Report on Kilauea (United States) — March 2018
Bulletin of the Global Volcanism Network, vol. 43, no. 3 (March 2018)
Managing Editor: Edward Venzke. Report research and preparation by: Liz Crafford.
Kilauea (United States) Activity continues at Halema'uma'u lava lake, and at the East Rift Zone 61g flow, July-December 2017
Please cite this report as:
Global Volcanism Program, 2018. Report on Kilauea (United States). In: Venzke, E (ed.), Bulletin of the Global Volcanism Network, 43:3. Smithsonian Institution.
19.421°N, 155.287°W; summit elev. 1222 m
All times are local (unless otherwise noted)
Hawaii's Kilauea volcano continued its eruptive activity, intermittent for thousands of years and continuous since 1983, throughout 2017. The summit caldera formed about 500 years ago, and the East Rift Zone (ERZ) has been active for much longer. Lava lakes were intermittent in and around Halema'uma'u crater at the summit until 1982. Lava has been continuously flowing from points along the ERZ since 1983, and the episode 61g flow was still vigorous through the end of 2017. A large explosion within Halema'uma'u Crater in March 2008 resulted in a new vent with a lava lake that has been continuously active through 2017.
The US Geological Survey's (USGS) Hawaii Volcano Observatory (HVO) has been monitoring and researching the volcano for over a century, since 1912. Quarterly Kilauea reports for July-December 2017, written by HVO scientists Carolyn Parcheta and Lil DeSmither, form the basis of this report. MODVOLC, MIROVA, and NASA Goddard Space Flight Center (GSFC) provided additional satellite information about thermal anomalies and SO2 plumes.
The lava lake inside the Overlook vent at Halema'uma'u Crater continued to rise and fall during the second half of 2017 with no significant lake level changes and a few periods of spattering. The lake level overall was lower at the end of the year than during much of the year, reflecting long-term deflation of the summit. There were no major explosive events from rockfalls, but smaller sloughs of veneer (thin layers of recently cooled lava that adhere to the vent walls) without accompanying explosions were common. Ongoing subsidence at Pu'u 'O'o, especially around the West Pit prompted moves of monitoring equipment, but little else changed at the cone.
The episode 61g lava flow continued with numerous surface breakouts from areas near the vent all the way down over the pali and into the ocean at the Kamokuna delta during July-December 2017. Changes in the subsurface flow in lava tubes contributed to changing locations of surface breakouts, which were still active at the end of the year. The lava flowing into the ocean at Kamokuna slowed and finally ended in November with changes occurring on the delta in the final weeks of its activity.
Activity at Halema'uma'u. For the second half of 2017, activity at the lava lake inside the Overlook crater continued with little change from January-June. The lake's surface circulation pattern was typical, with upwelling in the N and subsidence of the crust along the southern lake margin, but also around the entire edge of the lake depending on the upwelling location (figure 292). There were often "sinks" a few tens of meters from the SW edge of the lake where the crust folds in on itself and sinks, pulling material away from the wall. A noticeable lava veneer buildup often occurred on the southern margin, where the surface crust was most consistently subducting. Short-term spattering events lasted minutes to hours and occasionally altered the surface crust motion by creating localized subsidence. Throughout the period, spattering was often confined to a grotto at the SE sink. On most days, two or more spattering sites were active simultaneously.
The lava lake level generally rose and fell over periods of hours to days in response to gas-piston action and to inferred changes in summit lava pressure indicated by deflation-inflation (DI) events. There were a few periods with exceptions when the lake level remained constant for many days at a time, heating up the surrounding walls enough to produce thermal cracking and popping sounds. The total range of the lake level varied between 35 and 40 m during July-December 2017, with the highest level about 17 m below the rim in early September (elevation 1,020 m), and the lowest levels, about 57 m below the rim in late July and September (elevation 977 m) (figure 293).
