Annual Report / Draft 01-17-02

Thank you for your interest in our annual report. In the pages that follow you will find a summary of our research activities for calendar year 2001. This includes major highlights in each of the five main areas in which we work (Geomagnetism, Crustal Deformation, Topography and Surface Change, Planetary Geology and Geophysics, and Orbital-Rotational-Climatic Interaction), as well as publications submitted, accepted or published in 2001. We also show major presentations given by members of our Branch both at national and international meetings and elsewhere, and include a list of the science seminars we held during the year.

Our group consists of 11 civil servants, and about the same number of full-time contractors and visiting scientists. Last year we had about a dozen students and other visitors spend time working with us. Despite our relatively small size, we are involved in a wide range research and mission support activities. We have personnel contributing to the currently operating Mars Global Surveyor, and the Oersted and CHAMP magnetic satellites, and providing support to the future ICESat mission. Although not discussed in this report, we also contributed to a number of advance instrument and mission studies for possible future flight opportunities in both the Earth Science and Space Science Enterprises. Our personnel, in addition to their highly successful individual research activities, are heavily involved in proposal and paper reviews as well as mission design and readiness reviews in both areas as well.

A pictorial overview of our group is shown on the following page. Moving your computer mouse over a picture will provide the name and clicking on the picture will link you to that person’s individual home page or other contact information.

We hope you enjoy reading this annual description of our activities, and welcome any comments you might have with regard either to the content or the format of this presentation. Please feel free to contact me or any of the individuals in our group at any time.

--Herb Frey

Head, Geodynamics Branch

Who we are - (details)

GEODYNAMICS BRANCH RESEARCH

INTRODUCTION

Research in the Geodynamics Branch covers a wide range of subjects in the broad disciplines of geophysics, geology, geodesy and geodynamics for both the Earth and solid planetary bodies, especially Mars. Surface and satellite data, models derived from these, and other observational and theoretical information are used to help improve our understanding of the evolution of the Earth’s core, mantle and crust, and the character and evolution of planetary surfaces. Geodynamics includes studies of the processes which operate or have operated to produce the observed present-day state and motions. Of particular interest are core fluid motions and how they relate to changes in the Earth’s magnetic field; motions of the Earth’s crust and the relationship with earthquake hazard, especially in areas of active subduction; vertical rebound of paleolakes and the constraints it provides on lower crustal and upper mantle properties; long term orbital-rotational evolution and its relationship to long term climate change; magnetic properties of the Earth’s crust and, most recently, the nature of the sources of magnetic anomalies not just on the Earth but on Mars as well; topographic characterization of the surface of the Earth and Mars, in the former case to understand landforms associated with active faulting, in the later case to understand volcanic and tectonic structures, the origin of the fundamental crustal topographic dichotomy, and how large scale impacts affect early crustal evolution.

Brief descriptions of each of the major areas of work and the Geodynamics personnel involved in them are provided below. Below each summary, individual significant highlights of the past year are provided, along with contact information for the relevant author. Publications, major presentations, and other information are shown in subsequent sections.

Geomagnetism (Nazarova, Purucker, Sabaka, Taylor, Voorhies)

This area includes development of comprehensive models of the Earth’s total magnetic field, involving simultaneous estimation and parameterization of the core, crust and external (magnetospheric and ionospheric) components. The latest (Phase 3) version of this model has been accepted for publication and is being used to help study new satellite magnetic field data (see below). Main (core) field models are produced from satellite observations, including recent data from the Oersted and CHAMP missions. These descriptions include secular variations, and are used as constraints on models of the core dynamo. The lithospheric field is studied separately to infer properties of the Earth’s crust. As described below, such studies have provided new insight about the edge of the North American craton. Recently effort has been devoted to studying the crustal magnetic anomalies of Mars discovered by the Mars Global Surveyor spacecraft, using techniques originally developed for analysis of terrestrial Magsat data. Spectral studies of both Earth and Mars magnetic fields have been done, and are reported below (see Planetary Geology and Geophysics). Scientific expertise is also applied to calibration of in-flight magnetometers on current missions, and towards development of possible new terrestrial and martian magnetic gradient missions, using both satellite and balloon platforms.

Comprehensive Models of the Earth’s Magnetic Field

A single measurement of the Earth's magnetic field actually measures a combination of fields originating from different sources. Most of these sources are distinct concentrations of electrical currents which flow either inside the Earth or within several tens of Earth radii of its surface. The major sources contributing to the total magnetic field measured at the Earth's surface up through satellite altitude are the core, crust, ionosphere, magnetosphere, currents coupling the ionosphere and magnetosphere, and currents induced in the outer layers of the Earth's surface due to time-varying external (ionospheric or magnetospheric) fields. It is important to describe the extent and behavior of these current systems as well as to understand the processes by which they are established. Through the years, scientists have produced from available surface and satellite data increasingly sophisticated mathematical models of the spatial and temporal organization of these individual fields. The most accurate representation to date comes from a series of models developed at GSFC by a team of international investigators. Because these models include the effects of all the previously described sources, they are called "comprehensive models" or CMs. A paper describing a recent version of the GSFC comprehensive model was recently accepted for publication and the model itself is available for (electronic) distribution to interested investigators.

The key to the CM approach is understanding that the various magnetic fields in the near-Earth environment change or fluctuate on scales that overlap. This is true both in space and time. For instance, portions of the field from the Earth's core can change on scales of a decade, while on the same time frame fields in the crust induced by the magnetosphere will fluctuate as the magnetosphere fluctuates with the 11 year sunspot cycle. Thus, it is very difficult to separate one field from another by simply arguing that a particular field changes at a given rate in space or time. The CMs are developed by taking into account these "correlations" and overlaps. The goal is to produce an optimal model which fits all the data and any other physical information that might be included.

The most advanced CM to date is CM3d. It is derived from data collected by the NASA POGO and Magsat satellites and the recent Danish Oersted satellite, and also includes data from a global distribution of permanent ground-based magnetic observatories. The model includes about 20,000 parameters and spans the years 1960-2000, and represents the magnetically quiet field in the near-Earth environment (i.e., within about 2,000 km of Earth's surface). An illustration of its power in resolving individual components of the total field is seen in the Figure for the particular pass of Magsat data shown below.

The top panel shows the residuals (black boxes) with respect to the long wavelength main (core) and crustal fields. The predicted magnetospheric field is shown as the red line. The next panel shows residuals with respect to main and magnetospheric fields, with the predicted ionospheric field in red. The bottom panel shows residuals with respect to main, magnetospheric and ionospheric fields, with the predicted medium to short wavelength crustal field in red. Clearly the model does a good job accounting for the different components of the near-Earth filed measured along this pass. Note the large excursion in the crustal field at 40N latitude in the bottom panel. This is the well-known Kursk magnetic anomaly, which is the strongest of the terrestrial lithospheric anomalies and which has been studied by Taylor et al. (see article in this section).

Modeling of crustal anomalies has become a "Holy Grail" for much of the magnetics community. Proper separation of the crustal signal from that of the ionosphere is difficult because of poor surface data coverage and overlapping spatial scales. Most methods employed to date use along-track filters to remove ionospheric signal from satellite passes. However, this can remove and usually does remove crustal signal along the pass in the north-south directions (most magnetic field satellites are in near-polar orbits). The CM avoids this ad hoc approach and preserves more of the crustal field structure by parameterization of the ionospheric contribution. The crustal anomaly maps derived from the CM show many N-S anomalies associated with known large-scale structure. Examples can be seen in Figure 2, which is a global map of the radial component of the crustal (lithospheric) field at 400 km altitude. This model resolves several north-south trending features far better than previous models, including the Andean subduction zone and mountains and the mid-ocean spreading ridge in the south Atlantic.

Contact: Terence J. Sabaka, Geodynamics Branch, sabaka@geomag.gsfc.nasa.gov

New Satellite Magnetometer Observations Help Answer Some Old Questions

The recent launch of three new magnetometer satellites (the Danish Oersted, Argentina’s SAC-C, and Germany’s CHAMP mission, all of which have NASA involvement) is providing a renewed opportunity to investigate ways to separate temporal and spatial variations of the earth's magnetic field. This separation is difficult because any one measurement of the magnetic field is actually an observation of the combined contributions from the internal main (core) magnetic field, from the ionosphere and magnetosphere, and from sources within the crust. A comprehensive modeling approach, described elsewhere in this section, is one approach that has proven useful in separating the different components of the total field. But having multiple satellites making measurements at the same time from different orbital locations also helps in the separation of time-variable contributions from the longer term, more permanent components of the field. One product of this improved analysis capability is a much better representation of the lithospheric or crustal fields. This has, in particular, led to a better definition of the boundaries of the continental cratons, as shown in the Figure below.

