V51H-01
Kilauea's Ongoing Eruption: 25th Year Brings Major Changes
2007 marks the 25th year of nearly continuous eruption on Kilauea's east rift zone. Episodic high lava fountains, which built the Pu`u `O`o cone during the first three years of the eruption, ended in 1986. Activity then migrated downrift and the Kupaianaha shield was formed by passive effusion of lava. The change in eruptive style resulted in a switch at Pu`u `O`o from cone construction to cone collapse that has been ongoing for the last two decades. Activity at Kupaianaha ceased in 1992, and the eruption resumed at Pu`u `O`o. The eruptive style established at Kupaianaha continued, however, with continuous effusion from vents on the southwest flank of the Pu`u `O`o cone. The last 15 years have been characterized by the formation of relatively stable tube systems---broken only by a brief fissure eruption uprift of Pu`u `O`o in 1997---that have carried lava from the flank vents to the ocean about 9 km away. The Prince Kuhio Kalanianaole (PKK) tube, the most recent of these tube systems to develop, was active from March 2004 to June 2007. The PKK flow was emplaced almost entirely on older flows of this eruption and entered the ocean in several locations over a span of 6 km. The "Father's Day" intrusion of June 17--19, 2007, robbed the supply of magma to Pu`u `O`o and, thus, the active flow field. The floor of the Pu`u `O`o crater dropped 80--100 m, the PKK tube system drained, and the active flows and ocean entry quickly stagnated. On June 19, a short-lived fissure eruption broke out low on the east flank of Kane Nui o Hamo, about 6 km uprift of Pu`u `O`o, burying only 0.22 hectares. The eruption at Kilauea paused from June 20 through July 1 or 2, when lava returned to Pu`u `O`o and began refilling the collapsed crater. Near midnight on July 20--21, after at least 19 days of lava lake growth, the lava pond within the Pu`u `O`o crater drained suddenly when a series of fissures opened on the east flank of the cone and propagated ~2 km downrift. The new activity, dubbed the July 21 eruption, initially fed `a`a and pahoehoe flows but, within a few days, developed into a system of perched lava ponds. By the end of July, the perched pond over the easternmost fissure evolved into an open channel feeding a series of `a`a flows heading downrift around the north side of the Kupaianaha shield. The lava supply from other fissure segments declined and stagnated within a few days of the lava channel's formation. As of September 4, 2007, `a`a flows have extended up to ~6 km from the fissure and have covered more than 600 hectares. Unless the eruption supply rate increases, `a`a flows fed by the open channel will likely travel no more than a few kilometers from the vent and pose no threat to those living downslope. If the eruptive style changes, however, to one characterized by tube-fed pahoehoe, then downslope communities could be directly impacted.
V51H-02
Seismic Characterization of the June 17, 2007 East Rift Intrusion at Kilauea Volcano
An early morning earthquake swarm on Sunday, June 17, 2007 signaled the beginning of a sequence of seismic and deformational activity consistent with an intrusion of magma in the upper East Rift of Kilauea volcano, Hawaii. This culminated in an outbreak of lava from a discontinuous 160 meter long fissure, approximately 6 km west of Pu`u `O`o and 13 km southeast of Kilauea's summit. Here we detail the seismic characteristics and observed deformation accompanying this magmatic intrusion and eruption. Seismic activity began at 0216 Hawaiian Standard Time (HST - UTC minus 10 hours) with 38 events greater than magnitude 2 and over 80 located earthquakes in the first two hours. These earthquakes were centered 1.5-2 km southwest of Mauna Ulu (9 km southeast of Kilauea summit) between 1.5 and 3 km deep. At the same time, tiltmeters at Kilauea's summit began indicating rapid deflation of the summit area. Starting at 0730 HST there was a pronounced eastward shift (by about 4 km) of the center of seismicity to a location just west of Makaopuhi crater. At this time GPS stations began to show extension across the rift in that area. Over the next 7-8 hours, the locus of seismicity continued to migrate eastward with over 70 events located at Makaopuhi crater. Over the next day more than 40 earthquakes occurred with locations ranging from the uppermost east rift to Makaopuhi crater. The next afternoon (June 18) there was another pronounced eastward shift in seismicity with 23 earthquakes, occurring between 1459-1600 HST, centered between Makaopuhi crater and Napau crater. Kilauea summit area tremor levels rose throughout the day on June 17, reaching a peak of 24 times background levels mid-day on the June 18, before beginning a slow decline. Tremor levels along the East Rift showed brief periods of strong activity (4-6 times background levels), and lava was discovered at the surface shortly after one such tremor episode at 0625 HST on June 19. Shortly before noon on the same day, Kilauea tiltmeters indicated an end to summit deflation, and GPS stations indicated a slowing of rift extension which had reached nearly 1 meter of across rift extension. This series of punctuated eastward migrations of seismicity (along with the observed deformational signals) provide a detailed record of the space-time evolution of the June 17 magmatic intrusion and eruption. http://hvo.wr.usgs.gov/
