This study combines high precision Global Positioning System (GPS) and Interferometric Synthetic Aperture Radar (InSAR) measurements with seismic observations and mechanical modeling to better understand volcanic systems. The approach combines data collection and interpretation with theoretical and numerical analysis. Continuous GPS measurements collected over the past 10 years in Hawaii in a collaborative effort with the Hawaiian Volcano Observatory (HVO) have shown: 1) inflation and deflation cycles on Kilauea volcano correlated with eruptive activity at P'u 'O'o; 2) the onset of summit inflation at Mauna Loa volcano; 3) apparent temporal correlations between deformation at Kilauea and Mauna Loa; 4) seaward motion of both Kilauea and Mauna Loa south flanks; 5) eight hours of rift extension preceding an eruption in 1997; rapid deflation accompanying this eruption followed by re-inflation; 6) recovery of the P'u 'O'o eruption when the summit regained its pre-eruptive state; 7) six months of transient rift extension following the 1997 eruption; and 8) a series of four silent earthquakes.
The silent earthquakes are equivalent in size to a magnitude 5.7 but occur over a period of 1 to two days, rather than a few seconds. They appear to repeat in the same area, immediately seaward of a narrow band of earthquakes. The GPS data are being analyzed to determine the source locations with respect to the sources of long-term flank deformation and seismicity on Kilauea. The raw GPS phase data are used to determine event durations. Another aspect of the work is to see if the silent events are associated with non-volcanic tremor or other unusual seismic signals. The sources of longer-term transient deformation, including potential Mauna Loa-Kilauea interactions and magmatic tectonic interactions are also being analyzed. The latter involve the integration of InSAR and GPS and will test specific structural models inferred from seismicity and other geophysical and geologic data.
A second focus is imaging and modeling time dependent deformation accompanying dike intrusion. Analytical mechanical models are explored to elucidate the factors that determine whether a dike reaches the surface and erupts or ends with intrusion, and whether it is in principal possible to determine this while the process is occurring. Time dependent inversion methods have been developed, but the spatial resolution of geodetic data is limited. Including the locations of swarm seismicity accompanying intrusions helps to improve spatial resolution. Dieterich's (1994) theory provides the key link between dike induced stress and seismicity rate. This allows one to invert geodetic data and earthquake rates for the shape of the dike, the excess dike pressure, and other parameters.
The final objective is relating the observations to physical models of dike propagation. This work extends extends existing lumped parameter models by including thermal effects and magma density changes induced by vesiculation. This should allow better understanding of the growth of dikes and their induced deformation with time, and to quantify the factors that govern eruption vs. intrusion.
Lastly, the project will replace failing receivers in the collaborative GPS network, many of which are more than a decade old. This project will support two Ph.D. students, one focusing on seismic and aseismic processes including silent earthquakes, the other magmatic processes including dike intrusion. Four past or present Ph.D.'s (3 female students) were based at least in part on this project. In addition, 10 other graduate students or Postdocs have participated in this research. Our research is fully collaborative with HVO where we have worked with six project scientists over the past 16 years.