The Figure shows a vertically-integrated model of the induced and remanent magnetization of the Earth’s crust that explains the new satellite magnetic field observations. The model also incorporates information from surface magnetic field observatories in addition to the satellite data, and uses seismically-derived estimates of the thickness of the crust. Areas of negative magnetization are due to orientations in directions oblique and opposite to that of the present earth's field. Positive (induced) magnetizations generally align with the present day field.

The model shows in color long-wavelength magnetization features, which are dominated by the continent-ocean crustal thickness and composition contrast (e.g., compare North America and the Atlantic Ocean). Short-wavelength features shown by shaded relief are mostly sea-floor spreading anomalies. In a paper in Geophysical Research Letters, Purucker and colleagues suggest that the very large total field magnetic anomalies centered over Kentucky and the south-central United States are the manifestations of the magnetic edges of the continental craton along the southern boundary of North America. That is, their strong magnetization is due to the difference in magnetic properties and thickness between continental and oceanic crust.

The current generation of satellites include several which, in the future, will dip to lower altitudes, which will improve the spatial resolution of anomalies such as those shown in Figure 1. This should allow scientists to address even more localized questions, such as the age and heat flow of the central Greenland crust (see Figure 1), now the site of rapid ice flow (Fahnestock et al., Science, Dec. 14, 2001). While Greenland is typically classified within the same cratonic zone as the Canadian shield, ice cover masks all but the borders of Greenland. Recent seismic surface-wave observations and preliminary analysis of the current low-altitude satellite magnetic field observations suggest that only the northern and southern edges of Greenland are underlain by ancient, cold crust. Future lower altitude magnetic field observations should help refine this model of the Greenland crust.

Contact: Michael Purucker, Geodynamics Branch, purucker@geomag.gsfc.nasa.gov

Satellite Altitude Magnetic Anomaly Study of the Kiruna, Sweden Iron-Ore Deposit

Previous studies of satellite altitude (~400 km) crustal magnetic data, mainly from the Magsat mission, have revealed several significant (>15 nT) anomalies over known ore bodies in Russia , the United States and the Central African Republic. Additional satellite data is now available from the Danish Oersted satellite, and is becoming available from the newly launched German CHAMP mission. We recently investigated another of the world's largest known iron ore deposits in the area around Kiruna, Sweden, by looking at the relationship between satellite-altitude magnetic and aeromagnetic data of northern Sweden and the regional geology/tectonics of the Kiruna iron ore district. Figure 1 compares the satellite-elevation crustal anomaly field from Magsat with aeromagnetic data from the NORDKALOTT (Unified study by geological surveys of Norway, Sweden and Finland) upward continued to the same altitude.

In the Magsat map (Figure 1A) the marine regions surrounding Norway are represented by predominately negative anomalies. The land area is dominated by a positive bulls-eye anomaly (the Kiruna anomaly) with a maximum value (>9 nT) centered 67 deg. N latitude and 20 deg. E longitudes near the city of Kiruna, Sweden. Extensive aeromagnetic surveying has been done across this region and these extremely detailed magnetic data reveal a concentration of high near-surface anomaly values (>6000 nT). The Kiruna region is the center of a complex nexus of magnetic highs. The mathematically upward-continued aeromagnetic data (Figure 1B) shows these NORDKALOTT data at 400 km altitude have virtually the same anomaly pattern as the Magsat map. It is particularly noteworthy that the positive magnetic anomalies in both the satellite and upward continued aeromagnetic data are coincident and have the same value. This means that these data are both sensing the same anomalous magnetic body and that the Earth’s main field has been correctly and completely removed from both data sets.

We made an estimate of the depth-to-source of the magnetic body from an analysis of these satellite magnetic anomaly data. The method used was originally developed by D. Ravat and P. Taylor and is based on the property that the rate of decrease of the anomaly amplitude with distance from the source is a known function. This method, which is best suited to nearly circular, isolated, large amplitude anomalies, yielded a depth of between 20 and 30 km for the source of the anomalous mass responsible for the Kiruna anomaly. Since the thickness of the Earth's crust in the Kiruna area is known to be less than 46 km this would place the magnetic source body in the mid to lower crust.

The major fracture zone of this region, based on geologic mapping, are superimposed on the Magsat anomaly map (Figure 1A). The intersection of the major fracture zones lies very near the center of the Kiruna anomaly. It is, therefore, interesting to speculate that the anomalous magnetic mass causing the satellite-altitude anomaly is not only deep-seated but may also be related to the fracture zone pattern. It has been suggested by others that major lineaments play a role in the emplacement of the ore deposits of Northern Sweden and that major mineralization of the upper crust is controlled or localized by large and deep fractures in the crust. The fracture zones could provide migration routes allowing lower crustal material to be emplaced in the upper crust.

Contact: Patrick Taylor, Geodynamics Branch, ptaylor@ltpmail.gsfc.nasa.gov

Crustal Deformation (Bills, Cohen, Sauber)

GPS field measurements of point positions taken over several years provide crustal motions that are used to investigate the subduction zone process, especially where large earthquakes are a significant hazard. Of particular interest are shallow dip subduction zones where both earthquakes can be great and mountain building extensive. Active folds and faults on Kodiak Island, Alaska were identified using Landsat imagery, DTED and kinematic GPS elevation data. Anomalous motions not easily explained by either current plate motion or as long term post-seismic effects have been observed in the Kenai Peninsula area of Alaska. Crustal motion data has been used along with visco-elastic modeling to estimate the thicknesses of the converging Pacific and Australian plates near New Zealand, as described below. Ongoing studies of paleolakes have concentrated on Lake Bonneville and Lake Lahonten in the Western US where unusual upper mantle structure appears to exist. In these lake rebound studies GPS measurements provide the elevation of past shorelines and bio-stratigraphic markers provide a time reference that permits reconstructing the rebound history following loss of the water.

Crustal Deformation and Earthquake Hazard at Kodiak Island, Alaska

The Kodiak Islands are located in the eastern Aleutians of Alaska and lie approximately 140 to 250 km from the Alaska-Aleutian trench. The Pacific plate is subducting beneath the North American plate near Kodiak Island (Figure 1) at a rate of about 57 mm/yr. Since Kodiak Island lies within the southern extent of rupture from the great (M=9.2) 1964 Prince William Sound earthquake, we might not expect a large earthquake to occur in this region in the near future. But there is still a significant earthquake hazard in this region. The faults located above the plate interface could slip in large earthquakes. In fact, geologic work done by a co-investigator (G. Carver, Humbolt State University) shows that there are numerous crustal faults across the Kodiak Islands which have slipped in earthquakes in the last 100,000 years.

To quantify the rate of ongoing crustal strain, we established geodetic sites at a range of distances from the trench, and in addition established some coverage parallel to the trench, on the portion of the island accessible by roads. Changes in the rate and orientation of short-term strain across the northern, populated part of Kodiak Island have been estimated from GPS measurements made by us between 1993 and 2001. This field observation program included, in 1995, 1997, and 1999, an educational outreach program with Kodiak Island High School, funded in part by a GSFC Director’s Discretionary Fund (DDF). Students participated in setting up observing sites, recording data, and analyzing the results. Although most of these GPS observations were made using temporary portable stations, in 1999 a permanent GPS site (KODK) was established (using support from the DDF) that is now part of an international, global GPS network. KODK is run and maintained by Kodiak Island High School Earth Science teacher Eric Linscheid. A description of this collaborative research program with NASA by Linscheid can be viewed at http://www.koc.alaska.edu/nasa/index.html.

Figure 1 shows a shaded relief digital elevation map of the northeastern segment of Kodiak Island. The location of GPS sites in which horizontal velocities have been estimated are shown. Both orientation and magnitude of the horizontal motions are displayed. The location of the Kodiak launch facility is given by ^ and the city of Kodiak by #.

