GEOL205 - Kilauea Eruptive History

Slide Show

Lecture Notes

General Characteristics of Kilauea Eruptions

The image on the right is a computer generated digital elevation map of Kilauea and Mauna Loa volcanoes. Mokoweoweo can be seen on the left side of the image. The summit crater of Kilauea is located just below the center, and the East Rift Zone occupies the lower left quarter of the image. A red line shows the extent of fissuring occuring during the first few weeks of the Pu'u o'o eruptions (beginning January 3, 1983).

Most of the major flows on Kilauea Volcano have been roughly dated using Carbon 14. Flows classified by age are shown on the map on the left. Red flows (date certain) were erupted this century. Note that they tend to be predominantly associated with both the East and Southwest Rift Zones. 19th century eruptions are shown in orange. Only two, 1823 along the Southwest rift, and the 1840 eruption on the East rift zone. Both of these eruptions were rather unusual in ways that will be discussed below. Eruptions during the 18th century are shown in yellow. What is immediately obvious here is that there seem to be considerably more activity in the 18th century than in the subsequent 19th. Why is this so? The answer turns out to be rather straightforward. The activity along Kilaueas East Rift Zone preceded the major summit explosion and caldera formation event of 1790. Subsequently during the 19th century activity was concentrated at the summit, refilling the deep pit following the summit collapse late in the preceding century. Prehistorical flows are shown in olive (250-350 yrs B.P.), dark green (350-500 yr B.P.), dark blue (750-1000 yrs B.P.), light blue (1000-1500 yrs B.P.). Note that the youngest flows North of the East Rift are all in excess of 250 yrs B.P.

The map on the right shows the various types of lava flows found on Kilauea Volcano. The light red areas are extensive tephra units, generated primarily during fountaining events, either high fountaining or the lower fountains associated with a "curtain of fire". The dark units are a'a flows. As can be seen, they occur both along the rift zones as well as distally, primarily associated with flows descending the pali on the South Flank. There would be more visible along the rift, except that they have been covered by subsequent surface-fed pahoehoe. The remaining units are pahoehoe, either surface fed (dark grey), or tube fed (light grey). Note that surface fed pahoehoe is primarily proximal while tube fed is distal (distant from the point of effusion).

The map on the left classifies surface flows based on the distance of the vent from the summit of Kilauea. Note that flows mantling the South Flank are associated with rift zone events immediately upslope. North of the rift zone, however, things are totally different. The lightest grey lavas were erupted from vents less than 3 km from the summit. As shown here, the flows between the East Rift and Mauna Loa were erupted from a shield near Kilauea Iki, and much of this lava flowed through Thurston Lava tube. Rift zone derived lavas have historically not invaded the region north of the rift zone.

The map on the right shows yet another way to "cut the lava". Here the red zones are material erupted during explosive eruptions, in all cases shown here related to interraction between magma and water. The products near the summit were produced during the untravulcanian eruptions in 1790 and 1924. The smaller areas near Point Kumukahi were related to the interraction of magma and the water table. The dark areas are brief eruptions, those that satisfy the Epp eruption. Generally these tend to be more distal, however, amoung this group those in this class that are more distant from the summit do in most cases last longer and involve more material than those closer to the summit. The intermediate grey areas show unstable sustained eruptions, such as the one that has been ongoing with many fluctuation since early 1983. The lighter grey areas show the areas covered by stable, sustained eruptions. Again, the area of Kilauea north of the rift zone were derived from a vent near the current Kilauea Iki at the summit of Kiluea. This lava flowed stably from a summit shield for periods that may have lasted for several centuries. As we shall see below, there is some evidence of a kind of a cycle between sustained summit activity and briefer periods time when activity concentrates along the rift zones.

