Kilauea is justly known for its colorful lava flows and spectacular lava fountains. But there is another, darker, side to Kilauea?s volcanic activity (fig. 1).
Kilauea is an explosive volcano. It explodes about as often as does Mount St. Helens and many other volcanoes considered to be explosive (fig. 2). The explosions at Kilauea may not necessarily produce as much tephra (a general term for all fragmented material thrown from a vent), but they can be powerful, lethal, and hazardous to air traffic as well as to people near the vent.
The designation of Kilauea as an explosive volcano may surprise some readers, but it would have been nothing special to people living in the 16th-18th centuries, when explosions probably took place every few decades or, at times, more often. Chants tell of Pele, the volcano goddess, throwing rocks to drive away an erstwhile paramour, Kamapua`a. Pele is prone to temper tantrums, probably a reflection of her explosive nature. Since Europeans arrived on the island, however, Kilauea has been in a relatively quiet, effusive, period. Its frequent eruptions of lava flows lull the public, and many volcanologists as well, into thinking that Kilauea is a milquetoast volcano. Such a reputation is unjustified.
This presentation summarizes some of the past explosive activity of Kilauea. Research is proceeding at a rapid pace to learn more about what has happened and how the explosions occurred.
For this discussion, we define explosion as an eruption that throws out solid rock with or without liquid lava. By this definition, lava fountains are not explosions; typical Hawaiian lava fountains contain effervescing liquid lava but no solid rock. In explosive eruptions, the solid rock comes from the walls of the vent or conduit, either by quarrying of wall rock or by tossing out talus that fell into a vent. Many volcanologists would argue that a lava fountain is an explosion, too. But in this presentation we want to emphasize the most hazardous kinds of eruptions, and so we don?t discuss fountains.
| large |
Figure 1 (left). Eruption column of the 1115 explosion of May 18, 1924. This explosion killed one person. Photo taken
by K. Maehara from Uwekahuna Bluff and used with permission of Bishop Museum.
Figure 2 (right). List of explosive eruptions at Kilauea and Mount St. Helens. The list is not exhaustive but is meant
to show that Kilauea erupts about as often as a well-known explosive volcano.
Figure 2 (right). List of explosive eruptions at Kilauea and Mount St. Helens. The list is not exhaustive but is meant to show that Kilauea erupts about as often as a well-known explosive volcano.
Three small explosions took place at Pu`u `O`o, one in 1987 and two others in 1991 and 1996. A visit to Pu`u O`o today shows a landscape littered with solid rocks a few centimeters to 1 m across, some quite dense (fig. 3). Each of the explosions could have been lethal to someone on the cone at the time. It is not clear what caused the explosions, but most likely a vent became clogged and gas pressure built up below it; eventually the pressurized gas burst forth powerfully enough to carry solid rocks with it. There is probably no way to predict such a small explosion; that is just one of the reasons that Pu`u `O`o is a dangerous place to visit.
Figure 3. Rocks on north flank of Pu`u `O`o resulting from explosions in 1987, 1991, and 1996. Largest blocks in image are 30 cm across. Others not shown are 2-3 times that.
In May 1924, explosions took place from Halemaumau at intervals of a few hours during a 2-1/2 week period. The largest explosion was on Sunday, May 18 (the same date and day of the week as the large explosion of Mount St. Helens in 1980) and killed a bystander (fig. 1).
Ruy Finch, HVO seismologist at the time, calculated that ash rose as high as 9 km (30,000 ft) during one of the explosions. Large boulders, some weighing 10-12 tons, fell around Halemaumau and as far as 1 km away (fig. 4). Visitors today see examples of these boulders along the trail to Halemaumau (fig. 5).
Fine ash was mostly blown southwest from Halemaumau, dusting the village of Pahala and nearby agricultural land 30 km away. Some mud fell northeast of the volcano, causing railroad traffic in lower Puna, 40 km from Halemaumau, to be curtailed owing to slippery tracks. Ash washed from the roof of a store in the hamlet of Glenwood overloaded gutters, causing them to collapse.