The rate of ongoing crustal deformation across the Island ranges from up to 20 mm/yr along the eastern coast near the Kodiak Launch facility to 7 mm/yr east of the city of Kodiak. Locking of the plate interface at shallow depths is the process exerting the greatest influence on the horizontal, interseismic rate of deformation shown in Figure 1. Most of the deformation can be accounted for by this mechanism. Note in Figure 1 that the orientation of the short-term deformation rates of the sites located near the east coast of Kodiak Island are similar to the plate motion vector, as expected for a locked main thrust zone. The orientation of site vectors west of the fault "KIF", however, are rotated counter clockwise (more westerly). These results suggest that, in addition to eventual thrust faulting on the main plate interface and crustal shortening, some strike-slip motion may occur on faults within the overriding plate.

Contact: Jeanne Sauber, Geodynamics Branch, jeanne@steller.gsfc.nasa.gov

Tectonic Plate Coupling and Elastic Thickness from Space Geodetic Measurements of Crustal Movements

Measurements of crustal movements in seismic zones provide insight into earthquake processes and properties of the Earth’s crust and mantle. Cohen and Darby (2001) recently showed that geodetic measurements of crustal movement, can be used to simultaneously determine the strength of coupling between interacting tectonic plates (a measure of how large an earthquake might occur in the region) and the effective elastic thicknesses of the lithospheric plates (a measure of the depth range over which the earth is mechanically strong enough to support stresses without significant viscous flow).

In the past lithospheric thicknesses have been determined by examining the deformation of the earth in response to long time-scale loading. For example, measurements of the bending of tectonic plates under seamounts (underwater volcanoes) or sedimentary basins have shown that oceanic plates typically have an elastic thickness between 10 and 50 km and continental plates have similar or greater elastic thicknesses. An important aspect of these determinations is that the applied stresses persist for very long times, 1 million to 100 million years.

In the case of the earthquake cycle, however, stress accumulation and release takes place over much shorter time: 100 to 1000 years. It is possible that flow in the higher viscosity portions of the earth may not occur and tectonic plates may have a greater effective elastic thicknesses. Cohen and Darby investigated this possibility. A dense set of GPS-derived crustal velocities were used in combination with an elastic-viscoelastic model of crustal deformation to determine the coupling and elastic thicknesses near the Hikurangi trough at southern North Island, New Zealand, a part of the Australian-Pacific plate boundary region shown below.

At the southern Hikurangi Trough the Pacific Plate approaches the Australian plate with a relative velocity of ~39 mm/yr. The approach is highly oblique, with the angle between the relative plate motion and the normal to the Hikurangi Trough being about 60 degrees. The GPS-derived crustal velocity field for this region, published recently by Darby and Beavan (2001), and a finite element model of the obliquely convergent plate boundary, are shown below.

An important feature of southern Hikurangi region is that, although it is very active seismically on a geological time-scale, there has been no major earthquake since 1855. Thus, any transient response to the last earthquake has long since dissipated and the crustal velocities can be described using a steady-state model whose behavior depends on the plate coupling and plate elastic thicknesses. For given values of plate thickness, the coupling is obtained by mathematical inversion techniques applied to the finite element model and the GPS data. The thicknesses are then systematically varied to find a best fit between observed and predicted velocities (as measured by the reduced chi-squared parameter) as well as the range of model parameters that are statistically consistent with the observations. The Figure (3) below shows the set of reduced chi-squared contours derived by Cohen and Darby, using a seismological constraint that the Young’s modulus of the Australian crust is a factor of two smaller than the Young’s modulus of the mantle and Pacific Plate.

From this Figure and the associated analysis, Cohen and Darby concluded that:

(1) The “best” value for the thickness of the Pacific and Australian plates are, respectively, about 45-65 and 100-125 km (but the range of acceptable values, as indicated by the blue region in Figure 3, is large). Alternatively, smaller thicknesses were obtained when the Young’s modulus contrast was increased above the values derived from seismic velocity data. There are no estimates of long-term plate thicknesses in this region of New Zealand, but estimates of the thickness of the Australian plate at the Wanganui basin (200 km to the northwest) indicate a value of about 15 km. The geodetic results suggest that the apparent elastic thickness in response to earthquake stress is, as suggested above, greater than the thickness associated with long time-scale geologic (plate motion) processes.

(2) The results explain why simple infinite elastic half-space models of interseismic strain accumulation seem to “work”, i.e. provide a reasonable fit to the data, even though the plates do have finite thickness. The continental plate is thick, although not infinitely so and the oceanic plate is stronger than the continental plate. Both of these conditions suppress the surface effects of viscous flow at depth.

(3) The trench normal component of the crustal velocity field is much more diagnostic for discriminating between model parameters than the trench parallel component. This result has implications for planning future ground and space based observations of crustal motions in this kind of tectonic setting.

(4) The coupling coefficient (which can take on values between 0 and 1) is > 0.8 to depths of 20-25 km. This indicates that the plates in this region are currently strongly coupled. The coupling strength at very shallow depths cannot be estimated because there are no geodetic observations over the shallow part of the zone (which lies in the ocean). However, the estimates of coupling at seismogenic depths (where earthquakes are generated) are virtually unaffected by this ambiguity. This is an important conclusion for seismic hazard estimation.

Because some of the aspects of this work were surprising and have potential importance for both earthquake hazard studies and for basic tectonic modeling, efforts are now underway to determine whether the results for New Zealand are replicated elsewhere.

References:

Cohen, S.C., and D.J. Darby, Tectonic plate coupling and elastic thickness derived the inversion of geodetic data using a steady-state viscoelastic model: Application to southern North Island, New Zealand, J. Geophys. Res. (submitted), 2001.

Darby, D.J. and J. Beavan, Evidence from GPS measurement for contemporary interplate coupling on the southern Hikurangi subduction thrust and for partitioning of strain in the upper plate, J. Geophys. Res., 106, 30881-30891, 2001.

Contact: Steven C. Cohen, Geodynamics Branch, scohen@carnoustie.gsfc.nasa.gov

Topography and Surface Change (Lowman, Harding, Sauber, Still, Yates)

Airborne and spaceborne laser altimetry provide the basis for studies of areas where geological hazards exist (such as earthquakes, volcanoes, coastal erosions). In cooperation with USGS and local agencies, an intense study of the Puget Sound area in Washington State has revealed new details of previously unknown faults and better characterization of other landforms (e.g., beach terraces) related to active faulting. Many of these are hidden beneath dense canopy which airborne lidar can penetrate to provide information on both canopy structure and the “bare Earth” structure. Calibration and validation studies for the future laser altimeter instrument GLAS on ICESat have been ongoing and calibration sites in Alaska, Washington State and in the Western United States have been selected to help cover the major kinds of surfaces (mountainous glacial, faulted vegetated, flat dessert) the lidar will measure. Technology development is focusing on imaging lidar capability, such as single photon counting, scanning and swath mapping. Planning is underway for a third Shuttle Laser Altimeter flight to demonstrate this capability.

New versions of a Digital Tectonic Activity Map have been produced, and these have become very popular outreach items on the World Wide Web. Also, as described below, a major new book on the role of satellites in “Exploring Space, Exploring Earth”

Puget Sound Faults and Earthquake Hazards

The densely populated Puget Lowland of western Washington occupies a dynamic geologic setting in the North American plate above the Cascadia subduction zone. Oblique plate convergence along the subduction zone subjects the Lowland to damaging earthquake and volcanic hazards and helps create landslide-prone topography in the poorly consolidated materials typical of active convergent margins. Traditional remote-sensing technologies are of limited utility in identification and assessment of these hazards, due to the difficulty in imaging the ground surface through the dense cover of vegetation. Indeed, the subtle tectonic landforms associated with recent earthquake deformation are essentially invisible beneath the forest canopy.

“Finding active faults and tectonic landforms in densely forested regions using Airborne Laser Terrain Mapping (ALTM), Puget Lowland, Washington” is an ongoing collaboration between the USGS (Sam Johnson, PI) and the Geodynamics Branch which addresses this problem. The program is funded by the Solid Earth and Natural Hazards programs of NASA HQ. Mid-way through the project, the following have been accomplished:

(1) The project has fostered a partnership between the USGS, NASA and seven local government agencies, forming the Puget Sound Lidar Consortium (PSLC). The PSLC has pooled resources to acquire 4800 km² of ALTM data in 2000 and 2001, and is scheduled to collect an additional ~2500 km² of data in 2002.

(2) Discovery of five previously unknown linear scarps up to 6 m high in the ALTM “bald Earth” imagery that are interpreted as surface ruptures of large earthquakes, based on the results of trenching across two of the scarps. The ALTM imagery also enables detailed elevation measurements of uplifted and tilted shoreline terraces bordering Puget Sound that reveal actively growing folds produced by slip on the faults.