V51H-03
Father's Day dike intrusion and eruption reveals interaction between magmatic and tectonic processes at Kilauea Volcano, Hawaii
The June 17-19, 2007, Father's Day dike intrusion and eruption at Kilauea volcano brought to an end a seven- year period of steady state lava effusion at the Pu'u 'O'o vent. The event was observed by an unprecedented number of geophysical instruments, with temporary arrays of GPS and tiltmeters augmenting the continuous monitoring network. Envisat and ALOS SAR scenes were also acquired during this event and provide further information on the surface deformation as the event progressed. Fortuitously, the Envisat acquisition was during a pause in the middle of the sequence, while the ALOS PALSAR scene was acquired at the end of the sequence, allowing us to model each phase separately. Analysis of these data sets indicates that, in addition to three phases of the dike intrusion, a slow earthquake also occurred on the south flank of Kilauea. The slow earthquake apparently began near the end of the second phase of the dike intrusion. It was still underway the following day, when the third phase of the intrusion began and culminated in a small eruption. This suggests the possibility that the slow earthquake was triggered by the initial diking, and then in turn influenced the progression of the intrusion. Two of the largest previous slow earthquakes also hint at a connection between slow earthquakes and eruptive activity on Kilauea. The range of observations of the Father's Day events provides us with a unique opportunity to investigate the complex interactions between the tectonic processes of the south flank and magmatic processes within the summit and rift zones.
V51H-04
An Episode 56 Perspective on Post-2001 Comagmatic Mixing Along Kilauea's East Rift Zone
A significant change in the petrology of Pu`u `O`o -Kupaianaha lava occurred in April 2001 (3 years after the onset of the decade-long episode 55). Prior to that time all steady-state eruption products were olivine phryic. After that time and until the Kane Nui o Hamo eruption of June 19, 2007 (episode 56), all magma erupted from vents in and around Pu`u `O`o was olivine and pyroxene-phyric containing <1mm, isolated or clustered clinopyroxene (±olivine, ±plagioclase), usually with resorbed edges. Textures, phase chemistry and low-pressure phase relations define a pre-eruptive mixing environment that is driven by continuous recharge of a stagnant, near-cotectic shallow magma body. The comagmatic nature of the cooler component in the post-2001 hybrid magma is verified by low concentrations of incompatible elements relative to MgO. Since the eruption began in 1983, the olivine-saturated liquid-line-of-descent has progressively shifted toward the present-day low concentrations of incompatible elements. Superimposed on this long-term trend are shorter chemical cycles (months to years) which track fractionation and recharge between comagmatic endmembers of ~10 and ~7 wt% MgO. These shorter cycles correspond to heating and cooling events and imply magmatic recharge of a sustained shallow magma reservoir within the eruptive plumbing system. The longest cooling cycle of the entire eruption began in April 1998 after effusion of the hottest and most- primitive lava erupted since 1985 (episodes 30 and 31). Glass temperatures up to 1168°C and bulk MgO of 9.5 wt% steadily declined for 6 years until late 2004 when they bottomed-out at 1140°C and 6.8 wt%. This signaled a stable near-cotectic magma condition. MgO contents and glass temperatures stailized at ~7.1 wt% and 1146°C for the remainder of episode 55, as the hybrid magmas were stirred by a steady influx of summit-derived magma beneath a complex of intermittently active Pu`u O`o vents. During the June 19, 2007 Kane Nui O Hamo eruption (episode 56), as with the January 1997 Napua Crater event (episode 54), the summit deflated and Pu`u O`o collapsed as magma was drawn from either end of the active rift conduit toward a zone of extension. In both cases, magma returned to the Pu`u `O`o vent