There were no significant explosive events triggered by rockfalls, but smaller collapses of veneer and the wall were common, particularly during deflationary phases when the lake level was low and exposed larger areas of the walls. A few larger collapses in September 2017 were big enough to change the geometry of the lake slightly (figure 294). The first, on 8 September at 1806 HST, was a collapse of the large ledge attached to the wall in the southern corner of the lake. This event produced a plume containing ash, a composite seismic event, and lake surface agitation. The following day, 9 September, there was another collapse at 0509. This involved an area of the E Overlook rim composed of mainly lithic deposits, directly above the Southeast sink, which produced a dusty plume, a composite seismic event, and lake surface agitation. On 12 September a thin slice of the southwest lake rim collapsed at 1420, producing a dusty plume, an agitated lake surface for about 10 minutes, and a composite seismic event.
An interesting effect observed on two veneer collapses occurred on 24 October 2017 at 1617 and 1623. Both were silent events but were noticed because they visually depressed the lake as they fell in and sent a small "wave" propagating outward before spattering began a few seconds later. The wave did not make it more than half way across the lake in either case, and both spattering events lasted only a few minutes. Several veneer ledges built up and subsequently collapsed around the lakes perimeter but were most notable on the SW corner of the lake. Three collapses, on 5 December at 0400 and 7 December at 1856 and 2024, enlarged the NNE edge of the lake towards true N, but did not produce a spatter deposit or explosion (figure 295). Another rockfall occurred on the N margin of the lake on 23 December 2017 at 1552 and triggered a large spattering event.
Activity at Pu'u 'O'o. During July-December 2017, there were only minor changes in the main crater of Pu'u 'O'o as recorded by the PO webcam, PT webcam, and the West Pit time-lapse camera. Due to slight subsidence, altered ground, and widening cracks first noted in August, the West Pit time-lapse camera was relocated 20 m to the SE on 12 October, and roughly 25 m further back from the rim on 1 November after new crack expansion was observed.
During the month of August 2017 there was slight subsidence of the W portion of the crater floor, and around 20 August a crack opened up in the S embayment with three heat locations. There appeared to be slight subsidence of the E side of West Pit from the time-lapse imagery spanning 22 November to 12 December. This subsidence accelerated during 15-17 December, but then was slower through the end of the year. The deformation data confirmed subsidence at Pu'u 'O'o, but it seemed to be confined to the land bridge separating the main crater and the West Pit lava pond. The lava pond inside of the west pit rose slightly during the period from around an elevation of 847 m in early August to 849.5 m on 12 December when measured during site visits about every three weeks. A thick surface crust and sluggish plate motion was typical at the lava pond.
The time-lapse camera located on the E rim of the lava pond (through October) captured three rockfalls in July and two in August that disturbed the pond's surface. On 30 September 2017 a collapse of the west pit's SE rim also broke off a portion of the ledge below, as it was impacted by the falling rocks (figure 296). The collapse was large enough to agitate the pond surface for several tens of minutes, and produced a small step in the tilt at the POC tiltmeter.
The pond surface was also disturbed from rockfalls on 22, 28, and 31 October 2017. The first two events were on the N side of the West Pit rim, and the events on 31 October were on the S side of the rim. A small rockfall that triggered minor spattering was witnessed during an overflight on 1 November (figure 297). After 1 November, when the camera was moved away from the rim, it no longer had direct views of the pond. One of the E spillway spatter cones collapsed into the lava tube that was feeding the 61g flow on 20 November and provided a skylight into the tube for a day before it crusted over. On 12 December, a large talus pile on the NNE side of West Pit was evidence of rock falls near the original time-lapse camera site. The talus, likely resulting from several rock falls, piled up onto the lava coated bench.
