Geotimes Logo NEWS NOTES  June 1999 

by Kristina Bartlett and Devra Wexler

Platinum in the hills
What tangled networks they weave …
Instrumenting plate boundaries
Dry lake sediments stir up old theory
USGS to manage Landsat missions

Platinum in the hills

Sixty million years ago, amidst huge volcanic eruptions and the break-up of Eurasia and North America, the North Atlantic Ocean opened. Increased mantle plume activity caused increased partial melting, which in turn caused the formation of a large flood basalt province. The increased degrees of melting during the opening of the Atlantic also resulted in an enrichment of platinum group elements, or PGEs, in the magma. Recent research from geologists at Camborne School of Mines (Exeter University, Cornwall, United Kingdom) reveals that the Scottish Tertiary Hebridean Volcanic Province, covering much of northwest Scotland and Northern Ireland — in particular the Scots Isle of Rum — holds a rich legacy of platinum deposits.

The deposits aren’t big enough to mine — one can be traced for only five meters and another for two kilometers — and the mineral grains themselves are generally less than two microns in diameter. While not of economic importance, however, the deposits are a scientific gold mine.

Rum, a small island off the western coast of Scotland, is already recognized as a national nature reserve. The discovery of platinum enhances the significance of the island for geological research, says Greg Mudge, West Highland area manager for the Scottish Natural Heritage (SNH), which owns Rum. He adds that SNH has followed the recent news of the platinum deposits with great interest.

“The natural interests of Rum, particularly its geology, have attracted scientists from around the world. We are keen to facilitate appropriate research work within national nature reserves and Rum is in the forefront in this respect. We anticipate that this [new information] will lead to further important discoveries which will help our understanding of aspects of the earth’s history.”

Indeed, the scientists say the same thing.

Dr. Duncan Pirrie, senior lecturer at Camborne School of Mines, discovered the PGE deposits while studying rocks on Rum with his students. He believes that the find opens up the possibility that platinum could be discovered elsewhere in the area.

“The interesting thing is that the rocks on Rum are very similar to those in northeast Greenland and were formed at roughly the same time,” Pirrie says. “The implication is that if it is on Rum and in Greenland [where rocks have yielded gold and palladium] there may be other deposits in northwest Scotland.”

Pirrie and his colleagues, A.R. Butcher (CSIRO Minerals, Brisbane, Australia), H.M. Pritchard (Cardiff University), and P. Fisher (Cardiff), published a paper in the Feb. 12, 1999, issue of the Journal of the Geological Society. In it, they detail the platinum-group minerals they discovered on Rum and discuss the mechanisms that possibly caused the mineralization.

“The mineralization on Rum is due to: 1) initial concentration in the parental magma linked to a high degree of melting associated with mantle plume activity, and 2) localized concentration due to magma mixing and subsequent crystallization within an open-system magma chamber,” the authors state. They show that several areas of Rum contain abundant platinum-group minerals — in particular, one contact horizon (Unit 7-Unit 8) in the Eastern Layered Series (ELS) and the Transitional Member of the Western Layered Series (see diagram and map).

Within the Eastern Layered Series, which is located in the southeast region of Rum, the chromite-rich layers of the Unit 7-Unit 8 contact can be traced along strike for at least two kilometers. The minerals in the laminae are mostly between 0.2 and 2 microns, although some have diameters up to 10 microns. Generally anhedral and inequant, these platinum-group minerals occur at grain boundaries or enclosed within other minerals, such as chalcopyrite, magnetite, and plagioclase. A single thin section (up to 3 square centimeters) of chromitite could contain more than 50 PGE grains.

In the Western Layered Series, near Harris Bay on Rum’s southwest coast, the PGEs occurred within a 1-2 millimeter thick chromitite lamina of the Transitional Member, trace-able for approximately five meters. The platinum-group minerals tended to be much smaller than those in the ELS — usually less than 1 micron — and were also anhedral and inequant. They also occurred at grain boundaries and within host minerals such as clinopyroxene and millerite. One thin section of the laminae hosted more than 40 platinum-group minerals. Dr. Butcher is now utilizing QemSCAN, a fully automated state-of-the-art scanning electron microscopic system, to provide details
of the wider distribution of the platinum-group minerals within the Rum intrusion.

