New Internet Software Aids Paleomagnetic Analysis and Plate Tectonic Reconstructions

Vol. 82, No. 45, 6 November 2001

New Internet Software Aids Paleomagnetic Analysis and Plate Tectonic Reconstructions

Antonio Schettino and Christopher R. Scotese

For additional information, contact Antonio Schettino, Dipartimento Nuove Tecnologie, Istituto Tecnico Industriale "E. Molinari", Via Crescenzago 110, 20132 Milano, Italy; E-mail: aschettino@itis-molinari.mi.it

A new method for the generation of synthetic and smoothed APWPs was recently proposed by Schettino and Scotese [2000]. A synthetic APWP for a given reference plate is based on the transfer of paleopoles from several continents to the reference plate via the global plate tectonic rotation model. For example, paleopoles from North America, Australia, etc., may be combined with African paleopoles to generate a "composite" African APWP. In fact, if the total reconstruction poles for any continent with respect to the reference plate are known, then each "foreign" paleopole can be rotated into the coordinates of the reference continent and can contribute to the construction of a composite APWP. This method has been applied by several authors during the last 2 decades [e.g., Besse and Curtillot, 1991; Bocharova and Scotese, 1993].

Various attempts alternative to the conventional sliding time-window approach have been less successful at smoothing APW paths [e.g., Jupp and Kent, 1987]. Our new smoothing method [Schettino and Scotese, 2000] is simple and much more intuitive with respect to other mathematical treatments. In this method, a spline curve is fit to the plot of paleodeclinations to give the best estimate of declination though time. In a similar fashion, a spline curve is fit through the plot of paleolatitudes to give the best estimate of latitude through time. A synthetic APWP, which does not depend upon the selected reference point, is then generated by recombining information from the independently smoothed paleolatitudinal and declination curves.

A two-part set of software tools developed at the Istituto Tecnico Industriale "E. Molinari" of Milan, Italy, reflects this new smoothing method. The first part, the Paleomagnetic Subsystem (PS), comprises applications for manipulating a modified version of the Global Paleomagnetic Database (GPMDB), searching tables, generating listings of quality paleopoles according to filtering parameters such as plate identification, producing plots and spline regression curves of declination and paleolatitude for a reference site, and generating smoothed Apparent Polar Wander Paths (APWPs). All these tools, as well as the specific implementation of the GPMDB, have been designed for use in the field of plate tectonic modeling. The Plate Reconstruction Subsystem (PRS) comprises at present a single application that generates global maps of the past configuration of the continents and ocean basins in several map projections from Early Jurassic to recent time. For all these reconstructions, the reference continent, central Africa, is oriented according to a synthetic and smoothed APWP in a standard paleomagnetic reference frame.

The tools are accessible by URL at: http://www.itis-molinari.mi.it/Geo.html. A complete set of static paleogeographic maps for the entire Phanerozoic can be found at the Paleomap Project Web site: http://www.scotese.com/.

Software Implementation

The programs were developed as Common Gateway Interface (CGI) executables to be run on a Windows NT Web Server. For best performance, all the applications were written in C/C++, adapting corresponding modules of a desktop software package for plate tectonic modeling designed by the first author [Schettino, 1998]. Some of these programs generate HTML pages containing links to related results. For instance, a page containing information about a rock unit also contains a link to the corresponding journal reference as well as to related paleopoles. This method allows easy navigation across the GPMDB tables. Other applications also generate flat ASCII tables that can be downloaded and used by other programs. The graphic output (plots, APWPs, or paleogeographic maps) consists of downloadable images in Portable Network Graphics (PNG) format. This new, powerful graphic system is a patent-free replacement for GIFs. For more information about this format visit the W3C site at http://www.w3.org/Graphics/PNG/ or the official PNG site at http://www.libpng.org/pub/png/.