Details of Historical Eruptions

This is a blowup of the lava flows erupted from Kilauea during the past 250 years or so. The oldest flow shown here is the own labelled 1750. R. Holcomb has labelled this eruption as sustained-unstable which puts it in the same category as the eruption that built Mauna Ulu between 1969 and 1974 and the eruption at Pu'u o'o from 1983 through the present. In Holcomb's model there seems to be some relationship between sustained-unstable eruptions along the rift zones and impending summit collapse, although the evidence is still far too scanty to prognosticate such matters with much confidence. The next oldest eruption shown here is the 1823 eruption near the end of the Southwest Rift Zone near Pahala. Recall that during much of the time from the earliest observations and presumably going back to shortly after the 1790 phreatomagmatic summit eruption, Kilueua's summit caldera has been the home of an active lava lake.

This situation lasted shortly before a second phreatic eruption in 1924. However, in 1823 lava suddenly welled passively from the Great Crack near the distal end of the Southwest Rift Zone with little gas or fountaining activity. A pecular kind of liquid bomb, known locally as cow patty bombs, was produced by the efuption. These appear to be blobs of slightly more viscous lava wrapped in more liquid melt which deform to a characteristic shape after they land. The lava flowed rapidly and stealthily to the coast apparently engulging an Hawaiian fishing village with little or know warning. Several people were reported to have been killed as the lava swept through the village. At the summit the lava lake in the summit caldera dropped markedly leaving a black rim of previously solidified lava around the margins. Apparently degassed lava drained through the Northern strand of the Southwest Rift Zone at shallow depth, which accounted for the unusual nature of the erupted lava. The next most significant eruption occured in 1840 beginning near the summit but then ripping down the entire East Rift Zone before becoming localized above what is now Hawaiian Beaches destroying the coastal village of Nanawale. Apparently the summit lava lake dropped more than 100 meters at the time of the eruption, at least this was its level as first reported after the eruption several years later.

The 1840 flow, like its predecessor in 1823 was of an unusual character, although quite different than the 1823 flows, with a high percentage of green olivine phenocrysts. These were about the only significant flank eruptions from the East Rift Zone during the period of time between 1750 and 1955, a period of more than two centuries. It is little wonder that the land around and adjacent to the East Rift Zone was considered safe from lava flows when many of the housing tracts were laid out in the early fifties. This was soon to change, however, with the sudden eruption of lava from the lower East Rift Zone in 1955 preceeded by a few days of seismic activity. The summit, however, remained the site of a seathing, boiling lake of liquid rock. There were several more eruptions from the Southwest Rift Zone during this period, a fact the volcanologists to conclude that the Southwest rather than the East Rift Zone was the more active.

In December of 1919 and continuing into early 1920, lava again flowed from the Northern strand of the Southwest Rift Zone, draining the lava lake in Halemaumau and generating a flow nearly 20 km long which did not reach the sea. This flow was somewhat more gas charged than the 1823 flows, but still far less volatile rich than modern eruptions along the East Rift Zone. There is some shelly pahoehoe available on the small shield known as Mauna Iki that formed during eruption. This is the shield that is visible to the south at the viewing location on the Footprints Trail off of Highway 11 west of the summit. The end of summit magma lake (with some brief returns during the ensuing years) came with a bang, or rather a series of bangs during the phreatic summit crisis in 1924.

The map on the left shows the details of summit flows from the previous map. Note that the flows covering the floor of the summit caldera are complex and varied. This is a continuation of caldera filling that has been ongoing since the crater formed shortly before 1790. The black circle just off the east side of the caldera is the Ai-laau vent, which was the source of much of the lava mentioned above that covers the flank north of the East Rift Zone. Several major flows from this century are also shown. Along the Southwest Rift Zone is the upper part of the 1974 eruption which was associated with the end of the sustained eruption at Mauna Ulu. This eruption lasted from 1969 through 1974 and is shown on the left lower edge of the map. The green flows shown in the upper right were erupted from the observatory vent. Its possiont is shown inside the northwest region of the summit caldera. This event predated the caldera formation, and no longer exists.