The rocks containing the minerals, says Dr. Pirrie, are very well exposed, and therefore easy to study. They aren’t viable to mine, but they will provide an important addition to geological research on Rum.

The PGE enrichment is a good example of what increased partial melting associated with plume activity can do. The Rum magma chamber is even more important because it shows a classic open-system style of crystallization, similar to that of the Bushveld Complex of South Africa, which does hold economically significant concentrations of platinum-group minerals.

— D.W.

What tangled networks they weave …

Any airplane passenger can observe the meandering of rivers. It’s easy to see the Mississippi River carving a winding path into the landscape. Go higher into Earth’s atmosphere, where numerous Earth-observing satellites photo-graph the planet’s surface, and you see the networks (often resembling the branching patterns of trees or fern leaves) that streams, rivers, and tributaries carve as they erode Earth’s surface and flow to the ocean. Are these paths random, or do simple physical concepts govern them?

In Fractal River Basins: Chance and Self-Organization, published in 1997, authors Ignacio Rodríguez-Iturbe and Andrea Rinaldo, (currently a visiting professor at the Massachusetts Institute of Technology (MIT)), show how river networks, the evolution of river basins, and landscape formation can be analyzed with fractal geometry. As it applies to river networks, fractal geometry dictates that the pattern created by a group of small streams over a few miles essentially resembles the pattern a group of streams and rivers create on a continental scale. The smaller scale resembles the larger scale.

“These are dendritic networks, and they can be quantified by scaling laws. Scaling laws can be defined by numbers,” says Daniel Rothman, a professor in MIT’s Department of Earth, Atmospheric, and Planetary Sciences. Rothman leads a team of researchers who aim to apply physical and mathematical concepts to understand river networks and landscape formation. They presented some of their ongoing research during the centennial meeting of the American Physical Society, which took place Mar. 20-26 in Atlanta, Ga.

Geomorphologists don’t have basic equations governing water paths and erosion patterns, Rothman says. “Why do river networks have the dendritic form they do, and what does this tell you about the geologic form of the landscape?”

The team is currently focusing on theoretical physics. They combine analyses of digital elevation maps and computer simulations to quantify how closely and how often basic physical concepts apply to the forms river networks take. They plan eventually to work with geologists and geomorphologists, who could use the theory to understand how river meandering and erosion form certain landscapes; or, conversely, to judge to what degree certain landscape features (such as fault lines) determine the forms of river networks.

For example, one concept of mathematics and physics that applies to river meander is the random walk. Rothman illustrates random walk by describing a drunk man who wants to cross a street. The man takes a step to the left, falls, gets up, and then takes a step either left, right, straight, or backwards. His choice is random and independent of the step he took before. Overall, though, he tends to move toward the other side of the street, even if his path is winding and long.

But, Rothman says, what if the drunk man remembers with each step the general direction he took with the last step? He might then step left twice, because he remembered stepping left before. What then determines his path is not just randomness, but also memory. Could a similar, fundamental mechanism also determine the paths of flowing water?

“What we’re trying to find is, what’s the simplest thing you can think of that might give rise to these sort of structures?” says Peter Dodds, a graduate student in mathematics and a member of Rothman’s team (along with graduate student Kelvin Chan and postdoctoral fellow Romualdo Pastor-Satorras). The most likely culprit so far, they say, is the simple concept of inertia —  in this case, the tendency of water to continue flowing in its original path, no matter what landscape obstacles it encounters.

If two streams or rivers intersect, they will flow away as one stream or river in a new path. Is the new path random, or dictated by the characteristics of one of the rivers? If the Mississippi River encounters a tiny stream, the Mississippi will have more inertia than the stream. In that case, inertia will dictate the resulting path. But if two similarly sized streams encounter each other, one won’t have much more inertia than the other and randomness tends to determine the resulting path more than inertia does.