area after the conduit repressurized. However, in contrast to cool and porphyritic hybrid magma erupted through isolated and chemically evolved rift magma reservoirs at Napau Crater, the episode 56 lava is relatively primitive (8.7 wt% MgO) and 30 to 50°C hotter at 1160°C. This is likely to be summit-derived magma from within the active rift conduit beneath Kane Nui o Hamo. The episode 56 lava is ~15°C hotter than the late episode 55 hybrid magmas with consistently low incompatible elements and likely represents the recharge component that maintained a shallow reservoir at near-cotectic conditions beneath the vicinity of the Pu`u `O`o vents for the last several years. Both lava erupted from Pu`u `O`o in early June, 2007(episode 57), and lava the from the July 21-24 sequence of fissure eruptions down-rift of Pu`u `O`o (early episode 58) contain a distinctly hybrid phenocryst and glomerocryst assemblage, suggesting a flushing of cooler crystal-laden magma from the conduit.
V51H-05
The Birth and Growth of Kupaianaha Lava Shield, Kilauea Volcano: 1986-1992
Kupaianaha began to form on July 20, 1986, 3 km northeast of Pu`u `O`o, which had been the focus of Kilauea's east-rift-zone eruption for the prior 3.5 years. On July 18, Pu`u `O`o was primed for the 48th episode of high fountaining. Instead, fissures erupted first uprift and then downrift of the cone. This activity, which lasted until mid- morning on July 19, was preceded by an earthquake swarm and accompanied by 17.4 Ýradians of deflation at Kilauea's summit. On July 20, another small swarm of earthquakes heralded the eruption of the 200-m-long Kupaianaha fissure. Lava flows spread rapidly from the new fissure, advancing about 800 m southeastward during the first 2 days. The nascent shield was 4 m high by July 25, and a lava pond was forming over the vents. On July 26, a major breakout fed a channelized flow with an `a`a terminus that traveled 4.6 km southeast before stagnating on August 3. The upper end of the channel remained active on the shield after August 3 and evolved into the pond neck and the upper section of master tube that would direct most of the lava to the southeast during the next five years. The Kupaianaha shield attained a height of 33 m during August due to pond overflows, and expanded to cover an area of 1 x 1.6 km. By early October 1986, the lava pond had acquired its final shape and the shield was over 40 m high. Growth of the shield via intrusions also began in August and continued throughout the first year. Outpourings of intruded lava built satellitic shields, and extrusions of `a`a emanated from upwarped regions on the flanks of the shield. Intrusions were volumetrically less important than pond overflows, but they had a significant effect on the final shield morphology. The Kupaianaha shield reached a final height of 60 m early in July 1987, when a blockage of the master tube caused the pond to overflow in all directions for the last time. Two days later, the master tube broke open on the east side of the shield, building a satellitic shield nearly as high as the main shield in just 2 days. Lava flows from this shield constructed a new tube system to the southeast. On July 29, the new tube became blocked and lava overflowed from the summits of both the satellitic and main shields. The increased pressure reopened the connection to the original master tube buried within Kupaianaha. During this same period, the large (500 x 200 m) laccolith complex and `a`a flow field that formed on the north side of the shield in the spring of 1987 remained quiet, but a new domal laccolith (150 x 100 m) grew 15 m high on the south side of the shield. Repeated extrusions from this structure in early July built an apron of `a`a that extended 0.5 km to the base of the shield. On July 27, a 1.5-km-long `a`a flow erupted from the north laccolith, and four days later it subsided 3-4 m. This sequence of events ended the growth of Kupaianaha shield. Beginning in September 1986, well before shield-building activity diminished, tube-fed lava flows had been progressing slowly away from the shield. During the first year, flow activity alternated between the shield and the advancing flow field, as immature lava tubes formed and failed. By the end of 1987, most of the flow activity was located on the coastal plain, terminating at ocean entries 10-12 km from the vent. This was the status quo for the remainder of Kupaianaha era. The end of Kupaianaha came slowly. The pond remained unchanged until early 1990, when repeated pauses in the eruption caused the pond to crust over. Through 1991, the lava output diminished, and, in early February 1992, Kupaianaha stopped erupting. Within 10 days, the ongoing eruption returned to Pu`u `O`o.