As much as 750,000 cubic meters of old rock were explosively thrown from the summit area, some chunks still glowing red from volcanic heat. An eruption column rose between 6 and 7 km above the summit. The explosive activity continued for nearly two weeks, and when they were over the pit at Halemaumau was nearly 1/2 km deep with a rubble strewn floor. No fresh lava was observed at the summit, but a concommitant swarm of earthquakes near Kumukahi provided strong evidence that the magma drained from the summit was emplaced at depth withing the lower East Rift Zone. This is thought by many to be the source of the lava erupted at the beginning of the 1955 eruption that began the last few decades of intense rift zone activity along the ERZ.

Over the next ten years Halemaumau slowly refilled. Abruptly in 1934 lava disappeared from Halemaumau beginning a period of 18 years during which Kilauea appeared to have become completely inactive. Even earthquakes were greatly reduced, and no lava flowed anywhere on the surface. Then beginning early in 1950 a series of deeper than usual earthquakes announced Kilauea's emmergence from this period of seeming dormancy. The number of earthquakes slowly increased and by shortly before 1955 the summit region began to swell. These observations have been interpreted to indicate the arrival of a fresh batch of magma from the Hawaiian "hot spot". A curious but intense swarm of more than 4000 recordable earthquakes began well offshore, which may have indicated the beginning of rift zone dilation. The interpretation that this swarm originated from Loihi does not seem to be correct based on relative arrival times at the station of the rather sparse seismic network being operated at HVO during that time. By 1952 lava had again returned to Halemaumau. Beginning in late 1954 and early 1955 earthquakes became increasingly common along the lower East Rift Zone near Nanawale, apparently as rocks parted to make way for lava moving in that area. On February 28 lava began to flow from the fissures along the rift zone, and fissuring rapildy spread downrift. The first erupted lavas were highly differentiated containing long, needlelike crystals of feldspar, supporting the idea that lava had been stored locally and cooled for several decades, possible from the 1924 drainout of the summit. By early March, flows entered the outskirts of Kapoho, and the village seemed doomed. It was, but not until a subsequent eruption in 1960. An unusual observation made during the 1955 eruption was that summit deflation lagged the beginning of eruptive activity by several days. This suggests that the eruption began as a result of local volcanic processes along the rift zone and was not a direct result of summit pressurization. This is also consistant with the differentiated nature of the early erupted lavas.

The 1960 eruption began with a spectacular show of high fountaining at the summit of Kiluea along the west margin of Kilauea Iki in November, 1959. The eruption in Kilauea Iki lasted less than two months but produced a rapid series of high fountaining events with recovery periods shorter than those produced by Mauna Ulu later in the decade and much shorter than the 1 month intervals separating episodes of high fountaining at the beginning of the current eruption which began in 1983. Shortly after the eruption ended, earthquakes began near Kapoho becoming centered near the village by January 10, 1960. Kapoho occupied a graben which had become deeper during the earthquake of 1924, and the normal faults bounding the graben were reactivated. Lava broke out in a cane field three days later and Kapoho was doomed. Again activity began at the vent several days before the summit began to respond by deflation feeding fresh magma into the vent area. Also of note in the 1960 eruption was the interraction between the erupting lava and the groundwater table which produced some explosive activity and rooster tails of ash laced steam screaming from vents and fractures. Walls erected to block the flow and save Kapoho were unsuccessful. The eruption lasted just over one month.