So far, Dodds says, their observations support the idea that inertia would be a significant factor. Their models tend to match what they observe from digital elevation maps. “We’re still left with the question, is this what really happens in the real world, or is there another mechanism? … It’s still a developing area.”

Such a quantification could help geoscientists. For example, a quantitative model could make it easier for planetary geologists to determine if the canyons on the surface of Mars were indeed carved by water.

Rothman says it will soon be easier to pinpoint what fundamental force or phenomenon
determines river networks. The National Aeronautics and Space Administration will release digital topography maps from the Shuttle Radar Topography Mission, a network of satellites mapping Earth’s surface at horizontal resolutions of 30 meters. The satellites should launch in May 2000 and will collect three-dimensional measurements of almost 80 percent of Earth’s surface over 11 days with an accuracy of better than 53 feet.

“Having access to continental-scale forms and being able to access every river in the world — it could keep computers busy for a long time,” Rothman says.

— K.B.

Instrumenting plate boundaries

Plate boundary deformation zones account for 15 percent of Earth’s total surface and cover 30 percent of the continental landmass. Scientists representing many specialized disciplines in the earth sciences, and using a variety of techniques, study these zones. They work mostly in isolation. Paul Silver, a geophysicist at the Department of Terrestrial Magnetism (DTM), Carnegie In-stitution of Washington, believes that many of the approaches and techniques used to study plate boundaries need to be more closely coordinated. “It is very important to study these zones,” Silver says, “to understand the forces that drive plate tectonics and the connection between plate tectonics and earthquakes. But we will benefit only if there is integration of these scientific studies.” Silver is therefore proposing a new initiative for studying plate boundary deformation in a way that integrates the various methods of studying these important zones. He presented his ideas in March at a DTM seminar. “What forces are responsible for plate boundary deformation?” Silver asked. He discussed three regions whose characteristics shed light on the question.

Building Tibet

Silver first presented the case of the ongoing deformation at the Tibetan plateau and how he studies the forces driving the plates. The present-day surface deformation can be inferred from both the slip on Quaternary faults and the fault mechanisms of earthquakes that occur in the plateau. The deformation of the deeper mantle can be seen by studying seismic anisotropy, particularly with shear-wave splitting. “We can obtain two parameters from shear-wave splitting — the fast polarization direction and the delay time. These are related to the orientation of deformation and the thickness of the deformed region in the mantle.” This mantle deformation can then be compared to surface deformation. The mantle beneath nearly all continents is anisotropic, but it is particularly so beneath Tibet. This means that the region’s mantle has been strongly deformed.

What is most striking about Tibet is that the crust and mantle are deforming coherently to 200 to 300 kilometers depth. Normally, the collisional deformation of the crust and mantle portions of tectonic plates thickens the plates. Since plates are colder than the mantle below, a region with a thickened plate should be colder than average. But the Tibetan crust and upper mantle are unusually hot, as indicated by active volcanism, excess uplift, and low seismic velocities in the shallow mantle.

Silver contends that the coherent deformation of the plate, and the high crust and mantle temperatures, are most easily explained if the deformation itself is the source of the heat. This phenomenon, known as shear heating or viscous dissipation, is an efficient way to generate heat. Accounting for this excess heat requires that the stresses causing the deformation be several kilobars high — and consequently that the stresses driving the plates involved in the collision be higher than commonly thought.

Events in the Andes

In contrast to the Himalayas, the Andes are formed at a convergent (rather than collisional) boundary, where one plate slips beneath another. This process does not always produce mountains, however, as the absence of such structures associated with other convergent margins around the rest of the Pacific Rim demonstrates. “So what causes mountain building in the Andes?” Silver asked seminar participants. He proposed that “Cordilleran events may have distant causes.”