Activity at the East Rift Zone, episode 61g flow field. The 13 June 2017 breakout that had started on the upper flow field, approximately 1.1 km from the vent, was the largest area of active surface flows on the 61g flow during July-September. Ranging between 2.6–5.8 km from the vent, the breakout significantly expanded the upper flow fields western flow margin. This breakout remained active through the end of September (figure 298). On 26 June 2017 a breakout started near the top of Royal Gardens and quickly advanced down the pali, east of the main flow field. By 6 July the front of the breakout had extended 500 m beyond the pali base with fluid pahoehoe at the front, and a small a'a channel on the steep part of the pali. Slow advancement of the flow placed it approximately 1.5 km from the emergency road near the coast by 9 August before the flow front stalled. When mapped again on 15 August, the closest active flows were about 2.1 km uphill from the road. Intermittently during 1-20 September the breakout produced channelized flows on the steep part of the pali, sometimes as often as every 24 hours. By the end of September active surface flows had advanced to approximately 1.6 km from the emergency road (figure 298).
Two other breakouts that started near the episode 61g vent were also active during July-September 2017. The 5 March breakout, which had advanced downslope during its 4 months of activity, was weakly active on 10 July, with two small lava pads observed approximately 4.8 km from the vent. By the time of the overflight on 9 August, the breakout was inactive. On 26 July around 1025 HST, a new breakout started about 1.1 km from the vent and remained active through the end of September with flow activity located 1.1-2.5 km from the vent. On 27 August at roughly 0945 a breakout began on the steep part of the pali originating from the main 61g tube. By 1 September the breakout was at the base of the pali and spreading onto the coastal plain. A few other channels were reported on this area of the pali, and activity continued through the end of September with very little advancement across the coastal plain (figure 299).
The 26 June 2017 breakout remained active and stable through the end of 2017, forming a tube from its breakout point to midway down the pali on the E side of the 61g flow. The area where breakouts from 5 March, 13 June, and 26 July occurred (1.1 km from vent) also remained intermittently active through the end of 2017 (figure 300).
Numerous overflows originating on the sea cliff began in early October 2017. These breakouts occurred within 310 m of the sea cliff and persisted for nearly a month. There were also approximately 20 short-lived breakouts in October above the sea cliff, each lasting 1-3 days. They were located mostly in clusters on the upper flow field at 1, 2, and 3.5 km from the vent, along the top and base of the pali, and from the coastal tube.
An estimated 35 tube breakouts occurred during November 2017; they typically lasted 2- 10 days, and were located inland of the October breakouts. Locations of activity were in the upper flow field almost entirely between 2 and 3.5 km from vent, with three closer breakouts at 0.5, 0.8, and 1 km from vent. The two active tubes on the pali continued to have breakouts at the top and base of the cliff, but also started breakouts midway downslope (figure 301). At 0805 on 7 November, a viscous breakout occurred approximately 500 m above the sea cliff. The small breakout came directly from the 61g tube and lasted for roughly four and a half days. Another viscous breakout from the tube occurred approximately 950 m upslope of the sea cliff from 18-23 November. A week after that, a third viscous breakout occurred about 2 km from the sea cliff. By the end of November, there was no further breakout activity on the delta or the distal half of the coastal plain.
During December 2017, an estimated 30 breakouts were recorded from the 61g flow tube, however these were often longer, lasting up to a week on the upper flow field, and with near perpetual breakouts on the pali throughout the month, which made quantifying the exact number difficult. A new breakout occurred 500 m from the 61g vent on 1 December and lasted through 20 December. This breakout, and the whole area between 500-1,200 m from the vent, poured lava onto the eastern upper flow field (figure 300). Most of the upper flow field activity was focused very close to the vent, between 350-800 m; additional activity also occurred at the 1 km location and a few continued breakouts were noted from the 2-3.5 km region. The coastal flow field activity was sluggish and mostly a result of the near-constant pali tube breakouts reaching the base. On 9 December a new voluminous breakout began near the top of the pali that burned through the kipuka near the center of the flow field (figures 302 and 303). This major breakout lasted through the end of the year and produced mostly 'a'a channels on the pali with pahoehoe at the pali base. Pali tube breakouts occurred at nearly every elevation but seemed to move higher up the slope as the month came to a close. Activity did not advance more than 400 m from the base of the pali.
