Robert W. Collier, Jack Dymond and James McManus
College of Oceanography
Oregon State University
Corvallis, OR 97331
Introduction
<<
Table of
Contents
>>
Crater Lake sits within the caldera of Mt. Mazama, a center of volcanism in the
Oregon Cascades for more than 400,000 years. The morphology of the lake is
largely a consequence of a climactic eruption that occurred 6845 + 50 years ago;
however, intercaldera volcanism took place as recently as 4000 years ago. The
volcanic morphology provides a basin for what is now the deepest lake in the
United States (approximately 590 meters). The volcanic terrain strongly limits
the nutrient fluxes into the lake, mostly because the lake covers 78% of the
total drainage area. Consequently, the lake is highly oligotrophic and one of
the clearest lakes in the world.
Our studies of
Crater Lake began as an attempt to understand the important physical and
chemical characteristics of the lake and the processes which produced some
unusual sediment compositions, consistent with hydrothermal inputs to the lake.
Given the location of the lake directly above a relatively recent and major
magmatic source, thermal spring input to the lake would not be surprising. Such
a source was first suggested by the USGS scientist Van Denburgh in 1968, who
noted the relatively high sulfate and chloride content of Crater Lake as
compared to nearby Davis Lake, and suggested that these two constituents " ....
may have been contributed to the lake by thermal springs or fumaroles, probably
located below the present lake level. Such springs and fumaroles are a common
expression of hydrothermal activity at a site of volcanic eruptions."
|
 |
Since 1968, many other investigators have shown thermal, chemical, and isotopic
evidence of active hydrothermal inputs to the lake. But evidence of hydrothermal
sources to the lake have not gone unchallenged. Others argued against any
hydrothermal inputs to the lake and suggested that underestimated evaporation
rates, fumarolic inputs, and weathering of Mazama ash could account for the
anomalous composition of the lake. They suggested that conductive heating rather
than convective inputs accounts for the temperature anomalies.
The question of
whether there are hydrothermal sources to Crater Lake has important implications
for understanding the lake ecology. For example, the relatively rapid mixing
that has been suggested for Crater Lake may be a consequence of hydrothermal
inputs of heat to the bottom of the lake. If this is true, temporal variability
in this source could impact the nutrient cycling and the plant productivity of
the lake. In 1987, in response to the requirement of Public Law 99-591 for
identification of significant thermal features in national parks, the National
Park Service funded a three-year program to evaluate possible hydrothermal
sources to Crater Lake. The research, which is part of an ongoing limnological
study of Crater Lake, was designed to: (1) define the thermal and chemical
variability in the deep lake, (2) examine the data for evidence of a
hydrothermal source, (3) design and carry out a program that would find possible
venting sites and sample any associated fluids, and (4) evaluate alternative
mechanisms to explain the observed thermal and chemical variability. The text
and photos below offer a brief summary of the results from this research
program.
Field Methods
The results
summarized below, cover field programs and laboratory analyses which have been
carried out between 1987 and 1991. During the summer of 1987, our research group
spent 20 days at Crater Lake, during which time we performed extensive sampling
operations and made thermal and chemical measurements from the surface research
vessel. We introduced the use of a remotely operated vehicle (ROV) for making
detailed observations of the bottom of Crater Lake. Seven ROV deployments were
made which provided information on the sediment thickness, the presence of
benthic plants and animals, and the existence of unusual precipitates and crusts
which were in marked contrast to the normal buff-colored sediments that blanket
much of the bottom.
The hydrothermal
field program for 1988 was based on three separate expeditions to the lake
during the summer when the lake is accessible. These included extensive sampling
programs from surface boats and a 25 day dive program using the submersible
Deep-Rover (http://www.nuytco.com/deeprover.html).
The one-person submersible was used to locate, observe, and sample geological,
geochemical, and biological features in the deep lake that had been hitherto
inaccessible. The 1989 field programs continued the surface and submersible
studies between June-September. In all, we made over 47 dives to the lake bottom
with Deep Rover and 7 ROV dives.
Summary: Observations Related to Active Hydrothermal Features
Measurements of temperature and salt content within the South Basin of Crater
Lake show surprising variations over distances of a few meters. These thermal
and salinity gradients can only be maintained by a continuing input of anomalous
fluids.
