chain, future basaltic eruptions cannot be ruled out.