Time series thermal maps of the 61g flow field overlaid on all of the tubes mapped from the field to date suggested to HVO scientists that some of the many breakouts during October-December 2017 may have come from reactivation of an earlier tube thought to be inactive since at least April 2017 (figure 304). Breakout locations coincided with the former tube trace, and happened at least five times between 21 September and 5 January 2018.
Activity at the East Rift Zone, Kamokuna ocean entry. By the end of June 2017, flows from multiple breakouts had resurfaced the delta of the Kamokuna ocean entry, covering earlier cracks, and building up and steepening the delta's landward side. These surface breakouts continued into early July, but by 10 July several new cracks had appeared, two of which visibly spanned the width of the delta (figure 305). Slumping of the seaward half of the delta and expansion of the cracks was visible in time-lapse camera images until the end of September.
On 19 August 2017 around 0405 HST a breakout started on the sea cliff approximately 100 m upslope of the ramp, and five minutes later lava was spilling over the sea cliff and onto the delta. The breakout point and the lava falls over the cliff were both on the W side of the 61g tube. The lava produced a small 'a'a flow on the delta (figure 305), during its short-lived activity that lasted roughly 9.5 hours. Late on 19 August, the time-lapse camera also captured two images of littoral explosions in the center of the delta that produced a large spatter deposit on the delta's surface.
Three more sea cliff breakouts started on 23 September 2017. The first was brief "firehose-like" activity that began in the early morning hours. Based on the delta surface flows it produced, activity lasted less than 24 hours. Later views of the cliff face revealed that the "firehose" came out of a narrow horizontal crack E of the ramp, that was less than a meter below the top of the cliff. Later that day, on the sea cliff near the ocean entry, two new breakouts started, one to the E and one to the W of the tube. The E breakout originated roughly 70 m upslope of the sea cliff, and the breakout point had been fractured and depressed. Its thin pahoehoe flow spread out behind the littoral cone and came close to the edge of the cliff but did not spill over. The W breakout was visible in the time-lapse camera images on 23 September from around noon until midnight, producing only a few small dribbles of lava over the sea cliff. The breakout point was roughly 100 m upslope of the sea cliff, and buried the breakout from 19 August with thick, viscous pahoehoe. By the end of September, surface flows again covered much of the delta until most of the cracks were obscured, and only the ramp and a small area of the eastern delta close to the sea cliff were still uncovered.
Beginning in late August 2017, the ocean entry plume started to fluctuate regularly, and the plume was often weak or would briefly shut down. A shatter ring (a raised rim depression that forms over active lava tubes) began forming near the front of the delta on 21 August. By 30 August, the repeated uplifting and subsidence of the delta had broken the surface flows and built up a large rubble pile. On 26 September 2017 a bulge formed on the back half of the delta where the slope was steepest (figure 306). This inflationary feature produced steam and a delta surface flow from a crack at its base.
HVO scientists concluded that the bulge observed on 26 September 2017 was the result of the formation of a spreading-induced graben in the middle of the delta that obstructed the 61g tube between 23 and 26 September 2017 (figure 307, top row). During the first part of October, additional breakouts from the tube above the sea cliff produced lava falls that poured down on the W side of the tube (figure 307, middle row). A few breakouts in the latter half of October flowed to the E side of the tube (figure 307, bottom row). The delta did not expand much in area during October-December 2017, but it thickened greatly due to the added volume from the lava falls breakouts and several small sluggish breakouts on the delta. The maximum extent that the delta reached was a little over 4 hectares in October, and then it began to shrink from waves crumbling its edges. By the end of December, the delta had lost about 0.4 hectares (1 acre) of land.
The ocean entry was thought to have fully ceased activity shortly after 12 November 2017. The plume had its first pause in activity on 23 September, and quickly resumed but with decreasing vigor. By 26 September the plume was noticeably weaker and beginning to show intermittent pauses, which continued and became more prolonged through 4 November. The following day (5 November) was the first day with no plume visible in the HPcam, and 6 November was the last day an ocean entry plume was visible in the HP webcam. Ocean entry was active and observed during field visits between 6-11 November, but its weak, diffuse plume was not visible to the HP camera. The time-lapse camera stopped taking photos during the end of the Kamokuna delta activity in the late afternoon on 11 November (figure 308). This malfunction was discovered during a field visit on 12 November; the batteries were replaced a week later. The last photo of known lava activity on the delta was taken on 12 November, and the delta was likely completely inactive within a day or two.
During a 12 December 2017 overflight, an HVO scientist witnessed a collapse of a small portion of the sea cliff east of the tube into a yellow talus pile on the back portion of the delta, removing the evidence of the lava falls.
Satellite thermal and SO2 data. In addition to field observations, satellite-based thermal and SO2 data provide important insights into the ongoing activity at Kilauea. The many MODVOLC thermal alerts issued during July-December 2017 show the varying intensity and locations through time of the many breakouts along the episode 61g flow field from near the vent at the base of Pu'u 'O'o all the way down to the Kamokuna ocean entry delta (figure 309).
The MIROVA project thermal anomaly graph of distance from the summit also shows the multiple sources of heat at Kilauea and the migration of those sources over time (figure 310). The MIROVA center point for relative distances described here is about 10 km (0.1°) E of Halema'uma'u crater. The anomaly locations at about 10 km distance from this point correspond to both the lava pond at Pu'u 'O'o crater and the Halema'uma'u crater lava lake. Those about 20 km away correspond to the Kamokuna ocean entry. Anomalies that migrate over time between 10 and 20 km distance trace the movement of the many episode 61g flow breakouts between Pu'u 'O'o and the Kamokuna ocean entry during July-December 2017.
Kilauea emits significant SO2 that is recorded by both ground-based and satellite instruments. Sulfur dioxide emissions exceeded density levels of two Dobson Units (DU) multiple times every month during the period (figure 311). Increases in SO2 flux are caused by many factors including increases in the number and size of surface lava breakouts as well as activity at the summit crater.
Geologic Background. Kilauea, which overlaps the E flank of the massive Mauna Loa shield volcano, has been Hawaii's most active volcano during historical time. Eruptions are prominent in Polynesian legends; written documentation extending back to only 1820 records frequent summit and flank lava flow eruptions that were interspersed with periods of long-term lava lake activity that lasted until 1924 at Halemaumau crater, within the summit caldera. The 3 x 5 km caldera was formed in several stages about 1500 years ago and during the 18th century; eruptions have also originated from the lengthy East and SW rift zones, which extend to the sea on both sides of the volcano. About 90% of the surface of the basaltic shield volcano is formed of lava flows less than about 1100 years old; 70% of the volcano's surface is younger than 600 years. A long-term eruption from the East rift zone that began in 1983 has produced lava flows covering more than 100 km2, destroying nearly 200 houses and adding new coastline to the island.
Information Contacts: Hawaiian Volcano Observatory (HVO), U.S. Geological Survey, PO Box 51, Hawai'i National Park, HI 96718, USA (URL: http://hvo.wr.usgs.gov/); MIROVA (Middle InfraRed Observation of Volcanic Activity), a collaborative project between the Universities of Turin and Florence (Italy) supported by the Centre for Volcanic Risk of the Italian Civil Protection Department (URL: http://www.mirovaweb.it/); Hawai'i Institute of Geophysics and Planetology (HIGP) - MODVOLC Thermal Alerts System, School of Ocean and Earth Science and Technology (SOEST), Univ. of Hawai'i, 2525 Correa Road, Honolulu, HI 96822, USA (URL: http://modis.higp.hawaii.edu/); NASA Goddard Space Flight Center (NASA/GSFC), Global Sulfur Dioxide Monitoring Page, Atmospheric Chemistry and Dynamics Laboratory, 8800 Greenbelt Road, Goddard, Maryland, USA (URL: https://so2.gsfc.nasa.gov/).