A feature story on this web site describes in some detail the events that led up to the explosions, as well as the explosions themselves. The explosions appear to have been driven by heated groundwater and hence are called phreatic by volcanologists.
| large |
Figure 4 (left). Block weighing about 10 tons thrown 1 km from center of Halemaumau during an early explosion in May
1924. Later explosions deposited ash that filled any crater that might have been created when the block crashed to the
Figure 5 (right). Rocks ejected during 1924 explosions about 150 m from southeast rim of Halemaumau. Note car in parking
lot in background, and fence along Halemaumau trail in upper right. This scene is typical of areas near the crater that
have not been disturbed by human activities.
Figure 5 (right). Rocks ejected during 1924 explosions about 150 m from southeast rim of Halemaumau. Note car in parking lot in background, and fence along Halemaumau trail in upper right. This scene is typical of areas near the crater that have not been disturbed by human activities.
These explosions were responsible for the volcanic ash and coarser debris that visitors see when they enter Hawai`i Volcanoes National Park and drive around the caldera. Especially good outcrops occur along the walls of a wide crack at a favorite visitor stop, Southwest Rift (fig. 7).
Figure 6 (left). Part of cliff southwest of Keanakako`i Crater that is the "type locality" of the Keanakako`i Ash. A type
locality in geology is a location that displays important characteristics of a rock unit that serve as a reference for future
workers. The cliff is about 12 m high. The oldest beds in the Keanakako`i Ash are exposed along
the cliff but are not visible
in this photo.
Figure 7 (right). Beds of Keanakako`i Ash form the northwest side of large ground crack, known by the National Park's sign
as Southwest Rift, one of many such cracks in the southwest rift zone of Kilauea. Lava flow erupted in 1974 drapes part of
wall and caps the ground surface above the crack.
Figure 7 (right). Beds of Keanakako`i Ash form the northwest side of large ground crack, known by the National Park's sign as Southwest Rift, one of many such cracks in the southwest rift zone of Kilauea. Lava flow erupted in 1974 drapes part of wall and caps the ground surface above the crack.
Early geologists recognized that the explosions that formed the Keanakako`i Ash took place over a long period of time, but more recent workers mistakenly interpreted them all to have occurred in 1790 or at most a few years earlier. This has led to the common short-cut expression ?1790 explosions,? or some variant thereof, as a synonym for the entire Keanakako`i Ash. Ongoing work by scientists from HVO, Smithsonian Institution, and University of Hawai`i shows the prolonged nature of the explosive period. There were indeed explosions in 1790; in fact, one of them killed about 80-800 people (estimates differ) in the most lethal known eruption of any volcano in the present United States. The 1790 explosions, however, simply culminated (or at least occurred near the end of) a 300-yr period of frequent explosions, some quite powerful.
Many explosions and lava fountains were relatively small and formed deposits confined to the proximity of Kilauea?s caldera (fig. 8). Some deposited ash in what is now Volcano Village and other communities adjacent to the National Park. Most were certainly larger than the explosions of 1924 and would have been lethal to many people in the summit area. Some produced surges, one of the most dangerous kinds of volcanic eruption (fig. 9). But, most of the explosions and fountains were not large enough to produce high eruption columns or to have dispersed ash long distances from the summit.
Figure 8 (left). Coarse debris in Keanakako`i Ash exploded from the summit of Kilauea, probably in the 18th century,
exposed in road cut in southwest part of Kilauea's caldera. Most of this debris was erupted during relatively small
explosions and showered onto nearby areas. Note that the large block (about 40 cm across) in upper left indents the ash
Figure 9 (right). Thin beds containing glassy ash, scoria, and even pumice,
in the Keanakako`i Ash near Sand Hill, 2 km southwest of center of
Halemaumau. A shallow erosional contact, which may have formed
as a surge passed across the area, separates an older gray tephra unit from
a younger brown unit.
Figure 9 (right). Thin beds containing glassy ash, scoria, and even pumice, in the Keanakako`i Ash near Sand Hill, 2 km southwest of center of Halemaumau. A shallow erosional contact, which may have formed as a surge passed across the area, separates an older gray tephra unit from a younger brown unit.
At least three powerful explosions are great exceptions to this generalization, however, and they are very important in understanding the explosive hazards of Kilauea.
One such explosion in the late 1500s or early 1600s, termed ?layer 6? by volcanologists, spread scoria and some small solid rocks southeastward from Kilauea?s summit to the seacoast some 20 km away (figs. 10 and 11). The scoria has a median grain size of about 5.5 mm nearest its source, with some pieces as much as 6 cm across. A few dense rock fragments reach 3-4 cm diameter. At the coast, the scoria is of coarse sand size (with a median size of about 1 mm; fig. 14), and still finer ash must have blown much farther across the Pacific.
Figure 10 (left). Layer 6 scoria (in Keanakako`i Ash), product of a powerful explosion of crystal-rich
magma and a few pieces of solid rock. The scoria is interbedded with fine ash. The massive pink ash just below the knife has
numerous accretionary lapilli. Site is 1.9 km southeast of center of Halemaumau.
Figure 11 (right). Map showing margins of layer 6 scoria (black) and lithic layer 14 (magenta) in Keanakako`i Ash southeast
of Kilauea's summit. Note that layer 14 has a more easterly dispersal direction than
does layer 6. For comparison, the approximate
limit of fallout from the 1959 Kilauea Iki lava-fountaining eruption is shown in red; note that the 1959 tephra was dispersed
southwestward from its vent--down the trade wind--and covers a much smaller area than does the tephra of layers 6 and 14.
This map is a work in progress--hence the queries.
Figure 11 (right). Map showing margins of layer 6 scoria (black) and lithic layer 14 (magenta) in Keanakako`i Ash southeast of Kilauea's summit. Note that layer 14 has a more easterly dispersal direction than does layer 6. For comparison, the approximate limit of fallout from the 1959 Kilauea Iki lava-fountaining eruption is shown in red; note that the 1959 tephra was dispersed southwestward from its vent--down the trade wind--and covers a much smaller area than does the tephra of layers 6 and 14. This map is a work in progress--hence the queries.
A fine ash, commonly containing accretionary lapilli (?mud pellets?) 1 mm or more in diameter, was deposited some time after layer 6 (figs. 10 and 12). It has a similar or even more far-flung dispersal than that of layer 6.
|Figure 12. Keanakako`i Ash with abundant accretionary lapilli, largest of which are about 3 mm in diameter. This closely resembles, but may not be the actual, far-flung ash mentioned in the text.|
A third powerful explosion ejected dense rocks (?lithics?) ranging in size from boulders and gravel to medium sand that were deposited east-southeast from Kilauea?s summit (figs. 13 and 14). Deposits from this large explosion (interpreted as layer 14 for now, though this may change with more work) have been found as far east as Pu`u `O`o, and its eastward limit has not yet been discovered (fig. 11). Debris from this explosion fell at the coastline near the end of the Chain of Craters Road (figs. 11 and 14), where the heavy rocks are of medium sand size (about 0.5 mm median grain size); finer ash must have been blown far beyond the island. This explosion may have taken place in, or a short time before, 1790, though more work is needed to be sure. The edge of the explosion deposit buries human footprints south of Halemaumau.
Figure 13 (left). Bed of coarse, boulder- and gravel-size lithic ejecta in layer 14 of the Keanakako`i Ash
1.8 km east-southeast of center of Halemaumau. Largest block is about 1 m in diameter.
Figure 14 (right). Scoria of layer 6 (left of car key, popcorn-like) is separated by a thin bed of fine ash from the overlying
lithic sand-size ash of layer 14 (bed at level of tip of key, best shown in middle of image), only 1.5 km from the ocean at
the junction of Puna Coast Trail and Chain of Craters Road. Median grain size of the scoria is 1.1 mm and of the lithic ash,
Figure 14 (right). Scoria of layer 6 (left of car key, popcorn-like) is separated by a thin bed of fine ash from the overlying lithic sand-size ash of layer 14 (bed at level of tip of key, best shown in middle of image), only 1.5 km from the ocean at the junction of Puna Coast Trail and Chain of Craters Road. Median grain size of the scoria is 1.1 mm and of the lithic ash, 0.5 mm.
The deposits of these three explosions are relatively thin. Layer 6 is less than 50 cm thick even near the caldera (fig. 15), the accretionary lapilli-bearing ash is nowhere more than about 5 cm thick, and the lithic deposit is no more than a few tens of centimeters thick (fig. 16). In distant areas, the thickness is only 1 cm or less. The wide dispersal and direction of fallout imply high eruption columns, a critical point returned to later in this presentation.