Communities of
bacteria, which produce impressive mat features on rock outcrops and sediment
surfaces, mark sites of deep lake venting. The mats have internal temperatures
which are more than 15°C higher than lake bottom water. These communities
apparently use the abundant reduced iron in the advecting fluids to fuel their
metabolism. Although there were no visible indications of fluid flow through or
from the mats, fluid advection is necessary in order to provide the continuous
input of reduced chemical species which is required
for the survival of these prolific bacterial communities. The temperature
gradients within the mats indicate that the advection rates are as high as 100
meters per year. Consequently, the bacterial mats are visual markers of
thermally and chemically enriched fluid venting.
Pools of saline
water, with major element contents that are approximately ten times greater than
background lake values, have been discovered in two widely separated areas of
the lake. Sediment pore water compositions from some South Basin cores are
similar to those of the pools. The pore water measurements define non-linear
gradients which indicate vertical fluid advection rates of up to two
meters/year. These measurements as well as the major element compositions
suggest that the fluids advecting through the sediments, the brine pools, and
the bacterial
mats are derived from a similar source. Results from chemical geothermometry
determinations suggest that this source equilibrated with silicate rocks at
temperatures ranging from 40 to 165 °C. In addition to these expressions of
active inputs, we discovered inactive silica-rich spires over 10 meters tall,
located on the lake floor near the base of Skell Head. Their morphology and
chemistry suggests they were formed underwater during earlier hydrothermal
episodes.
|
 |
Sampling of
the mat fluids, the brine pools, and sediment pore waters has dramatically
increased the known range of anomalous water compositions within Crater Lake. In
the most anomalous fluids manganese is enriched by as much as a million times
and Radon is enriched 100,000 times over typical lake values. Helium-3, perhaps
the most distinctive indicator of a magmatic source, is enriched 500 times over
values for waters in equilibrium with the atmosphere. Striking depletions of
C-14 in pool fluids and the deep lake waters indicate a magmatic source of "dead
carbon" is entering the deep lake. Rare earth element concentrations in lake and
sediment pore waters have an abundance pattern which indicate a hydrothermal
source. Isotopic compositions of hydrogen in the saline pools clearly show that
these anomalous fluids are highly modified lake water and could not have
originated outside the lake.
The enhanced
salt content of the anomalous fluids enables us to account for the bulk
composition of the lake by elucidating the sources of chemical species which
were previously unexplained by known water sources such as precipitation and
caldera springs. We have identified a third source—a hydrothermal component—as
the major influence on lake composition. Using sensitive analytical methods, we
have monitored the active accumulation of heat and salt in the deep lake that
results from this source. Various mass balance models indicate that a net heat
flow of 15 to 30 megawatts (MW) is carried into the lake by thermal fluids. The
calculated flow rates for a thermally and chemically enriched fluid are
approximately 200-400 liters/second—roughly two billion gallons per year.
Conclusions
As a result of more than three years of field studies and our interpretation of
these and other data from the literature, we conclude that there are active
inputs of hydrothermal fluids into the bottom of Crater Lake. The dissolved
materials associated with these thermally and chemically enriched fluids,
coupled with the overall hydrologic balance, control the observed chemical
composition of the lake. Because the hydrothermal input dominates the flux of
most dissolved chemicals into Crater Lake, the hydrothermal process is highly
significant. Furthermore, the geothermal inputs have a direct effect on the
density structure of the deep lake, and therefore can profoundly affect the rate
of heat transport and the redistribution of dissolved salts and nutrients within
the body of the lake.
Primary
funding for this project was provided by the U.S. National Park Service (Pacific
NW Region, Cooperative Agreement # 9000-3-0003, subagreement #7). Additional
support was provided by the OSU College of Oceanography, the OSU Foundation, the
US Geological Survey, and the National Geographic Society. Other contributions
to this work were made by: H. Phinney, D. McIntire, G. Larson, M. Buktenica,
Oregon State University; C. R. Bacon, C. H. Nelson, J. H. Barber, Jr., U.S.G.S.;
D. Karl, U. Hawaii; J. Lupton, NOAA PMEL; M. Watwood, C. Dahm, U. of New Mexico;
A. Soutar, R. Weiss, U. C. San Diego; C. G. Wheat, MBAR