As in the case of the Andes, the descending slab serves as a “wall” that resists the westward motion of the South American plate. The rate of this motion should determine how fast the Andes grow. If this is true, changes in Andean tectonic activity should correspond to changes in the velocity of the South American plate.

Silver noted that Andean tectonic activity increased intensely starting about 30 million years ago. He hypothesized that there would be a corresponding increase in the motion of the South American plate. He calculated the motion of the South American plate in a hot-spot reference, and showed that a westward speedup in South American motion did begin about 30 million years ago, as he hypothesized.

But what ultimately produced this increase in velocity? Silver demonstrated that the speedup occurred at the same time that the African plate slowed down as a result of its collision with Eurasia. He suggested that these two events were causally related, that because of the coupling of the South American and African plates to mantle circulation, the velocities of the two plates interact in what he terms a “flow-coupled plate interaction.” The speedup in tectonic activity predicted by Silver’s model would correspond to about 1 centimeter a year of shortening across the Andes, if all of the speedup were converted to shortening. Seminar participants argued, suggesting the buoyancy of the Cordilleran slab, the dipping of a shallow slab, and a fold thrust as alternative explanations. Silver agreed that these were valid possibilities, and again stressed the need for a cohesive, integrated study. But he noted the power of modern geodesy by saying that, “GPS [Global Positioning System] work in the central Andes did indeed show a present-day rate of shortening of about 1 centimeter a year, comparable to the model predictions” (see Geotimes, April 1998). He added, “Let’s increase our focus on plate boundary deformation, combining GPS with strainmeters and INSAR [Interferometric Synthetic Aperture Radar]. We need a boost in instrumentation.”

Earthquakes in California

Silver believes that the best place to begin studying plate boundary deformation with this combination of instruments and techniques is the western United States, along the zones in California. “Let’s completely instrument the plate boundary,” he proposes.

With each earthquake that occurs in California — Landers, Parkfield, Kettleman Hills — geologists have used the available instruments to measure the motion of the San Andreas. But the instruments aren’t always in the best locations. After Kettleman Hills in 1985, geologists tried to use strainmeters to look for precursors to the quake, but the instruments were over 30 kilometers from the epicenter — too far to determine if preseismic signals were related to the earthquakes or were more local. “Right now we have missing data,” says Silver. “We can’t predict where to place instruments. If we had more comprehensive geodetic coverage, we could resolve these issues.”

Yet, some locations are “well instrumented.” After the Landers earthquake, strainmeters and GPS tools were available to look at changes in seismicity from postseismic strain-pulse migration. And the Parkfield area had concentrations of instruments that showed transients. Silver reported, “Every dataset — from strainmeters, creepmeters, two-color Geodimeters, rainfall measurements, and seismic movement —suggested the existence of a multi-year strain transient that included a speedup of the San Andreas fault in 1993 and the accompanying increase in seismic activity.”

Silver stressed that geologists need to work together to understand how plate motions connect to earthquakes. He believes that the comprehensive study of plate boundary deformation, through the deployment of a plate boundary observatory consisting of a large number of geodetic instruments, may provide the key to understanding earthquake occurrence and active tectonics as well. “The San Andreas fault is the best area for this,” Silver said. “It’s close by, it’s a small environmental area, and it’s on land (whereas most plate boundaries are underwater).”

He adds, “The future ought to bring an increased focus on plate boundary deformation, one that requires an integrated approach.” While the cost to support this proposal may seem high, “the increased knowledge and understanding will be worth it.”

 — D.W.

Dry lake sediments stir up old theory

More than 40 centuries ago, the Harappan-Indus culture (2600-2000 B.C.E.) was evolving into one of the great civilizations of South Asia. Gurdip Singh, a palynologist who did extensive work in India in the 1960s and ’70s, suggested that the civilization arose during a period of moist, temperate climatic conditions — a concept known as the “climate-culture hypothesis.” A recent study by Yehouda Enzel, Lisa Ely, S. Mishra, R. Ramesh, R. Amit, B. Lazar, S. N. Rajaguru, V. R. Baker, and A. Sandler in the Apr. 2, 1999, issue of Science refutes this theory. These researchers present sedimentological and chemical evidence suggesting that temperate conditions had already ceased — and in fact lake dessication had occurred — before the birth of the Harappan-Indus culture.

