Search

Hydrothermal eruptions are the most violent and hazardous phenomena within geothermal fields. The largest of these may produce kilometer-sized craters and breccia deposits that are tens of meters thick. The geological and hydrothermal priming that leads to these types of eruptions is poorly understood. To understand large hydrothermal eruptions, we investigated a series of prehistoric events at the Rotokawa geothermal field in New Zealand. By revising the stratigraphy and distribution of hydrothermal breccia deposits and correlating these with componentry, crater morphology, and subsurface geological structure, we estimated the frequency, priming processes, triggers, and dynamics of multiple eruptions. Seventeen large hydrothermal eruptions occurred centuries to millennia apart in the period from ca. 22 cal ka B.P. to ca. 3.4 cal ka B.P. Of six hydrothermal eruptions since ca. 7 ka, four produced oval-shaped craters up to 2 km in diameter, creating a broad, shallow depression within the geothermal field. The two youngest eruptions occurred northeast of earlier eruption centers and have narrower and elongated vents. We infer that in the central depression, newly formed craters rimmed by breccia deposits and high-relief country rock hosted temporary lakes tens of meters deep. Crater-lake breakout(s) and/or seismic events caused sudden pressure reduction above the hydrothermal aquifer, triggering hydrothermal eruptions. Northeast of the basin, hydrothermal alteration produced caprocks above intensively fractured areas. In this case, earthquakes are the most likely trigger for cap-rupture and eruption. All eruptions excavated shallow and large craters mostly within partially altered Oruanui Formation and pre-fragmented breccias. The size and localization of the eruptions was likely due to a combination of (1) availability of undisturbed porous ignimbrite hosting large thermal aquifers, (2) efficient crater excavation within or alongside pre-fragmented breccia, and (3) the location of fracture and fault zones that channeled deep fluid upflow, favoring priming processes. This study highlights how an interplay of tectonic, magmatic, and hydrologic processes is responsible for the timing, dynamics, and ultimate size of hydrothermal eruptions in geothermal fields. Some events may be very large and destructive depending on the right priming and geological conditions. [ABSTRACT FROM AUTHOR]

Steam-driven eruptions are caused by explosive vaporization of water within the pores and cracks of a host rock, mainly within geothermal or volcanic terrains. Ground or surface water can be heated and pressurized rapidly from below (phreatic explosions), or already hot and pressurized fluids in hydrothermal systems may decompress when host rocks or seals fail (hydrothermal eruptions). Deposit characteristics and crater morphology can be used in combination with knowledge of host-rock lithology to reconstruct the locus, dynamics, and possible triggers of these events. We investigated a complex field of >30 craters formed over three separate episodes of steam-driven eruptions at Lake Okaro within the Taupo volcanic zone, New Zealand. Fresh unaltered rock excavated from initially >70 m depths in the base of phase I breccia deposits showed that eruptions were deep, "bottomup" explosions formed in the absence of a preexisting hydrothermal system. These phreatic explosions were likely triggered by sudden rise of magmatic fluids/gas to heat groundwater within an ignimbrite 70 m below the surface. Excavation of a linear set of craters and associated fracture development, along with continued heat input, caused posteruptive establishment of a large hydrothermal system within shallow, weakly compacted, and unconsolidated deposits, including the phase I breccia. After enough time for extensive hydrothermal alteration, erosion, and external sediment influx into the area, phase II occurred, possibly triggered by an earthquake or hydrological disruption to a geothermal system. Phase II produced a second network of craters into weakly compacted, altered, and pumice-rich tuff, as well as within deposits from phase I. Phase II breccias display vertical variation in lithology that reflects top-down excavation from shallow levels (10-20 m) to >70 m. After another hiatus, lake levels rose. Phase III hydrothermal explosions were later triggered by a sudden lake-level drop, excavating into deposits from previous eruptions. This case shows that once a hydrothermal system is established, repeated highly hazardous hydrothermal eruptions may follow that are as large as initial phreatic events. [ABSTRACT FROM AUTHOR]