(3) The USGS and NASA, together with the Terrapoint, LLC, have developed new algorithms to derive “bald Earth” topographic maps in heavily forested terrain (see Figure 1) and have made major improvements in data acquisition and quality control procedures.

(4) ALTM images have proven to be an essential tool in regional landslide hazard assessment.

An example of the ALTM data acquired along a river valley is shown in perspective views of the canopy top and the underlying “bald Earth”, derived by filtering the data to remove laser returns from vegetation and buildings (Figure 1). The images are produced from digital elevation models at 1.8 m spatial resolution and depict an area 2.1 x 1.3 km in size. No vertical exaggeration is applied. The “bald Earth” image reveals a previously unmapped landslide deposit that was hidden beneath the vegetation cover.

Contact: David Harding, Geodynamics Branch, harding@core2.gsfc.nasa.gov

ICESat Laser Altimetry: Calibration Techniques for Precision Geolocation for ICESat-Waveform and Profile Matching to Digital Elevation Models

In the thirty years since the launch of the Skylab radar altimeter, satellite altimetry has proven to be a powerful tool for mapping the topography of the Earth and other planets. To fully exploit the data obtained from an orbiting altimeter, it is necessary to calibrate certain parameters not only before launch but also after the altimeter is in orbit. In preparation for the launch of the ICESat mission, ICESat science team member Jack Bufton and his LTP co-investigators (D. Harding, S. Luthche, D. Rowlands and J. Sauber) have developed a number of calibration techniques for the ICESat laser altimeter. These include the integration of multiple tracking data types with planned pointing maneuvers over oceans, as well as matching waveforms and profiles to land DEMs (Digital Elevation Models) [Rowlands et al., 2000]. The analysis of residuals for geolocated laser footprint tracks can assess the relative accuracy and reproducibility of the resulting geolocation. These analyses do not make possible the assessment of the absolute accuracy or systematic errors of a geolocation result, however.

Comparison of geolocation results to accurate DEM’s does provide a means to assess the absolute accuracy and systematic errors of the laser footprint position, and to evaluate alternate geolocations methods. The comparison can be done based on differencing elevation profiles or waveforms with respect to an accurate DEM. In profile matching, the elevation for each footprint along a profile is differenced with respect to the corresponding DEM elevation. With the ability of laser altimeters to digitize the backscattered energy (and therefore produce waveforms), the waveform matching approach potentially has greater sensitivity in assessing footprint geolocation than the profile matching. Waveform matching is accomplished by minimizing the residual between within-footprint surface height distributions as recorded by the observed waveform and a simulated waveform derived from an accurate DEM.

The waveform matching approach is being assessed prior to launch in order to establish the DEM characteristics required for validation purposes. We are conducting sensitivity studies with various DEMS. The studies consider DEM resolution, accuracy, extent, vegetated versus non-vegetated, rugged versus smooth topography, and uniform versus spatially varying reflectance. One of the DEM’s that has been used is a 1.8 m resolution DEM of a 2350 sq km area of western Washington State acquired for the Puget Lowland Lidar Consortium. Results of airborne lidar studies in that vegetated area are described elsewhere in this section. To assess ICESat performance over Alpine glaciers we are also using a DEM from part of the Bagley Ice Field and Seward Glacier in southern Alaska.

References:

Rowlands, D.D., C.C. Carabajal, S.B. Luthcke, D. J. Harding, J.M. Sauber, and J.L.

Bufton, Satellite Laser Altimetry: Orbit Calibration Techniques for Precise Geolocation,

The Review of Laser Engineering, vol. 28(12), pp. 796-803, December, 2000.

Contact: Jeanne Sauber and David Harding, Geodynamics Branch, jeanne@steller.gsfc.nasa.gov

"Exploring Space, Exploring Earth: New Understanding of the Earth"

A book which summarizes the impact of space research on solid earth geophysics and geology is now in production and will be published by Cambridge University Press in 2002. The book is written from the unique perspective of Paul Lowman, the first geologist hired by NASA (in 1959) who has been involved in both space exploration and study of the Earth from orbit. This dual career provided access to almost all aspects of geological and geophysical research carried out by NASA and GSFC scientists.

A foreword by Neil Armstrong summarizes classic efforts to determine the size and shape of the Earth in ancient Greece. The first major chapter reviews space geodesy, a field founded by the late John A. O'Keefe (to whom the book is dedicated) with discovery of the "pear-shaped earth" from Vanguard tracking data in 1959. The next chapter covers crustal magnetism as revealed in data from satellites such as Magsat, the launch of which in 1979 propelled Goddard to the forefront in this new field of study. Remote sensing from Earth orbit follows, with extensive examples of Landsat imagery and that from other satellites, including Terra (see Figure 1).

To this point, the book has been primarily focussed on the direct results of space research in geology and geophysics. Later chapters cover a wider range topics that apply, perhaps less directly, to the study and understanding of the Earth. Among these are impact cratering and comparative planetology as they relate to evolution of the Earth's crust. The final chapter discusses the effect of life on terrestrial tectonics. This is interpreted through the Gaia Hypothesis, in which the Earth is thought to have various parameters, such as temperature, regulated by life. The book describes a "biogenic theory of tectonic evolution," in which the broad structure of the earth is the result of biologically-promoted sea-floor spreading and related phenomena superimposed on a crustal dichotomy formed by catastrophic impacts on the early Earth. The final line of the book describes the Earth as it might be reported by an interstellar visitor: “It is unique, a uniqueness that may be due primarily to its life. "

"Exploring Space, Exploring Earth" is a synthesis of scientific progress over many decades from many contributors, but also provides a comprehensive summary of the achievements of scientists and engineers at Goddard Space Flight Center.

Contact: Paul Lowman, Geodynamics Branch), lowman@core2.gsfc.nasa.gov

Planetary Geology and Geophysics (Frey, Purucker, Roark, Sakimoto, Taylor, Voorhies)

The Mars Orbiter Laser Altimeter (MOLA) on board the Mars Global Surveyor (MGS) mapped the topography of Mars until the end of June, 2001, producing more than 600,000,000 precision ranges to the surface and clouds of the planet. The MOLA Science Team released in 2001 a Gridded Data Product of the topography of Mars, at a resolution 64 pixels/degree (better than 1 km) and accuracy of 1 meter with respect to Mars' center of mass. MOLA data is the basis for numerous studies of Mars, of both its topography and its gravity field. Discovery of a very large population of buried impact basins below the plains in the northern lowlands has provided important constraints on the origin of the crustal dichotomy and the age and nature of the lowland crust. As described below, it appears the lowlands formed extremely early in Martian history, which has implications not only for crustal evolution but perhaps also for the possibility of life. Accumulating evidence for extremely young volcanic flows and features on Mars, observed not only by high resolution images but also shown by their fresh topographic character as revealed by MOLA data, raises the possibility that Mars may still be, at least locally, warm and active. Constraints on eruption rates and inferences of volcanic style have also been made, supported by laboratory and theoretical work constrained by terrestrial data on known flows. Much of this work has been carried out in collaboration with students from various high schools and universities, and has benefited enormously from in-house graphics software development using IDL. A program GRIDVIEW, as described below, has reached a mature state in which staff and even short-term visitors can rapidly study and analyze MOLA gridded data to address a wide variety of problems.

Experience with magnetic field satellites in low orbit around the Earth (e.g., Magsat) has been applied to the study of crustal magnetic anomalies discovered on Mars. Techniques for representing the anomalies in terms of dipole sources were used to produce an improved description of the anomaly field. A comparison of the spectral characteristics of the Earth and Mars fields dramatically shows the difference between the mostly induced crustal anomalies of the Earth and the remanent anomalies on Mars, and emphasizes the much greater strength (and inferred magnetization) of the martian sources.

Ancient Lowlands on Mars

The BIG problem in martian geologic evolution is the origin of Mars’ fundamental crustal dichotomy: the separation of the crust into two quite distinct “hemispheres”. The southern parts of Mars are generally high and heavily cratered, an indication of their great antiquity. The northern lowlands are 3-5 km lower and covered by plains which, based on their much smaller number of craters, must be much younger that the highlands.