Prior to the current eruption, the flows that built Mauna Ulu from 1969 through 1974 was by far the most voluminous historical eruption. It was also, as the eruption in 1750 and the current activity, an unstable sustained flank eruption. From early in the 50s through the onset of activity in 1969, the summit had been steadily swelling. Summit tilt reached a maximum about midway through the Mauna Ulu sequence and has been steadily deflating ever since. During the preceding decade following the 1960 eruption, small eruptions in and around the upper rift zones had been increasing in frequency. Lavas were erupting from tensional fissures, which near the summit tend to cut the rift zone axis at an oblique axis. In May, 1969 one of these began along three parallel cracks near the current site of Mauna Ulu. For the most part this eruption was similar that had been occuring in the preceeding years with one very telling exception -- this one didn't stop and continued for five years building the edivice of Mauna Loa which is visible on the left as one drives down Chain of Craters Road. Much of the bleak landscape after passing Mauna Ulu are the lava fields erupted during this period of time. Much can be said of the eruption of Mauna Ulu, and it probably deserves a complete page unto itself. Details of the lavas emplaced are shown on the right. As noted in a previous lesson, the dark flows shown here in many cases represent pahoehoe flows that converted to a'a as they spilled over the Hilina Pali. The eruption paused briefly in late 1971 followed by fissuring uprift and in the summit region. Activity at Mauna Ulu, however, resumed several month later. Eruptive activity again slowed in mid 1974 and again fissuring occured in the summit region. This time, though, activity was not resumed, and the volcano's plumbing was disrupted late the following year by a magnitude 7.2 earthquake deep in the South Flank (we'll visit this later in the course). This earthquake was associated with a seaward shifting of the flank which apparently created enough void space that little lava found its way to the surface until the beginning of the current eruption in 1983. The singular exception was a brief eruption in 1977 above Kalapana. Activity continued for only three weeks, and lava never reached the sea, although Kalapana was briefly evacuated.

The map on the right summarizes the evolution of eruptive activity since the resumption of volcanism around 1955. The period following the 1975 earthquake seems to have been dominated by intrusive activity. The upper left panel shows the distrubution of surface flows in 1955. The flows in 1960 are the dark areas of the middle left panel and the red area marks a massive intrusion that first entered the upper Southwest Rift Zone, backed up into the Koae fault system, and then appeared to intrude back towards the summit following the upper East Rift Zone. As shown here and remarked earlier, magma is sometimes stored in the Koae fault system and later erupted through southern strand of the Southwest Rift zone. The 1964 event represents one of these storage events. The lower left panel shows much of the lavas erupted during the first major portion of the Mauna Ulu eruption which continues in the upper right panel. Not the frequency of uprift intrusions and occasional fissure eruptions during this period. The middle panel on the right shows the 1975 Kalapana Earthquake (line with arrows) and subsequent activity. Note that the intense volcanism of the previous decaded ended with only the 1977 flows shown in black. During this period the remaining activity was intrusion. With time the intrusive activity increased in intensity leading up to the seismic events preceding the eruption in January of 1983 that began the long-lived flank eruption that has continued through the present.

Living on an Active Volcano

What does it mean to live on an active volcano? How does it come to be that many of the most densely populated areas of the Big Island also seem to be in areas of high probability of lava flows. Some of this mystery is revealed in this section. As discussed above, the situation through the mid 50s, as shown on the left, was radically different than the current one. The only eruption during a 200 year period shown here was the one in 1840. There had not been a single eruption this century, so it isn't all that unreasonable to assume that the East Rift Zone was less hazardous than has since proven to be the case.

Subdivisions were layed out and people moved into the area. Several residential areas actually straddle the rift zone, seeming to defy Pele's dominion of this region as revealed in the Hawaiian oral tradition. Other lessons could also have been learned from the Hawaiian tradition. For living on a volcano, the Hawaiian notion of land stewardship as opposed to the western notion of land ownership seems more appropriate. For the Hawaiians, most flows represented more of an inconvenience than a personal disaster, as long belonged to the Ali'i in large pie shaped wedges running from the ocean to the summits of adjacent volcanoes. These apua'a were large enough that most flows did not destroy them entirely. Also there was an attitude amoungst the early Hawaiians that Pele gives and takes, and it is for us to accept the use of the land as it is provided. For European immigrants, however, land ownership is much more finely defined, and even a small flow can be a personal disaster.

Consider the map on the left showing the plotting of most of the lower Puna subdivision in the early 50s. Its hard to imagine that planners in those days would have developed the region the way they did were they able to see a few decades into the future. However, despite the evidence of hazard along the South Flank adjacent to the East Rift, a new subdivision was laid recently, in part to accomodate those displaced by the destruction of Kalapana. One is compelled to think that there are more appropriate places on the Big Island to encourage people to live.