In 1991, Enzel, Ely, and Baker, then colleagues at the University of Arizona, became interested in studying paleoflood hydrology in India. They drafted a proposal for the National Science Foundation, and received a grant in 1994. “During the writing of the proposal,” says Enzel, “we figured out that we rely on the seminal pollen research by Gurdip Singh [as do] many other researchers that needed information from this part of the world.”

Enzel, Ely, and Mishra, who had joined the group, asked Prof. Rajaguru, a colleague of Singh and then the director of Deccan College (Pune, India), to guide them through several of Singh’s sites. “When we were in Lunkaransar, the premier site of Singh’s pollen research, we noticed the extremely well-laminated deposits and decided to test [a] few ideas,” says Enzel. “Rajaguru’s excitement, knowledge, and guidance helped us to pursue this research.”

With results from these first tests, Enzel, Ely, and Mishra enlisted colleagues from India, Israel, and the United States to join the project. The fieldwork was intense, says Enzel. “The efforts and dedication involved in extracting the samples in northwestern India are tremendous. We could not have done it without the full collaboration of all the researchers and the research institutes [Deccan College, Hebrew University, and India’s Physical Research Laboratory] involved.”

Lunkaransar, the team’s study area, is an ephemeral lake basin in the Thar Desert, India. Today, the influx of water into the basin is entirely dictated by meteoric and groundwater influences. Since no streams carry water into the basin, incoming sediments are derived solely from eolian sand from local dunes and eolian clay and silt dust. But 10,000 years ago, the environment was far more mesic, and far more hospitable.

While analyzing thin sections from Lunkaransar lake sediments, Amit noticed the evidence of extensive microbial mat communities. The algae, present in the sediment thin sections, also dominated the ð13C record. Boaz Lazar, Jonathan Erez, and Naama Gazit-Yarri of Hebrew University discovered that the ð13C composition in the shallow lake environment is controlled by water depths. As a result, the researchers were able to interpret the algal ð13C influences from Lunkaransar in terms of water depth.

The algal mats thrived in the lake during the early Holocene, when the water levels were generally shallow, although unstable. The algal communities did not incorporate lighter C values (meaning more 12C) until a period of high lake levels around 6,300 radiocarbon years ago (the deeper waters allowed the incorporation of 12C into the growing mats). When lake levels fell again around 800 radiocarbon years later, ð13C values increased. By 3500 B.C.E., long before the rise of the Harappan-Indus civilization, the lake had undergone complete dessication. It has remained dry through the present.

In addition to stable isotope evidence from organic material marking the environmental shift, the sediments from Lunkaransar indicate the climate change as well. During periods of increased aridity, the researchers noticed an increase in >125 µm eolian sand grains and subangular quartz grains, both derived from surrounding dunes. Since no similar input of eolian material
was observed earlier in the Holocene, Enzel and others say that the influx suggests a major environmental change into the arid environment of today.

Enzel and his colleagues have therefore concluded that the drying of the region occurred before the rise of the Early and Mature Harappan phases of the Indus civilization (2600-2000 B.C.) — which would suggest that the climate-culture hypothesis is invalid. Although this conclusion was reached from research at only one site, further paleoenvironmental studies in northwestern India and Pakistan will soon accompany the Lunkaransar work. Enzel, Ely, and others have analyzed several other lakes in the region, and the results may be compiled as early as next year.

Joshua A. Chamot
Department of Geology, University of Tennessee — Knoxville

USGS to manage Landsat missions

The Landsat satellites have been photographing Earth’s surface since 1972. After a long delay, the National Aeronautics and Space Administration (NASA) and the U.S. Geological Survey (USGS) launched the latest Landsat satellite, Landsat 7, on April 15. This satellite brings new technology into orbit and starts a new kind of management for the Landsat program. By 2001, the U.S. Geological Survey will be the primary manager of Landsat.