The origin of this crustal dichotomy is a problem of both how and when it occurred. Some have suggested formation of the lowlands occurred in the middle period of martian history (the Hesperian epoch) while others believe the dichotomy dates to the Noachian, or early portion of Mars history. How the crust of Mars became separated into these two distinct provinces has been a matter of controversy for as long as the dichotomy has been recognized. Explanations have ranged from purely exogenic (a single giant impact or several very large impacts) to purely endogenic (internally driven processes including mantle convection, subcrustal erosion, or perhaps plate tectonic processes). One problem with understanding the origin of the dichotomy is that the nature of the lowland crust below the plains was previously unknown.

Frey and coworkers showed that MOLA data reveal the presence of what are most likely buried impact basins below the visible surface of Mars (Frey et al., 1999, 2001). These “Quasi-Circular Depressions” (QCDs) are generally not visible in imaging data but well shown by stretched color and contoured MOLA data. A search for these in the relatively young lowlands revealed a very large number (644) larger than 50 km in diameter (Figure 1). Of these, only 90 were visible in Viking imagery.

That means that 85% of the population lies below the plains and that underlying crust is very old. It also implies the plains are relatively thin, so that relic topography still shows through. Cumulative frequency curves (cumulative number larger than a given diameter per unit area versus diameter) show that the buried lowland surface (below the plains) is older than the visible highland surface (but not as old as the buried highland surface). Furthermore, comparison with crater counts done on smaller areas of designated geologic units whose relative stratigraphy is known suggest the buried lowland crust is Early Noachian in age, dating from the earliest time period in martian geologic history, the time of intense cratering. But when did it become low? Unless there is some way to lower an already cratered surface without destroying the craters (that today we see as the buried QCDs), the lowlands must have been lowered during the period of intense cratering, as indicated in Figure 3.

This constraint on the lowlands is only temporal, but it may limit the likely mechanisms that can form the lowlands (and therefore the crustal dichotomy). It certainly favors processes which occur early and quickly (such as large scale cratering), and may make longer-lived processes such as whole mantle convection and plate tectonics unlikely.

There are implications of this work that go beyond the question of the martian crustal dichotomy. It is sometimes assumed that the oldest terrains on Mars date from its origin 4.6 billion years ago. If buried impact basins were found in these very old units, the oldest known terrains would then have to be younger than 4.6 billion years ago. Also, it appears there was a northern lowland essentially throughout martian history. That means that at whatever early time conditions on Mars permitted water to run across the surface, there was a basin into which the water could drain. So there may well have been an early (shallow) ocean on Mars, with all the implications that can have for the possibility of life.

This work was supported by the Mars Global Surveyor Project.

References:

Frey, H., S.E.H. Sakimoto and J. H. Roark, Discovery of a 450 km diameter, multi-ring basin on Mars through analysis of MOLA topographic data, Geophys. Res. Lett., 26, 1657-1660, 1999.

Frey, H.V., J.H. Roark, K.M. Shockey, E.L. Frey and S.E.H. Sakimoto, Ancient lowlands on Mars, Geophys. Res. Lett. (in press), 2001.

Contact: Herbert Frey, Geodynamics Branch, frey@core2.gsfc.nasa.gov

New Perspectives on the Enigmatic Medusae Fossae Formation, Mars

One of the most enigmatic formations on the surface of Mars is called the Medusae Fossae Formation (MFF). Its young age, distinctive surface texture, localized setting, and lack of obvious source have prompted a variety of proposals concerning its origin. Mars Orbiter Laser Altimeter (MOLA) data in conjunction with high resolution Mars Orbiter Camera (MOC) images from the Mars Global Surveyor (MGS) mission have provided important information on the extent, volume, roughness, orientation and superposition relationships that help limit the likely nature of the MFF.

Figure 1. Location map for Medusae Fossae Formation (MFF). Base map is MOLA topography. Scale is on the right. Outlined areas are location of MFF deposits based on earlier Viking image mapping. Hatched region is the crustal dichotomy boundary separating old cratered terrain to the South from younger smooth plains to the North. Our studies indicate the MFF may have had wider extent than shown by the above map, but the deposits may be thinner than previously assumed.

The MFF is a wind-scoured deposit located near the equator of Mars between the Tharsis and Elysium volcanic centers at 130-240E and 15S to 15N. Located along the major crustal dichotomy boundary (Figure 1), it overlies both old cratered highlands and young plains-forming units, which include very young volcanic flows and small cones. At large scales the formation appears smooth, but at smaller scales there are lineations that appear to be wind-eroded features (see below and Figure 3). Overall the MFF has a MOLA pulse width roughness 2-3 times that of the Mars global average.

Earlier studies based on Viking imagery led to a variety of hypotheses for the MFF, including ignimbrites or ash flows, carbonate platforms, shoreline terraces, rafted pumice deposits, paleo-polar deposits, and uplifted and exhumed ancient terrain. Geologic mapping earlier identified three distinct units (upper, middle and lower members) of the formation, but more detailed images available from MOC have shown multiple layers and cohesive caprock overlying much more friable (easily eroded) material. Studies by Bradley et al. (2001) using MOLA and MOC data have now reduced the likely origins to two quite different but atmospherically-related possibilities: volcanic airfall or aeolian deposits.

In particular, MGS data have shown that the MFF is draped over but does not always fully mask peaks and valleys in older terrain. There is clear indication that ancient fluvial systems underlie the MFF but also that MFF material appears to deflect some channels, suggesting a complex interplay between two quite distinct processes. MOLA topography has significantly reduced the estimates of the thickness of the MFF from an earlier 3 km to about 1 km on average, suggesting only about half the volume previously believed. On the flip side, the areal extent of MFF materials in the past may have been much larger than what currently survives.

Although some researchers have resurrected the suggestion that the MFF is similar to polar layered terrain units and may represent paleo-deposits of this unit (located near the present-day equator!), the MOLA topography does not support this.

In particular, the width, depth and wall slopes of valleys in the MFF materials (Figure 2b) are quite different from those of valleys in polar layered terrain (Figure 2a; also Figure 2c). Inter-valley roughness is much greater in MFF materials and the valley spacing much more irregular than for polar layered terrains. On the basis of these quantitative comparisons it appears unlikely the MFF are paleo-polar layered terrains.

Yardangs (irregular grooved ridges produced by wind erosion of weakly consolidated sediments) suggest multiple episodes of wind-related deposition and erosion in the MFF. They have a variety of orientations, including bi-directional patterns where the deposits are thin, but the lack of a consistent pattern overall suggests neither wind direction nor underlying topography alone controls their emplacement. But the observation of a common angle between intersecting sets of yardangs in some areas of the MFF (Figure 3) does suggest that material properties such as jointing may control their orientation in these areas.

The new data provide little support for some of the suggested origins of this unusual deposit. In particular, the lack of widespread horizontal layering virtually eliminates ideas such as carbonate platforms, paleo-shorelines or other water-related origins. As described above, paleo-polar deposits like present-day polar layered terrains are also not viable. The two most likely remaining candidates are volcanic airfall deposits such as ashes and tuffs or non-volcanic aeolian deposits such as loess. Volcanic airfall is most consistent with the whole range of characteristics of the MFF, especially: presence of widespread resistant (to erosion) zones, ability of MFF to drape pre-existing topography but also to apparently deflect development of a fluvial channel, widespread occurrence of yardangs with different orientations, and apparent evidence for jointing.

References:

Bradley, B.A., S.E.H. Sakimoto, H. Frey and J.R. Zimbelman, The Medusae Fossae Formation: New Perspectives from Mars Global Surveyor, J. Geophys. Res. (Planets) in press, 2001.

Contact: Susan Sakimoto, UMBC/GEST at the Geodynamics Branch, sakimoto@core2.gsfc.nasa.gov

Recent Floods and Volcanism in the Cerberus Plains, Mars

The Cerberus Plains on Mars are a smooth low area just north of the planet's dichotomy boundary, located along the equator between the two large volcanic centers of Elysium and Tharsis. A long series of sub-parallel fissures called the Cerberus Rupes stretches from Elysium across the plains to their center. Based on studies done using 1970’s vintage Viking Orbiter data, explanations for the Plains included deposits laid down by water, a dry lakebed, or simply lava flows. Since Mars Global Surveyor arrived in 1997, and began mapping in 1999, the area has been under renewed scrutiny which has taken advantage of the much higher resolution imagery and elevation data now available.