V51H-06 INVITED
Crater Floor and Lava Lake Dynamics Measured with T-LIDAR at Pu`u`O`o Crater, Hawai`i
We used a near-infrared (1.2 micron wavelength) tripod-based scanning LiDAR system (T-LIDAR) to capture crater floor and lava lake dynamics in unprecedented detail at P`u`u `O`o crater on Kilauea volcano, Hawai`i. In the ~40 days following the June 17-19 intrusion/eruption, Pu`u `O`o crater experienced substantial deformation comprising 2 collapse events bracketing rapid filling of the crater by a lava lake. We surveyed the crater floor with centimeter-scale spot-spacings from 3 different vantage points on July 13 and from one vantage point on July 24. Data return was excellent despite heavy fume on July 24 that obscured nearly all of the crater features, including the walls and floor. We formed displacement fields by aligning identical features from different acquisition times in zones on the relatively stable crater walls. From July 13, over a period of several hours, we imaged ~2 m of differential lava lake surface topography from the upwelling (eastern) to downstream (western) portion of the flowing lava lake. From July 13 to July 24, the lava lake level dropped by as much as 20 meters in a zone confined by flanking levees. Our results confirm the utility of T-LiDAR as a new tool for detailed volcano geodesy studies and suggest potential applications in volcano hazards monitoring.
V51H-07
Monitoring Changing Eruption Styles of Kilauea Volcano Over the Summer of 2007 With Spaceborne Infrared Data
On June 19, 2007 episode 56 (the Father's Day intrusion) of the ongoing eruption at Kilauea Volcano culminated with a small eruption of lava from a 250 m long fissure approximately 6 km west of Pu'u 'O'o. The event was preceded by an earthquake swarm and attributed to the intrusion of magma. This intrusion was also associated with cessation of activity at Pu'u 'O'o and deflation of its summit region. On July 21, 2007 new lava then erupted along a set of fissures that extended eastward from Pu'u 'O'o toward the old Kupaianaha vent. By early September, this eruption continued to supply a lava channel approximately 1 km long, which has fed two 'a'a flow lobes advancing to the northeast and southeast. We describe the application of spaceborne imaging data from the visible to the thermal infrared (TIR) wavelengths for monitoring activity throughout this period. Satellite thermal infrared (TIR) data with low spatial resolution (i.e., kms/pixel) have been used for years to monitor changes in surface thermal features such as volcanic flows. However, the use of higher spatial resolution data allows for the extraction of physical parameters at meter to sub-meter scales. The Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) provides TIR, shortwave infrared (SWIR), and visible-near infrared (VNIR) data ideal for this type of analysis, hazard assessment, and smaller-scale monitoring of active lava flows. From June-August of 2007, ASTER was scheduled 23 times and collected 11 independent scenes of the new flow activity at Kilauea. Of these, 7 were clear to partly-cloudy and show excellent coverage of the activity following the Father's Day intrusion. TIR and SWIR data, converted to atmospherically corrected emitted surface radiance, have been used to extract flow extent, areal coverage, flow advance rate, and maximum brightness temperature. These data correlate well with descriptions of the flow activity documented by Hawaiian Volcano Observatory field crews. For example, the ASTER night time image collected on July 19 (22:42:56 HST) had a maximum SWIR-derived temperature of 305 C, and a total thermally-elevated area of 0.19 sq. km. Within that region, 3 distinctly hotter zones were identified as most likely the West Gap pit craters, which were described as intermittently overflowing to form a small lava lake at the time. Following the July 21 fissure eruption, ASTER observations were augmented with non-standard approaches such as collecting visible night time data in order to accurately extract the higher temperature of the open lava channel. Although clouds partially obscure the August 30 night image, a maximum pixel-integrated temperature of 750 C was detected using the VNIR night- time data for the first time. Such a monitoring program coordinated between NASA and a USGS volcano observatory can provide important data on hot spot detection, eruption rate, and flow advance at times where it may be too costly or risky to send scientists into the field.