Many satellites orbit Earth and photograph its surface, and several commercial satellites image the planet’s surface at much higher resolutions than what the Landsat satellites can achieve, says R.J. Thompson, Landsat 7 program manager for USGS. But Landsat is the United States’ oldest land-surface observing system. Its age is its strength, says Bonnie McGregor, associate director of USGS.

While other systems pinpoint specific regions of Earth at specific times, Landsat provides a pool of raw data that spans the entire globe and stretches back continuously for 26 years. The satellites record images of continental surfaces in the visible, near-infrared, short-wave, and thermal infrared regions of the electromagnetic spectrum. The satellites relay these images to international ground stations, and the images are stored at the EROS (Earth Resources Observation Systems) Data Center in Sioux Falls, S.D. Researchers or commercial users can buy images of particular regions. The age and continuity of the images make them a crucial archive for monitoring the natural and anthropogenic changes of Earth’s surface, McGregor says.

The Landsat satellites have been consistently collecting data, despite some problems. Landsat 5, launched in 1984, is still returning images. Landsat 6 was destroyed in 1993 because of rocket-firing problems. Technical difficulties delayed Landsat 7’s launch, originally scheduled for July 1998. McGregor says launching Landsat 7 in April became crucial to retaining the continuity of the Landsat database.

The management of Landsat on the ground hasn’t been as constant. The missions started under NASA in 1972. Under legislation enacted by the Carter Administration, Landsat was transferred to the commercial sector in 1985. But the company that managed it, the Earth Observing Satellite Co., said launching and managing a third satellite wouldn’t be profitable, Thompson says. Legislation passed in 1992 transferred Landsat back to the federal government, first under NOAA’s management and then under management of NASA and the Department of Defense (DOD). NASA became the lead agency in 1994 when DOD withdrew.

With the launching of Landsat 7, NASA and USGS will manage the Landsat mission together under a dual-agency contract. Starting in 2001, Thompson says, USGS will take over the daily operations of Landsat 7, managing the satellite and its data from the EROS Data Center and from a Missions Operations Center at NASA’s Goddard Space Flight Center in Greenbelt, Md.

“The agreement that’s emerging between [USGS] and NASA is that the Department of the Interior and USGS would evolve toward a land-resource specialty responsible for land-oriented satellites,” Thompson says. The National Oceanic and Atmospheric Administration (NOAA) would be responsible for monitoring the oceans and the atmosphere, and NASA would be responsible for building new technologies and launching the satellites. “That’s what’s been happening now for many years.”

Landsat 7 will fly in formation with Terra, the first satellite in NASA’s Earth Observing System (EOS). Terra should be launched in July, and will meas-ure physical and radiative properties of clouds; air-land and air-sea exchanges of energy, carbon, and water; trace gases; and volcanoes.

Part of the Survey’s new direction for Landsat is working with the commercial sector, Thompson says. Landsat will supply raw data on a global scale. Individual companies can take that raw data and make it more detailed or cater it to specific needs (such as installing pipelines or building transportation systems). The private sector would then sell these value-added images, Thompson says. In such a partnership, the government supplies continuous raw data, and the private sector distributes that data.

Landsat 7 will contribute to this mission because it can collect more data than the other Landsat satellites. A solid-state data recorder can store 100 images, allowing the satellite to update a complete, global view of Earth’s surface seasonally, or about four times every year. Landsat 7 can also download its data to ground stations and the EROS Data Center at a faster rate.

More details about the Landsat 7 mission are on the web at <>.

— K.B.


Geotimes Home | AGI Home | Information Services | Geoscience Education | Public Policy | Programs | Publications | Careers

© 2022 American Geological Institute. All rights reserved. Any copying, redistribution or retransmission of any of the contents of this service without the express written consent of the American Geological Institute is expressly prohibited. For all electronic copyright requests, visit: