Kilauea Volcano, Hawaii
Slide set compiled by Pete Mouginis-Mark, University of Hawaii
Many of the thermal studies of volcanoes that we will conduct with ASTER and MODIS are being
investigated using the 12-year long eruption of Kilauea at the Pu'u O'o/Kupaianaha site on
the East Rift Zone. We have been using a variety of field, airborne, and spaceborne instruments to
study the thermal characteristics of lava lakes, active lava flows, active lava tubes, and
recently erupted lava flow fields.
Almost every time we visit Kilauea, something is different about the eruption. When it was
active (1986 - 1991), the Kupaianaha lava lake varied in its degree of vigor (rate of overturning)
and depth beneath the rim (shallow at first, progressing to more than 30 meters below the rim late
in its lifetime). Active pahoehoe flow fields are transient in their location and rate of
movement. A'a flows, while more unusual, are also quite variable. Even the location and level
of activity of lava tubes changes on timescales from a few days to a year.
Here we show some of the typical eruptions that we have seen during the course of collecting
data to test our EOS algorithms. These views can generally be divided into four categories: lava
lakes, plumes, lava flows, and the practicalities of field measurements of lava temperature. These
topics closely match some of the phenomena we are studying as part of our preparation for EOS
data. All of the following pictures were taken by Pete Mouginis-Mark unless otherwise
SLIDE #1 (151K)
During the first year of its activity, the Kupaianaha lava lake was frequently full, with
numerous sheet pahoehoe flows overtopping the rim. In this view (taken in January 1987) we see the
lake in a period of inactivity with large plates of cooling lava crust on the surface. When this
image was taken, there had been no overturning event for more than 10 minutes, so the lake surface
was relatively cool.
SLIDE #2 (190K)
This image was obtained in December 1988 when the lake surface was about 25 meters below the
rim. The field of view is about 40 m wide. Numerous plates are on the surface, with one active
shear zone (at left) and some recent (< 10 seconds) spatter at lower center. The majority of
the lake surface had been constructed over the last ten minutes, with the oldest parts of the
surface no more than 30 minutes old. This view is typical of the lake as described in Flynn et al.
SLIDE #3 (164K)
Another view of the lake in December 1988, showing a rare (a few times per hour) vigorous
episode of overturning of the surface.
SLIDE #4 (172K)
Spattering of incandescent lava on the surface of the lava pond in October 1987 illustrates the
degree of difficulty in interpreting satellite thermal observations of a lava lake. Here we see
spatter 3 to 4 m high over several parts of the lake surface. We know from spectrometer data that
these small areas of very hot material dominate the thermal emission from the surface, making it
very difficult to characterize the lava lake without some addition field knowledge.
SLIDE #5 (118K)
We have not spent much time performing spectral studies of fire fountains, primarily because
this type of activity at Pu'u O'o ended in 1986 prior to the development of suitable
instrumentation. Here we see the Phase 34 eruption of Pu'u O'o in August 1984. The plume
is estimated to be about 200 m high to the top of the glowing spatter. Above this is a very
prominent thermal plume that has entrained lots of ash and water vapor from the adjacent
SLIDE #6 (161K)
In early 1984, the area uprift of the Pu'u O'o cone was vigorously fuming above the
rift zone. If this were to occur during the EOS mission, we would expect that thermal IR data from
ASTER could be used to study the spatial and thermal variations in these fumaroles.
SLIDE #7 (114K)
There are many instances when vigorous degassing at the Pu'u O'o vent sends dense
clouds of sulfur dioxide and water vapor over the Chain of Craters Road of Kilauea. As we develop
more skills in using the FieldSpec spectrometer, which operates at a wavelength of 0.35 to 2.4
micrometers, we hope to study the distribution of SO2 using ultraviolet measurements. Other
instruments that Paul Lucey at the University of Hawaii is developing also collect data between
1.0 to 5.0 micrometers, so we can intercompare data that cover the 3.9 micrometer SO2 absorption
band as well as the 0.36 micrometer SO2 feature.
SLIDE #8 (197K)
To date, no high resolution spectral data have been collected over active a'a flows.
However, such observations are a high priority for us, since numerical models developed by Joy
Crisp and Steve Baloga use the temperature distribution to predict the emplacement characteristics
of these flows. Image by Scott Rowland.
SLIDE #9 (201K)
Many of our FieldSpec spectrometer measurements in late 1994 were concentrated on pahoehoe
flows similar to the one seen here. We are trying to collect data on the cooling rate of an entire
flow lobe, so we select features that can repeatedly studied over a period of about an hour
without drastic changes in viewing geometry.
SLIDE #10 (148K)
There is a lot of interest in the rate of postemplacement inflation of pahoehoe lava flow
fields in Hawaii. Field observations by the U.S. Geological Survey indicate that several meters of
uplift can take place after initial emplacement of the flow. Here we see an example of this
inflation, with the top surface of the flow having been raised about 20 cm above its original
level due to the continued injection of new material into the flow lobe. The thermal
characteristics of inflating flow lobes is another high priority target for our future
spectrometer observations, and we have been studying the daily and weekly thermal characteristics
of Hawaiian flows using AVHRR data. These results have been reported in Mouginis-Mark et al.
SLIDE #11 (151K)
Our interest in active lava tubes was initiated during our Landsat TM study of Kilauea during
July, 1991, which we describe in Flynn et al. (1994). We have since returned to the flow field and
collected FieldSpec data of the active lava in the tube and the roof over the tube. We hope to
prepare a manuscript for publication on these observations in early 1995.
SLIDE #12 (194K)
One of our early attempts to study the temperature of active flows had Scott Rowland using a
thermocouple to obtain direct measurements. These data give temperatures in the 1120 - 1150
centigrade range when the flow is active, but are impossible to obtain over the following few
hours as the flow starts to cool.
SLIDE #13 (140K)
The first high spectral resolution instrument that we used for temperature measurements was a
GER Field Spectrometer. Here we see Jonathan Gradie collecting supplemental observations while the
spectrometer studies the Kupaianaha lava lake in October 1987. These data were subsequently
described in Flynn et al. (1993).
SLIDE #14 (171K)
Another instance where we used the GER spectrometer was during the February 19, 1992 eruption
of Pu'u O'o. We were really lucky to have planned a field trip that took us to the
eruption site just hours after the onset of a major flank eruption on the uprift side of the
Pu'u O'o cone. Here we see part of our group collecting data about 30 meters away from a
highly active lava channel. Results from the observations made from this site were reported by
Flynn and Mouginis-Mark (1994). Image by Henning Haack.
SLIDE #15 (174K)
The size of the February 19, 1992, flow was really impressive. Here we see the flow extending
more than 1 km away from the Pu'u O'o cone (note for scale the group of people at the
bottom of the image). Note the large amount of fume coming off of the flow. Steve Self has been
studying the degassing history of comparable flows in Iceland, and concludes that basaltic flows
such as these add significant amounts of sulfur dioxide and aerosols to the troposphere for
periods that may be as short as a day (as in this case) to as long as 9 months (in the case of
Laki in Iceland). Mauna Loa is in the background.
SLIDE #16 (146K)
For our thermal studies, this is the next evolutionary step in our field instrumentation. The
FieldSpec spectrometer is much lighter than the GER and collects data more than two orders of
magnitude more quickly. Here we see Luke Flynn checking the spectrometer before studying active
lava pahoehoe flows in August, 1994.
SLIDE #17 (132K)
We have also been investigating the best method to use to remove solar radiance from spectral
measurements of lava flow temperatures. The best way to reduce this problem is to collect the data
at night, and we reported these results in Flynn and Mouginis-Mark (1992). Here we see the active
Pu'u O'o cone illuminated by both surface flows associated with the February 19, 1992,
eruption, and a full Moon rising over the volcano. Without doubt, this view also shows why we
enjoy doing field work!
If you are interested in reading more about our thermal studies of Kilauea, you might be
interested in some of our recent publications:
- Flynn, LP, and PJ Mouginis-Mark (1992) Cooling rate of an active Hawaiian lava flow from
nighttime spectroradiometer measurements, Geophys. Res. Lett., 19: 1783-1786.
- Flynn, LP, and PJ Mouginis-Mark (1994) Temperature of an active lava channel from spectral
measurements, Kilauea volcano, Hawaii, Bull. Volcanol., 56: 297-301.
- Flynn, LP, PJ Mouginis-Mark, JC Gradie, and PG Lucey (1993) Radiative temperature measurements
at Kupaianaha lava lake, Kilauea Volcano, Hawaii, J. Geophys. Res., 98: 6461-6476.
- Flynn, LP, PJ Mouginis-Mark, and KA Horton (1994) Distribution of thermal areas on an active
lava flow field: Landsat observations of Kilauea, Hawaii, July 1991, Bull. Volcanol., 56:
- Mouginis-Mark, PJ, H Garbeil, and P Flament (1994) Effects of viewing geometry on AVHRR
observations of volcanic thermal anomalies. Remote Sens. Environ., 48: 51-60.
Back to Volcanoes Under Study
Return to Main EOS Volcanology Page