One such region is shown here south of the summit. For hundreds of years this region has been sheltered from flows by a series of summit facing fault scarps associated with the Koae Fault System. These faults are visible as a series of intertwining lines at the top of the map. Of course this region is within Hawaii Volcano's National Park and protected, but there are other regions that are equally safe. Note the coastal indentation which is also an indication of a portion of the coast that has not received as large a share of surface flows as those adjacent.

Long Period Cycles of Kialuea Volcano?

We conclude by trying to assemble a broader view of the changes and cycles that describe volcanism on Kilauea over periods of hundreds to thousands of years. The map on the left shows the character of activity during the 18th century culminating with the sustained-unstable eruption of 1750 shown in grey. Note the similarities with the character of activity during the past three or four decades, also with sustained-unstable eruptions after a period of brief fissuring along the rift zone.

The map on the right tries to put these changes into the context of changes that occurred over the preceding millenia. The upper left panel shows the caldera infilling following an earlier caldera forming eruption during approximately the 6th century. This event is associated with the production of the Uwekahuna Ash member and the formation of the Power's Caldera. Eventually this caldera filled and lava spilled from the summit from a series of events that are no longer active. The formost of these was the Ai-lau vent at the head of Thurston Lava Tube which flooded much of Puna for several centuries. This activity continued into the 15th century when rift zone activity became the dominant eruptive style, and summit eruptions became far less frequent. As mentioned above this period of reactivation of the East Rift Zone culminated with the sustained-unstable eruption of 1750 followed by a collapse of the summit and the formation of the current summit caldera in 1790. The 19th century saw a return of caldera filling with a resumption of rift zone activity only two centuries later. So, some kind of a pattern seems to emerge, although the time scale for recent activity is greatly accelerated.

Holcomb in U.S. Geological Survey Prof. Paper 1350 proposed a model (shown to the left) to explain these changes. His proposal begins in the upper frame with a sustained period of summit activity during which the elevation of the summit increases and the pressure to erupt lava at the summit increases. Eventually the increase in pressure leads to intrusion of material into the rift zones, beginning with fissure eruptions near the summit followed by increased activity along more distal sections. Eventually, as a result of thermal erosion of the rift zone a large, distal eruption drains the summit beginning a period of caldera infilling (lower panel) with little if any material finding its way into the rift zones. They cool, become more resistant to intrusions until pressure again builds sufficiently at the summit to begin the cycle again. Regardless of whether this model proves to be correct or not, it is clear that the eruptive character of Kilauea Volcano changes radically with time. Most of the time the predominant activity is massive overflows from the summit area. The second most common eruptive style appears to be caldera infilling. Least likely, it seems, is the kind of rift zone fissuring that we are currently experiencing on both Kilauea and Mauna Loa. In other words, the current situation is not at all normal for Kilauea Volcano!

Examination Questions

  1. What are the main historical eruptions along Kilauea's East Rift Zone? How are the similar, and how are they different?
  2. What are the principal Southwest Rift Zone eruptions, and how do they seem to differ from those on the East rift zone?
  3. Discuss the cyclic variations in eruptive characteristics of Kilauea Volcano during the last 2000 years. What are the implications for the next few decades?
  4. Why is the flank of Kilauea south of the summit caldera essentially free of recent flows, and how might such considerations bear on land use planning?
  5. Why is there a caldera at Kilauea's summit and how and when did it get there?
  6. Describe the evolution of volcanic processes at Kilauea’s summit emphasizing its relationship with variations in the distribution of flows on the volcano’s flanks.
  7. Describe the phreatic and phreatomagmatic processes involved in explosive summit eruptions of Kilauea Volcano? What is the evidence for these? Why do these processes seem to be absent from Mauna Loa Volcano?

If you have comments or suggestions, email me at