Detailed images from the Mars Orbiter Camera (MOC) have revealed fresh-appearing, young lava flows, possibly less than 10 million years old (based on crater counts). Mars Orbiter Laser Altimeter (MOLA) data have provided stunning, detailed topography. Susan Sakimoto has used the topography to trace young lava flows across the plains back to their sources at the Cerberus Rupes fissure, and to characterize the range of vent types observed in the region. The region is mostly lava-covered and clearly not simply a lakebed. Sakimoto has also found that the volcanic flow rates are similar to those expected for plains volcanism regions on Earth. This work was reported at the annual Geological Society of America meeting (Sakimoto et al., 2001), and was reviewed in Science, (Kerr, 2001).

Cerberus Plains on Mars, showing fissures, vents, shield volcanoes, lava flows, fluvial channels and inferred directions of flows of both water and lava. Elevation shown by color (dark blues and purples are lows; greens are higher elevations).

The lava flows in Cerberus (Figure 1) are topographically some of the best preserved on the planet. This makes them excellent candidates for modeling what types and rates of lava might have flowed through the observed channels. This modeling work is funded by the NASA Mars Data Analysis Program. As a terrestrial validation of lava flow models, Sakimoto is also working with Dr. Tracy Gregg at the University at Buffalo using a combined approach of analytic models, laboratory simulations, field data, and computational fluid dynamics simulations (Sakimoto and Gregg, 2001). Figure 2 shows the excellent agreement reached so far with the field velocity measurements matched by channel model predictions of lava flow velocities.

Sakimoto and collaborators are extending this terrestrial work into related flow field and shield analyses with the help of a NASA-funded Space Grant collaboration with Scott Hughes and Glenn Thackeray on a study of the Snake River Plains volcanism. The Snake River Plains serves as both as a terrestrial field case study, and as a basis for comparison for the Cerberus and other volcanic fields on Mars.

One of the more interesting terrestrial/planetary comparisons related to this work is in the realm of lava-water interactions, which many consider prime sites for life. Sakimoto, in collaboration with Devon Burr (a NASA Graduate Student Research Fellow at the University of Arizona), and her advisor Alfred McEwen, has shown that water likely also originated from the Cerberus Rupes, flowed across the western plains in the recent geologic past, and then debauched onto permeable lavas, where it may still exist as shallow ground ice (Burr et al., 2001, 2002). The story of the lava-water interplay for this region is expected to have very interesting implications for the preservation of hydrothermal environments late into martian history, since these are some of the youngest lava and water flows yet found on the planet. Sakimoto and Burr suggest that the fissures, water and lava flows are the surface manifestation of subsurface magma that has kept the local subsurface warm for tens of millions of years. As the magma rises towards the surface, it probably encounters and helps release subsurface water. This erupts onto the surface before or after the lavas to create the intermingled lava and water channels and flows so clearly seen in the recent topography and images.

Since much of this activity is thought to be geologically recent (based on impact crater counts by other researchers), magma may well still be present at depth. The Cerberus fissures could erupt again. Given that the local conditions (likely warm and wet) are both interesting and perhaps hospitable for possible subsurface life, this region is one of four sites in consideration as a landing site for one of the two Mars Exploration Rovers scheduled for launch in 2005.

References:

Burr, D. M., A.S. McEwen, S.E.H. Sakimoto, Recent aqueous floods from the Cerberus Rupes, Mars, Geophysical Research Letters, in press.

Burr, D., A. McEwen, S. Sakimoto, Recent Aqueous Floods From the Cerberus Rupes, Mars, Eos. Trans. AGU, 82(47), Fall Meet. Suppl., Abstract P22A-0537, 2001.

Sakimoto, S.E.H. and T.K.P. Gregg, 2001,Channeled flow: Analytic solutions, laboratory experiments, and applications to lava flows, Journal of Geophysical Research-Solid Earth, Vol. 106., No. B5, p. 8629-8648.

Sakimoto, S.E.H., S.J. Riedel, D. Burr, Geologically Recent Martian Volcanism and Flooding in Elysium Planitia and Cerberus Rupes: Plains Style Eruptions and Related Water Release? Presented at the Fall GSA meeting, Boston, Nov 5-8, 2001.

Kerr, R., Life-Potential, Slow, or Long Dead, Science, vol 294, p1820-21, 2001.

Contact: Susan Sakimoto, UMBC/GEST at the Geodynamics Branch, sakimoto@core2.gsfc.nasa.gov

GRIDVIEW: Interactive Software for Analyzing Gridded Data

The Geodynamics branch has over the last several years developed a scientific visualization tool called GRIDVIEW that can be used by researchers and students alike to study gridded data sets. The program was created originally because researchers within the Laboratory needed a way to quickly and easily view and study the newly acquired topographic profile and gridded data from the Mars Orbiter Laser Altimeter (MOLA), one of the instruments on the Mars Global Surveyor satellite. The software was created with IDL (Interactive Data Language, from Research Systems, Inc.) and has evolved over several years to become a highly versatile yet user-friendly analysis tool. It is currently used by researchers and students in the US and other countries, including Great Britain and Japan.

GRIDVIEW features a variety of functions and capabilities, some of which are described below, and uses simple pull-down menus for function selection by point-and-click. The program is extremely easy to use. After given a short (< 30 minutes) introduction to the program’s capabilities, high school and undergraduate students have been able to use this tool to analyze the topographic characteristics of Mars in support of several education and research projects currently ongoing in the Geodynamics Branch. A graphical mouse driven interface allows the user to rotate a planetary globe (Figure 1a) and zoom into areas of interest and view the data in numerous ways (Figure 1b, 1c). Color representations of the data can be interactively stretched and manipulated to highlight specific details in the data (Figure 1b, 1c). A profile picking and display tool provides interactive options for distance, height and slope measurements along the profile (Figure 1d).

A list of the major features available includes:

Global Rotation and Zooming

Selectable Color Representation of Data

Shaded Relief Viewing Options

Latitude / Longitude / Data Value Tracking

Color and Contrast Stretching

Contouring

Profiling

Interactive Distance, Height and Slope Measurement Tools

Basin Center and Diameter Measurement (Circle Fitting)

Plotting Basin Ring Files

Overlay Contours of other Data Sets

Postscript and Image Figure Output

The combination of color stretching and contouring capabilities of GRIDVIEW to highlight subtle details of the data has made it a very powerful tool for mapping and discovering previously unknown buried basins on Mars. This is described elsewhere in this report (“Ancient Lowlands on Mars”). In addition, GRIDVIEW is routinely used to measure heights and slopes of small volcanic features, depths of craters, channels and canyons, and geographic correspondence between different features (geologic units, gravity anomalies, magnetic anomalies, topographic features). It has also become a valuable tool for investigations of the crustal dichotomy boundary, characterization of geologic formations, calculation of sediment fill in and around large craters, and measurement of the thickness of volcanic flows and debris aprons.

The program was specifically designed to work with MOLA topography data but can be used to analyze any gridded data. It has been used in the Geodynamics branch to study gravity and magnetic data as well as Earth topographic data. Although development continues and additional features may be added, the utility of the program is so great that we have made it available to anyone wishing to use it. GRIDVIEW requires the installation of IDL software (http://www.exelisvis.com/ProductsServices/IDL.aspx) and will run on any computer system supported by IDL (Windows, Mac, UNIX). The program is available for download on the web at http://core2.gsfc.nasa.gov/mola_pub/gridview .

Contact: James Roark, SSAI at the Geodynamics Branch, roark@core2.gsfc.nasa.gov

Magnetic Spectra of Earth and Mars

It has been known for decades that the Earth’s crust contains magnetic anomalies, regions that, in the presence of the Earth’s strong internal magnetic field, take on enhanced magnetic intensity or regions that were magnetized as they cooled and now preserve a remanent magnetization. These were first mapped in detail from orbit by NASA’s Magsat in 1980, and more recently by the Danish Oersted and German CHAMP missions. It was only in the last few years that crustal magnetic anomalies were discovered on Mars by Mars Global Surveyor (MGS), but that has turned out to be one of the major findings and surprises of that mission. Techniques previously developed for the study of Magsat crustal anomalies have been used to better represent and help study the anomalies on Mars. One of these techniques is a spectral method that allows a separation of the contribution from crustal sources from that due to the main core field of the Earth.

When astrophysicists measure light from a distant star, they can break up the starlight into its component colors, creating a spectrum from long wavelength red to short wavelength blue, by passing the light through a mechanical prism. Similarly, when geophysicists measure a planet’s magnetic field, they can break up the field into long and short wavelength components by running the data through a kind of digital prism called “spherical harmonic analysis.” And just as astrophysicists can learn much about a star’s surface and interior from its optical spectrum, geophysicists can infer something about a planet’s crust and core from its magnetic spectrum.

The Earth’s magnetic spectrum is determined from measurements made at magnetic observatories on the surface, from aeromagnetic data, and from global surveys by orbiting satellites. Recently, satellite data has also become available for Mars. Thus it is now possible for the first time to compare the magnetic spectra of two different terrestrial planets, and to use techniques that separate out the core and crustal components as has traditionally been done for the Earth. This comparison is shown in Figure 1, which plots magnetic power (actually mean square magnetic induction) versus harmonic degree, a measure of the wavelength scale. Wavelength decreases from left to right along the horizontal axis. Dots and crosses show the actual magnetic spectrum of Earth and Mars, respectively. The curves are fits to specific types of source models, as described below. The Earth’s magnetic spectrum has two distinct branches: a powerful, long-wavelength, rapidly decreasing (with harmonic degree) core-source magnetic field, caused by electric currents in its liquid iron outer core. The second flatter branch is due to a shorter-wavelength crustal-source field, caused by magnetization of crustal rock. These two branches cross at about degree 14. For the Earth we find best fitting fields from a core of radius 3512 + 64 km (very similar to the seismologic core radius of 3480 km), and from a crust represented by a shell of random dipolar sources at radius 6367 + 14 km, near Earth’s mean radius of 6371.2 km.

The spectrum for Mars has no sign of a core-source field. This is consistent with the lack of direct observation of a global field by MGS. There is only a field from crustal sources, represented here by a shell of random dipoles, but now at radius 3344 + 10 km, and with sources about 10 times stronger than Earth’s. The shell radius is about 46 km below Mars’ mean radius of 3390 km, and agrees well with the mean depth of Mars lithosphere inferred by independent modeling of MGS topography and gravity data. Our results indicate Mars has a thicker, more intensely magnetized crust than Earth. This may in part be due to an iron-rich crustal magnetic mineralogy for the Red Planet, magnetized in the past by a now defunct core-dynamo.

These results are described more fully in a paper now in press in Journal of Geophysical Research – Planets.

Contact: Coerte Voorhies, Geodynamics Branch, voorhies@geomag.gsfc.nasa.gov

Valles Marineris May Provide a Key to Understanding Early Mars

On the Earth, magnetic field observations and in particular crustal magnetic anomalies can be used for a variety of studies, including geologic reconstructions, plate tectonic interpretations, and resource exploration. Crustal anomalies on the Earth are of two kinds, those induced by the present-day main core field and remanent anomalies produced when magnetic rocks cooled in the presence of the past magnetic field. The alternating pattern of anomalies that flank mid ocean ridges on the Earth and which characterize the seafloor spreading part of plate tectonics are of the latter type.

There exists a debate as to whether Mars ever experienced plate tectonic processes such as produce seafloor spreading on the Earth. The discovery by Mars Global Surveyor (MGS) of a strongly magnetic crust on Mars and the pattern of the anomalies seen has been interpreted by some as evidence for an early episode of plate tectonics on that planet. Whether or not this is so, techniques developed for the interpretation of crustal magnetic anomalies on the Earth, including those seen at satellite altitude by Magsat and other satellites, can be applied to the magnetic anomalies on Mars. These martian anomalies are of the remanent type since there is no present-day intrinsic field to induce anomalies of the sort known on Earth.

Figure 1 shows two recent magnetic maps of Mars derived using these techniques. The anomalies have a pattern strongly suggestive of faulting and perhaps offset along faults along a major tectonic structure. The Vallis Marineris on Mars is a series of large, fault-bounded canyons which have been compared with major rift structures on the Earth. The pattern in the magnetic maps, especially the abrupt truncation of the anomalies at the wall of the canyon, supports the idea that the Valles Marineris canyon is a tectonic graben. The maps also suggest that highly magnetic source rocks exist at the intersection of Coprates and Capri Chasmata, on the northeast corner of the canyons, and there is a good possibility that these magnetic rocks may be exposed along the fault wall.

It is generally accepted that the magnetic anomalies on Mars are very old. It is also likely that the Valles Marineris, though formed in the middle part of martian history, exposes much older rock, as suggested by the extensive layering revealed by the Mars Orbiter Camera on MGS. If so, the possible exposure of ancient highly magnetized rocks in the Valles Marineris may provide important clues to the early tectonic evolution of Mars and the nature and demise of the martian core dynamo.

Contact: Michael Purucker, RSTX at the Geodynamics Branch, purucker@geomag.gsfc.nasa.gov

Orbital-Rotational-Climatic Interactions (Bills, Liu, Rubincam)

There is important coupling between the tilt of the axis (obliquity) of the non-spherical (oblate) solid but deformable Earth and its orbital motion that plays into the long term change in climate. Obliquity-oblateness feedbacks exist which can be important. Frequency modulation of the obliquity may be a significant forcing on changing the insolation and therefore the climate of the Earth. Similar effects exist on other planets. For Mars, for example, there are possible long term secular changes in the size of the polar cap. And sublimation and freezing of atmospheric constituents on planetary satellites like Triton and Charon are yet more examples of this interplay.

The diurnal and seasonal Yarkovsky re-radiation effect may contribute to the evolution of asteroid and meteroid orbits and could help populate the near-Earth space with small objects. Also, tidal dissipation and how it changes with and helps change orbital eccentricity is an important source of thermal energy for Mercury, which may help explain how that small and slowly rotating body still maintains an internal (core) magnetic field.

Yarkovsky Effects and the Orbital Evolution of Asteroids and Meteoroids

Ivan Osipovich Yarkovsky (1844-1902), a civil engineer who worked on scientific problems in his spare time, first proposed an effect which now bears his name. Writing in a pamphlet around the year 1900, Yarkovsky noted that the diurnal (day-night) heating of a rotating object in space would cause it to experience a force. This force, while tiny, can lead over time to large secular effects in the orbits of small bodies, especially meteoroids and small asteroids. In particular, the orbits of the meteoroids evolve towards resonances where gravitational effects of other planets cause them to be delivered into Earth-crossing orbits and eventually to Earth.

Yarkovsky's effect is a radiation force, and is the photonic equivalent of Fred Whipple's rocket effect. The basic idea behind Yarkovsky's diurnal effect is shown in Figure 1, which shows a spherical meteoroid in a circular orbit about the Sun. For simplicity the meteoroid's spin axis is taken to be perpendicular to the orbital plane, so that the Sun always stands above its equator. Insolation heats up the sunward side; the heat is ultimately re-radiated into space by the meteoroid (typically in the infrared part of the spectrum, unless the meteoroid is very close to the Sun). When it leaves the meteoroid, an infrared photon carries away momentum p according to the relation p = E/c, where E is energy, and c is the speed of light. Because more energy (and therefore more momentum) departs from the hotter side of the meteoroid than the colder side, the meteoroid feels a net kick in the direction away from the hotter part.

If the meteoroid had no thermal inertia, then the temperature distribution would be symmetrical about the subsolar point and the meteoroid would experience only a net force radially outward from the Sun. The only consequence of this force would be to weaken the Sun's gravitational grip on the meteoroid. However, all bodies do have some thermal inertia, which causes a delay in heating and cooling of its material. The hottest part of the meteoroid is actually its afternoon side rather than the subsolar (noontime) point. This is similar to the Earth, where afternoon is the warmest time of day, instead of noon. As a result, the force on the meteoroid has both a component radially outwards from the Sun and also along-track.

This along-track component causes a long term increase in the size of the orbit for the direct or prograde sense of rotation shown in Figure 1. But the sign of the diurnal Yarkovsky effect depends on the sense of rotation. If the meteoroid shown in Figure 1 rotated in the backwards or retrograde sense, then the orbit would shrink instead of expand. The important point is that over time the tiny Yarkovsky force can profoundly change the orbit of meteoroids, affecting the number that finally strike the Earth. The effect, of course, is generally greater for smaller bodies (unless they are so small they lack significant temperature gradients across them).

Nearly a century after Yarkovsky wrote his pamphlet a second Yarkovsky effect was recognized. While searching for the cause of the secular decay of the orbit of the LAGEOS Earth satellite, we realized that in general there had to be a seasonal effect in addition to Yarkovsky's original diurnal effect. The seasonal effect applies not just to Earth satellites like LAGEOS, but also to meteoroids and asteroids orbiting the Sun.

The seasonal Yarkovsky effect is illustrated in Figure 2. As in Figure 1, a spherical meteoroid is assumed to be in a circular orbit about the Sun; but in this case the spin axis lies in

(not perpendicular to) the orbital plane. It is the component of force lying along the spin axis which gives rise to the seasonal effect. When the meteoroid is at the bottom of the figure, the Sun shines most strongly on its northern hemisphere. As with the diurnal effect, there is a delay in cooling due to thermal inertia, so the northern hemisphere is hottest further on in its orbit. For a body without thermal inertia the along-track force is periodic and averages to zero when integrated over one revolution about the Sun, but for real objects the average of the along track force is non-zero and leads to secular changes.

For small orbital eccentricities (orbits that are nearly circular), the average along-track force always opposes to the motion of the meteoroid: it acts like drag and causes orbital decay. Unlike the diurnal Yarkovsky effect, the seasonal Yarkovsky effect is independent of the sense of rotation of the meteoroid; reversing its spin does not change the effect's sign.

These Yarkovsky effects may be an important mechanism by which meteoroids and small asteroids reach orbital resonances. Once the bodies reach the resonances, the eccentricities are rapidly pumped up so that after a few million more years their orbits cross that of the Earth, and they may fall to the ground to be picked up as meteorites. The drift times due to these effects for stony meteorites are on the order of 20 million years, in good agreement with the cosmic ray exposure ages.

These results are in press as a chapter in the book ASTEROIDS III, to be published by the University of Arizona Press, with W. F. Botke, D. Vokrouhlicky, D. P. Rubincam, and M. Broz as co-authors.

Contact: David P. Rubincam, Geodynamics Branch, rubincam@core2.gsfc.nasa.gov

Obliquity Modulation of the Incoming Solar Radiation

For several decades, geophysicists have been trying to develop a physical understanding of how the 100-Kyr cycle in the ellipticity of the Earth’s orbit around the Sun shifts the pattern of insolation, triggering the growth and decay of great ice sheets in the high latitudes of the Northern Hemisphere. This effort has encountered great difficulty: The 100-Kyr orbital eccentricity cycle is too small in magnitude and too late in phase to produce the 100-Kyr climate cycle during the last 1 million years. This is one of the most perplexing and enduring puzzles in science. The second puzzle is how to explain the ice sheet cycles in the Northern Hemisphere from 1 to 2 million years ago when the global ice volume and deep sea temperature varied at an almost metronomic 41-Kyr period of the Earth’s obliquity. Thirdly, we still do not understand why the 100-Kyr ice age cycle became dominant about 0.9 million years ago.

Numerous hypotheses and models have been proposed to explain these climate puzzles. Proposed explanations have invoked: (1) Stochastic resonance of orbital eccentricity forcing. (2) Internal oscillations of the climate system near the 100-Kyr period that can get phase-locked to orbital eccentricity forcing. (3) High nonlinear response of climate system to weak forcing by the orbital eccentricity. (4) Variations in the inclination of the Earth’s orbital motion. (5) A climate system with three steady states and a set of pre-defined rules for moving between them. However, the physical mechanism of the climate cycles remains a scientific mystery.

In a paper published last year (Liu, 2001), it was shown, from a dynamics point of view, that the 100-Kyr periodicity of the ice age is not due to orbital eccentricity at all. Numerical climate models were used to demonstrate that frequency modulation of the Earth’s obliquity (tilt of the rotation axis with respect to the orbit plane) is responsible and accounts for major climate changes during the past 2 million years (See Figure 1). Calculations of the variation in solar energy flux at the top of the atmosphere showed that the insolation flux deficit is the physics behind the ice age glaciation. According to this idea, it is unnecessary to invoke internal feedbacks such as ocean circulation and atmospheric CO2 concentrations in the Earth system to understand the climate cycle puzzles. Results from model simulations are in good agreement with geological climate records for the past 2 million years and extension of the model simulations reveals that the current warming trend of the climate is almost over and will give way to a small ice age.

Reference:

Liu, H.S., Obliquity modulation of the incoming solar radiation, Recent Res. Develop. Atmos. Sci. 1, 15-37, 2001.

Contact: Han Shou Liu, Geodynamics Branch, liu@core2.gsfc.nasa.gov)

Tidal Dissipation in Mercury

Two outstanding characteristics of Mercury are its high density and its strong internal magnetic field. The high density suggests a very large fraction of the planet is in the form of an iron core, which is difficult to explain in terms of its likely origin. Thermal history models for the planet suggest that the iron core of Mercury is likely to be frozen solid at present, in which case there should be no active dynamo to maintain its known magnetic field. Tidal dissipation may provide an answer to both of these puzzles.

Tidal dissipation within the planet Mercury is highly variable in both space and time, but may contribute significantly to the overall thermal budget of the planet. Tidal dissipation is the process by which periodic gravitational deformation of one body by another is converted to heat. The intensity of tidal dissipation depends on the material properties of the deforming body and the intensity of the gravitational interaction. Tidal dissipation in the planet Mercury has previously been recognized as having played a role in the planet’s capture into a 3:2 spin-orbit resonance (see below), but has generally been ignored as a currently viable heat source. The present work suggests that such dissipation may in fact be far more important than previously realized, and may help explain why Mercury has a present-day main (core) magnetic field.

The path of the Sun, as seen from Mercury, is rather unusual. This is due to a combination of two factors, the 3:2 spin-orbit resonance (in which Mercury complete 3 rotations for every 2 orbits), and the high eccentricity of the orbit. The rotation rate is quite uniform, but the orbital angular rate changes significantly, with higher rates near perihelion and lower rates around aphelion. Though the rotation rate is 3/2 the average orbital rate, the instantaneous ratio varies throughout the orbit. For all orbital eccentricities in excess of 0.191059, the maximum orbital rate exceeds the rotation rate, and the Sun (as seen in the Mercury sky) makes a small retrograde (backwards) motion for part of the Mercury day. The exact path depends on the eccentricity of Mercury’s s orbit, which at the present time is 0.206 (highest of all the planets). Due to exchange of angular momentum with the other planets, Mercury’s eccentricity fluctuates significantly on time scales of 1 million to 10 million years (Figure 1), and has been as low as 0.1 and as high as 0.4 within the past 20 million years. In some simulations of the long term dynamical evolution of the solar system, the eccentricity of Mercury has been calculated to exceed 0.7 several times within the past 10⁹ years. At e = 0.8685, the orbits of Mercury and Venus intersect and a chance of collision becomes possible.

The two points on the equator of Mercury where the Sun is directly overhead at closest approach (perihelion) receive the most radiative heating, and also the most tidal heating. If the orbit were perfectly circular (e=0), the sub-solar point on Mercury’s surface would move at uniform rate, and the associated peak in tidal potential would remain constant in amplitude. As the eccentricity increases, the potential at the sub-solar point varies over the orbit, and the rate of motion of the sub-solar point along the equator becomes more irregular. So tidal effects becomes more variable.

The internal structure of Mercury is largely unknown, and as a result, details of the dissipation pattern and intensity are still quite uncertain. However, using conservative estimates of the density, rigidity, and viscosity structure for Mercury, it appears that the current global rate of dissipation is in the vicinity of 3 x10¹² Watts. This is about the same as for the Earth, but significantly less than for Jupiter's tidally heated and highly volcanic moon, Io (10¹⁴ Watts).

The dissipation rate depends strongly on orbital eccentricity, and so changes as the eccentricity evolves. Figure 2 illustrates that dependence for a particular model of internal structure, assuming a 50 km thick elastic lithosphere, a 500 km thick viscoelastic mantle with a viscosity of 10²¹ Pa s (similar to the Earth’s mantle), and a fluid iron core. These material properties are highly uncertain, but are believed to be representative. The dissipation rate is nearly constant for eccentricities at or below the present value, and then exhibits a dramatic increase as the eccentricity increases.

Episodes of extremely high orbital eccentricity may have lead to tidal heating rates large enough to evaporate away significant amounts of volatile constituents. This could account for the observed high density, as has been suggested to explain a similar larger-than-expected density for Io. Tidal dissipation may also provide the thermal energy to keep the core liquid and convecting, and thereby maintain the magnetic field.

Contact: Bruce G. Bills, Geodynamics Branch, bills@core2.gsfc.nasa.gov)

Back to front page