Figure 15 (left). Layer 6 scoria bed in Keanakako`i Ash is massive dark layer about 1 one-third of way
from bottom of image. This exposure is one of the closest to the vent that is preserved. The thickness is 36 cm. White
specks are secondary minerals deposited by acidic moisture in this volcanically warm area.
Figure 16 (right). Gravel-size lithics in layer 14 of the Keanakako`i Ash, about 1.5 km from center of Halemaumau. Base
of bed is at bottom of image, covered by loose gravel sloughed from cut. Note car key for scale.
Figure 16 (right). Gravel-size lithics in layer 14 of the Keanakako`i Ash, about 1.5 km from center of Halemaumau. Base of bed is at bottom of image, covered by loose gravel sloughed from cut. Note car key for scale.
Research during the past 10 years has uncovered a previously unknown period of Kilauea explosions. Their deposits are being called the Kulanaokuaiki tephra. The Kulanaokuaiki is currently defined as the upper part of the previously named Uwekahuna Ash. The nomenclatorial details are nettlesome to specialists, but we can ignore them here.
Material erupted during the first explosion consists mainly of glassy pumice and scoria with a small percentage of solid rocks. The explosion was perhaps a very powerful lava fountain, so powerful that it could carry with it dense lithics probably derived from the walls of the conduit. We call it Kulanaokuaiki 1 (or K-1 for short).
Perhaps the most powerful of all known Kilauea explosions took place in the A.D. 800-1000 time period. This tremendous explosion, known as K-3, threw rocks weighing more than 4 kg farther than 7 km from Kilauea?s summit (fig. 17). Dense rocks the size of golf balls fell at the coast (Halape and Keauhou Landing) 15 km southeast of the summit (fig. 18). Coarse-grained rocks from deep within the volcano, known by their geologic name, gabbro, were caught up in the explosion and thrown out with the rest of the debris (fig. 19). Very little liquid, mainly crystal-rich scoria closely resembling that of Keanakako`i layer 6, was ejected; much of the material was already solid rock.
Figure 17 (left). Slightly vesicular ejecta block in
Kulanaokuaiki tephra (unit K-3), photographed in situ on Kilauea's south flank 7.6 km from
center of Halemaumau. The block weighs
Figure 18 (right). Collections of the 12 largest lithic ejecta (Kulanaokuaiki
tephra, unit K-3) found during
18-minute hunts at two locations on Kilauea's south flank. Site 16 (top) is 7
km, and site 103 (bottom), 15.3 km, south-southeast of Halemaumau. Same
ruler in both images
(short divisions in centimeters, long divisions in inches).
Comparison of images shows striking falloff in ejecta size with distance.
Figure 18 (right). Collections of the 12 largest lithic ejecta (Kulanaokuaiki tephra, unit K-3) found during 18-minute hunts at two locations on Kilauea's south flank. Site 16 (top) is 7 km, and site 103 (bottom), 15.3 km, south-southeast of Halemaumau. Same ruler in both images (short divisions in centimeters, long divisions in inches). Comparison of images shows striking falloff in ejecta size with distance.
Another powerful explosion (K-5) followed. It was dominated by crystal-rich scoria but still contains a significant lithic component.
Debris erupted during the K-3 explosion fell only southeast of Kilauea?s summit (fig. 20). The other two explosions were apparently somewhat smaller than K-3 and produced fewer ejecta; debris from each of them is also known only from the area southeast of the summit.
Figure 19 (left). Block of gabbro weighing 1.73 kg found in K-3 unit of Kulanaokuaiki tephra 6.9 km
south-southeast of center of Halemaumau. The block is about 15 cm long by 11 cm wide.
Figure 20 (right). Generalized contours (isopleths), in centimeters, of average nominal diameter of
the 2nd,- 3rd,- and
4th-largest lithic clasts found in K-3 unit of the Kulanaokuaiki tephra at locations indicated by red stars on Kilauea's
south flank. The shape of the isopleths indicates a source at Kilauea's summit. Colors denote various age groupings of lava
flows on Kilauea. All colors except green indicate lava flows younger than K-3 unit. Note sample site in tiny kipuka
just west of big bend in Chain of Craters Road; it helps constrain the 2-cm contour.
Figure 20 (right). Generalized contours (isopleths), in centimeters, of average nominal diameter of the 2nd,- 3rd,- and 4th-largest lithic clasts found in K-3 unit of the Kulanaokuaiki tephra at locations indicated by red stars on Kilauea's south flank. The shape of the isopleths indicates a source at Kilauea's summit. Colors denote various age groupings of lava flows on Kilauea. All colors except green indicate lava flows younger than K-3 unit. Note sample site in tiny kipuka just west of big bend in Chain of Craters Road; it helps constrain the 2-cm contour.
The Kulanaokuaiki deposits overlie weathered volcanic ash on Kilauea?s south flank southeast of the summit (fig. 21). This ash, probably part of the Uwekahuna Ash, indicates older periods of explosions but younger than about 2,300 years ago, the ago of the underlying lava flow.
Other workers have described still older ash deposits more than ten thousand years old, exposed in Hilina Pali and other steep parts of Kilauea?s faulted south flank. We have not studied these deposits but mention them to show that Kilauea?s explosive past reaches far back in time.
|Figure 21. Red-brown ash underlies dark brown Kulanaokuaiki ash on south flank of Kilauea, 7.5 km southeast of center of Halemaumau. The older ash is younger than about 2,300 years but older than the Kulanaokuaiki deposits.|
Bob Dylan had it right for Hawai`i. ?You don?t need a weatherman to know which way the wind blows.? There are really only two directions of wind that people feel: the dominant northeast trade wind, and the south kona wind. Most ash blown by these winds will therefore fall southwest of a vent during trade winds, and some will fall north of the volcano during kona winds.
But Kilauea's most powerful explosions produced debris that fell east-southeast to south-southeast from the site of the explosions at Kilauea?s summit. This is a fundamental characteristic that distinguishes the powerful explosions from their weaker cousins, which erupted debris that was blown southwestward or northward. Another important characteristic of the powerful explosions is the relative thinness of their deposits. A third characteristic is the distant dispersal of the ash.
We are currently working on ways to explain all three of these signatures. We are far from understanding the mechanism, but the following represents our current thinking. It is not the final answer, but we believe it is a step in the right direction.
"The answer is blowin' in the [high-level] wind." All of these characteristics can be explained by an eruption column that went to great heights, far higher than those reached by the more common lava fountains and small explosions. A premise is that ash must rise to such a height in order to be distributed far from the source. That alone suggests that the eruption columns of the powerful explosions rose far above the vents to produce fallout at least 20 km from source and doubtless much farther beyond the coastline.
How high? At heights of about 9 km or more, strong tropospheric winds are encountered above Hawai`i that typically have very different directions from those of the trade and kona winds (fig. 22). Those directions are out of the southwest to northwest quadrants, often at speeds of 40-80 knots (75-150 km/hr). Winter winds are more often out of the northwest quadrant than are summer winds. Frequent air travelers from the U.S. mainland to Hawai`i know these winds as the jet stream and realize that travel in the summer often takes a bit longer than in the winter, because the plane is more likely bucking a strong southwest wind.
Figure 22. Diagrams showing predicted wind directions and velocities over Mauna Kea at an elevation of 16.5 km (54,000 ft) above sea level, taken from daily data on the University of Hawai`i web site (http://mkwc.ifa.hawaii.edu/forecast/mko/index.cgi) managed by Ryan Lyman. Velocities are indicated by concentric circles, in 10-knot intervals. (A knot is 1.15 mph, 1.85 km/hr, or 0.51 m/s.) The diagrams, showing predicted winds for January-June 2005, suggest that winter wind (left diagram) is more often from the northwest than is summer wind (right diagram), but that the dominant direction during all seasons is from the west rather than the north or south. Diagrams starting at about 9.5 km (31,000 ft) show similar patterns. We are currently analyzing measured wind speeds and directions from 10 years of radiosonde data above Hilo to get a better understanding of high-level wind directions and velocities.
These high-velocity, high-elevation westerlies provide a convenient way to distribute ash east of Kilauea?s summit, and northwest winter winds could even blow the ash into the southeast quadrant. Ash falling from these winds will generally drop into the familiar low-elevation trade wind, which will blow the ash back toward the southwest. The net result should commonly be a distribution southeast of Kilauea?s summit for ash erupted to heights above 9 km (fig. 23). Most likely, though, the height would have to be considerably above 9 km, for some of the ash must be transported as far as 20 km or more before it encounters the surface winds. We suspect that eruption column heights of 15-20 km are likely possibilities, given the mix of wind velocities and dispersal distances available.
