A REVISED MODEL FOR THE FORMATION OF TRONA DEPOSITS
Modeling using the USGS computer code SNORM shows that groundwaters from the San Bernardino Basin (AZ) would yield over 60% trona (by mass) upon evaporation, but waters from the four adjacent basins would yield none, in spite of the fact that they are dominated by granitic/rhyolitic lithology. The Hardie-Eugster model for evaporite formation indicates that San Bernardino waters could form trona because they have a much higher ratio of alkalinity:Ca2+ than do the waters of the adjacent basins. This higher ratio is evidently a result of magmatic CO2 injection related to the Geronimo volcanic field.
SNORM modeling of tributaries of the Owens River (CA) draining the granitoid Sierra Nevada shows significant differences in the mineral assemblages resulting from evaporation of different tributaries. Waters associated with Long Valley caldera (currently emitting significant CO2) tend to yield significant trona; waters not associated with the caldera yield none. The episodic nature of trona deposition in the terminal basin of the Owens River, Searles Lake (CA), suggests that trona formation there is not solely dependant upon lithology, but rather on a combination of lithology and episodic changes in CO2 injection from Long Valley caldera. Finally, a survey of major trona-forming lakes and trona deposits shows that all are/were likely associated with excess CO2 from magmatic gas or microbial respiration.
Based on these observations, we propose that, in addition to the presence of significant volcanic rock or volcaniclastic sediment, the addition of excess CO2 (from the mantle, microbial respiration, or some other source) is an important condition for the formation of trona. The association of trona deposits with CO2 injection could be useful as a prospecting tool.