Seimoacoustics from Kilauea Volcano using multiple arrays David Fee and Milton Garces Infrasound Laboratory Univ. of Hawaii, Manoa email@example.com Tim Orr, David Wilson US Geological Survey Hawaiian Volcano Observatory
Overview • Introduction to Kilauea • Arrays • Observations in Chronological Order • Multiple types of tremor • Skylight Detection • Two fissure eruptions • Acoustic energy vs. seismic/tilt event • Conclusions
Pu`u` O`o Crater Complex • Pu`u O`o crater complex is the active section of Kilauea Volcano • Nearly continuous effusion and degassing from multiple vent system since 1992 • Lava flowed from the crater complex ~13 km through a complicated conduit system until it met the ocean • Multiple open vents and lava tube skylights appear active • System changed in June 2007 fissure eruptions
Infrasound Arrays – MENE MENE Array • Semi-permanent • 12.5 km from Pu`u O`o • Dense Jungle • Since 9/06
Infrasound Arrays – KIPU KIPU Array • 2 week deployment in 4/07 • 2.5 km from Pu`u O`o • High S/N • Collocated BB Seis
STC Seismometer Infrasound Arrays - ZION ZION Array • Since 9/07 • 7 km from Pu`u O`o • 6.5 km from Fissure D • Better azimuthal resolution • Results preliminary
Infrasonic Tremor for 12/1/2006 • Usually constant between ~0.5-3 Hz • Harmonic, gliding tremor as well • Broad harmonics (~2 Hz) • High Attenuation? • Gliding (spectral peaks change frequency over time) • Magma level may define effective length of conduit • Seismic spectrogram from same period has relatively uniform tremor, some harmonics • Tremor not visible in waveforms
KIPU KIPU results – Tremor and Skylight • PMCC Detections: 4/22/07 1510-1525 • Nearly all signal originated from Pu`u O`o • Discovered skylight along lava tube system • Intermittently acoustic active, both audible and infrasound • Possible to discriminate between the two sources • Need confirmation of detection…
KIPU results – 4/22/07 0600-0800 UTC • A sharp, complex spectral peak of ~0.6 Hz is present at KIPU and MENE • Tends to bifurcate and shift frequency over time • Seismic wavefield is complex and path effects appear to be present, but spectral peak is also manifested • Broadband infrasound signal as well Black-2.5 km inf Red-13 km inf Blue-seismic
Episode 56 • EQ swarm begins around 6/17 1415 UTC on upper east rift zone near Mauna Ulu • Magma intrusion 1.5-3 km deep • Puu Oo Vent collapses on 6/18 • First impulsive acoustic events at Puu Oo • Relatively long duration
Fissure Opening? Episode 56 – Fissure Opening • Infrasound signals around 1015-1045 UTC from ~132° • Fissure Opening?
Episode 56 - Resuts • 200 m long fissure opens 6 km west of on 6/19 • Steam, gas, and lava • Infrasound recordings from 0900-1100 UTC pm 6/19 • Focused between 2-10 Hz, doesn’t look like surf or cultural • Not constrained well by any other technology • Harmonics present • Similar to Puu Oo
Fissure Eruption – 7/21/2007 • Magma drained from Pu`u O`o around 1000 UTC • Eruption moved eastward along a set of fissures – strong infrasound! • Seismically quiet, video blocked by poor weather
Fissure Eruption- 8/22/07 • Increased tremor signals in late August • Activation of harmonics around 2000 UTC • Increased activity prompted installation of second array in Glenwood
Fissure Eruption – 7/21/2007: Single array • Fissure signal concentrated between 1.5-5 Hz, similar to Puu Oo • No harmonics
PO FD ZION Results – 2007/9/17 • Tremor signals switch on to Puu Oo on 9/17? • Puu Oo still actively degassing • Standing Hypothesis • Magma interacting with the gas • Many ways of producing this process • Want to model broad spectral peak
Fissure Eruptions Daily Acoustic Energy Calculation • Calculate acoustic source energy for tremor • To differentiate between signal and noise, the acoustic energy is only calculated during times of coherent signal from the specified azimuth. • Split data into 60 sec windows • Mean correlation value for window must be above threshold • Minimum number of pixels per window EAcoustic=2πr2/ρc ∫ΔP(t)2dt r=source-receiver distance ρ=air density C=sound speed ΔP=change in pressure
Acoustic Energy vs. RSAM/Tilt • Event 1: 8/24 00:02 UTC (8/23 14:02 HST) • Collapse Event: • Significant drop-off in acoustic energy (tremor) in the days following this collapse event. • Event preceded by 3 days of high acoustic tremor and RSAM values. • No impulsive acoustic signal recorded at the time of the seismic/tilt event. Minor earthquake recorded by the infrasound sensors around 0000 UTC. • Collapse event may have changed magma/degassing dynamics.
Acoustic Energy vs. RSAM/Tilt • Event 2: 8/29 1100-1400 UTC (8/29 0100-0400 HST) • Tremor Event • Strong acoustic and seismic tremor during this 3 hour time period, originating from P‘u‘u ‘O‘o, not Fissure D. • Infrasound signal concentrated between 1220-1400 UTC, is below 3 Hz, Collapse event may have changed magma/degassing dynamics. • Infrasound signal decreased rapidly at 1400 UTC and returns to similar azimuth as Fissure D. • No longer-term variation in acoustic energy or RSAM apparent. • Similar “burst” of tremor occurred on 9/17 around 0900 UTC.
Acoustic Energy vs. RSAM/Tilt • Event 3: 9/3 0152 UTC (9/2 1552 HST) • Collapse Event: • Nothing out of the ordinary recorded acoustically at the time of this collapse event. • However, tremor energy levels decreased to near zero in the days following. Collapse event may have changed dynamics as well.
Conclusions – Future Work • Identification of new sources (skylight, fissures) made at Kilauea • Tremor observations consistent with pressure oscillations resulting from unsteady magma flow within tubes/conduits • P‘u‘u ‘O‘o tremor between 0.5 – 3 Hz fairly constant in both infrasound and seismic recordings before 6/18 • Higher frequency harmonic, gliding tremor more intermittent and possibly due to excitation of conduit • Sharp spectral peaks in infrasound and seismic – same resonating source? • Comparison of seismic, infrasound, and tilt data help explain dynamics and overall volcanic activity levels