V51H-08 INVITED
Setting of the Father's Day Eruption at Kilauea
The Father's Day eruption and associated intrusion took place within a 10-km segment of Kilauea's east rift zone between Hi`iaka and Napau Craters--a segment that has had more numerous eruptions and intrusions than any other of comparable length during the past 200, probably the past 1000, years. Fifteen known eruptions started in this area in the past 200 years: 1840, 1922, 1923, 1962, August and October 1963, March and December 1965, August and October 1968, February and May 1969, May and November 1973, and March 1980 (only 3 cubic meters!). Three others, not previously designated as distinct eruptions despite having all the appropriate characteristics, took place during on-going eruptions: two in `Alo`i Crater in 1970 and 1972, and one in Napau Crater in 1997. Two of the largest shields on the east rift zone formed during long-lasting eruptions within this areaKane Nui o Hamo at an unknown date, perhaps the 11-12th century, and Mauna Ulu (1969-1974). In addition, many small intrusions without eruptions are known. Seven short eruptions punctuated a prolonged eruption: four within the segment during the Mauna Ulu eruption, two at the summit and southwest rift zone during that same eruption, and one in Napau Crater in 1997 during the Pu`u `O`o eruption. Thus the Father's Day eruption is not unique by virtue of taking place during an ongoing eruption elsewhere along the rift zone. The increased frequency of activity in the segment during the 20th century is obvious, particularly after 1962. For most of the past 1,000 years, eruptions were centered at Kilauea's summit, with significant but lesser activity along the rift zones. A large summit deflation in 1924 ended the nearly continuous lava lake in Halemaumau, eventually leading to the past 5 decades of dominantly east rift zone activity. This segment of the rift zone contains most of the pit craters on Kilauea and gradually changes from a SE trend near the caldera to an ENE trend that characterizes the rest of the zone. The Koa`e fault system joins the east rift zone at the curve. The complex structural setting likely affects the frequency of magmatic activity in the segment. All of the eruptive and intrusive activity results in storage of isolated magma bodies. Not surprisingly, petrologists find evidence that summit magma mixes with stored, fractionated magma. The area near Makaopuhi Crater and Kane Nui o Hamo is a particular focus, inferred since the mid-1960s to harbor a shallow magma reservoir. All of the eruptions and intrusions are accompanied by sharp deflation and shallow seismicity at the summit and shallow seismicity and uplift along the intrusion or eruptive fissures. Most often, no seismicity occurs between the summit and the area of intrusion or eruption. Within that area, seismicity commonly migrates downrift but occasionally uprift. Similarly, crack opening generally progresses downrift, with a few exceptions. Cracks generally trend about 65 degrees and can be either left- or right-stepping. Cracks open along azimuths of 155-175, with local exceptions. There is no structural difference between eruptive cracks (fissures) and non- eruptive cracks. Single eruptive fissures rarely exceed 200 m in length, instead stepping within en echelon zones above a presumably linear dike. Since the late 1960s, widening has been measured across the active area during eruptions and intrusions, first by EDM and then by satellite-based systems. The opening is nearly symmetrical within the rift zone, but farther out the north flank barely responds whereas the south flank moves seaward. Available leveling data show uplift on either side of the dike and subsidence along the crest. Examples of deformation in the 1960s and 1970s will be presented, and bibliographic references to past activity will be available.