Figure 23. Plan-view sketch showing how, initially, high-level westerlies blow tephra eastward. Eventually the tephra falls into the low-level trades, which blow it southwestward. The net result is distribution of the tephra southeast of the vent. The exact trajectories of the tephra reflect interplay between column height, high-level wind direction and velocity, and low-level wind velocity and, to a lesser extent, direction.
These heights are well above those at which passenger jets travel. A cruising elevation of 35,000 ft is about 10.5 km. Interisland traffic flies at lower elevations. It is well known that jet aircraft and ash don?t fare well together. Ash taken up by jet engines may cause engine failure or deterioration in performance. Ash may severely abrade windows and any forward-facing surfaces. As of 2000, seven encounters of commercial aircraft had caused in-flight loss of engine power the led to near crashes.
Ash from Kilauea explosions, the powerful kind or even the smaller ones, can be hazardous to air travel as well as to people on the ground.
The more powerful explosions in the Keanakako`i and Kulanaokuaiki deposits are probably a different matter. The size of ejecta blocks in the deposits far away from the vent is too great to be explained by existing phreatic models. The great column height is probably too much as well; the largest phreatic column in 1924 just reached 9 km, though even that would have been enough to impact present-day interisland air traffic.
We are looking into ways to get such explosions from pressurized volcanic gas within the volcano. Carbon dioxide is a reasonable candidate, given its abundance at Kilauea (8000 tons or more are emitted from the summit daily) and its low solubility in basaltic magma. The K-3 eruption brought up gabbro that was in the process of crystallizing when it was entrained in the explosion at a depth of several kilometers (fig. 24). Water is still dissolved in magma at that depth and so can't serve as a driver, but CO2 can exist in a gas phase and hence could become the driving force for the explosions.
Figure 24. Sketches illustrating one way to produce the Kulanaokuaiki K-3 explosions. Panel 1: CO2 pressure builds deep within volcano at top of magma body, which has crystallized substantially (as shown by the crystalline scoria erupted.) Gabbro is forming in pockets adjacent to the magma. Panel 2: Volatile pressure overcomes overburden load, and explosion commences, carrying pieces of gabbro, wall rock, and magma in a process analogous to eruption of a diatreme. Panel 3: Eruption column reaches elevation possibly more than 13 km, and tephra is blown southeastward by strong wind. It falls onto Kilauea's south flank, the finer ash possibly influenced by low-level trade wind (not shown).
The bad news about any deep-seated explosion is that we don?t know what, if any, early warning there might be. Nothing such as this has occurred during the period of scientific study of Kilauea. If CO2 is the culprit, perhaps we would see a downturn in the emission rate of CO2 at the summit as the gas is being retained at depth. But could we distinguish that from decreased flux of magma to the volcano? Possibly, because decreased magma flux should lead to deflation whereas increased pressure should lead to inflation. But the unfortunate bottom line is that we will probably have to experience a large explosion before we can develop a means to anticipate the next one.
One thing is sure, however. Kilauea will lose its reputation as a milquetoast volcano after the next big explosion.
Decker, R.W., and Christiansen, R.L., 1984, Explosive eruptions of Kilauea Volcano, Hawaii: Explosive volcanism: inception, evolution, and hazards, National Academy Press, Washington, D.C., p. 122-132.
Easton, R.M., 1987, Stratigraphy of Kilauea Volcano, in Volcanism in Hawaii, chap. 11 of Decker, R.W., Wright, T.L., and Stauffer, P.H., eds., Volcanism in Hawaii: U.S. Geological Survey Professional Paper 1350, v. 1, p. 243-260.
Jaggar, T.A., and Finch, R.H., 1924, The explosive eruption of Kilauea in Hawaii: American Journal of Science, Fifth Series, v. 8, p. 353-374.
McPhie, Jocelyn, Walker, G.P.L., and Christiansen, R.L., 1990, Phreatomagmatic and phreatic fall and surge deposits from explosions at Kilauea volcano, Hawaii, 1790 A.D.: Keanakakoi Ash Member: Bulletin of Volcanology, v. 52, p. 334-354.