The Paleomagnetic Subsystem (PS)

The modified version of the GPMDB used by the paleomagnetic subsystem includes only three of the six main tables that are defined in the original database [Lock and McElhinny, 1991]. These are the Reference, Rockunit, and Pmagresult tables. Two other "stand-alone" help tables from the original GPMDB, Information and Timescale, were also incorporated to give further information about rock units and paleopoles. An additional table, not included in the standard GPMDB release, contains data about the set of Mesozoic and Cenozoic tectonic elements and plates that were identified by the authors. Each block is defined as a rigid tectonic unit of continental lithosphere bounded by faults, or, in few cases, by folds, with an independent tectonic history during the Mesozoic or the Cenozoic. The Plates table includes 223 tectonic blocks or plates defined in the global and regional paleotectonic models of Scotese [1990] and Schettino and Scotese [2000]. This data set will be further discussed in two publications being prepared by Schettino and Scotese. Figure 1 illustrates the structure of this implementation of the GPMDB.

**Fig. 1.** Structure of the paleomagnetic database installed on the Web server. Arrows represent one-to-many relationships.

The main difference between this version and the original version of the GPMDB is in the paleopole grouping criterion. In the original data set of Lock and McElhinny [1991], paleopoles are grouped by continent, and each result may be assigned to a terrane. Terranes are in turn grouped by continent. This continent/terrane model does not easily integrate with existing global plate tectonic models that describe the motion of a large number of independently moving tectonic elements. These tectonic elements may or may not correspond to terranes defined in the GPMDB. For this reason it is difficult to combine paleopoles from different continents to produce composite APWPs. Our implementation allows a better integration of paleomagnetic data and plate tectonic modeling and reconstruction tools.

The Web server described in this paper provides independent paleolatitude or declination plots, as well as synthetic and smoothed APWPs, by the method described above. All of these tools are included in the paleomagnetic subsystem. In an example of paleolatitude plot for a point at (0N, 25E) in Central Africa (Figure 2a), 165 filtered paleopoles with an age range 0-200 Ma from Central Africa, Northeast Africa, Madagascar, Somalia, India, Arabia, Australia, Antarctica, the Brazilian Craton, Parana Block, and Patagonia were used to generate the curve. Filtering was performed according to the following reliability criteria (default values): B4 (number of sites), N/B4 (mean number of samples per site), ED9515^o (95% confidence interval), DEMAGCODE2 (cleaning procedure code), t20 Ma (half-interval of age uncertainty). A subsequent post-filtering procedure was then applied to exclude paleopoles that gave a paleolatitude determination of more than 10^oabove or below the spline regression curve. Figure 2b shows an example of declination plot for the same reference point and set of plates. In this instance, 189 filtered paleopoles in the age range 0-200 Ma were used in the regression analysis. Finally, Figure 2c illustrates the APWP resulting from these paleolatitude and declination spline regression plots.

**Fig. 2.** A: Paleolatitude plot for a reference point in Central Africa at (0N,25E). B: Declination plot for the same location. C. Synthetic and smoothed APWP for Central Africa, obtained from the paleolatitude and declination plots. 165 and 189 filtered paleopoles in the age range 0-200 Ma from Central Africa, Northeast Africa, Madagascar, Somalia, India, Arabia, Australia, Antarctica, Brazilian Craton, Parana Block and Patagonia were used to generate plots (A) and (B).

The Plate Reconstruction Subsystem

Plate tectonic reconstruction maps require the association between a) a rotation model; b) a compilation of tectonic elements of continental lithosphere; and c) a compilation of additional paleogeographic features that are to be displayed (for example, mountains, magmatic provinces, age of the ocean floor, etc.).

A rotation model is built starting from a set of identified marine magnetic anomalies and associated fracture zones. Fracture zone trends determine, for each time interval, the location of the instantaneous Euler pole of rotation of a plate with respect to a conjugate plate. Magnetic anomalies determine the amplitude of motion, that is, the size of the angular velocity vector. However, actual rotation models include finite total reconstruction poles at anomaly times rather than instantaneous poles. These rotations represent the total rotation poles and angles of rotation that must be applied to a tectonic element to get the relative positions with respect to a reference plate at the anomaly times. Rotation models are tree structures that include total reconstruction poles for each pair of conjugate plates. Collisional settings at regional scale require a special treatment, because relative motions cannot be calculated using magnetic anomalies and fracture zones. In such difficult areas--for example, the Mediterranean region--indirect methods of analysis must be applied.

The rotation model used by the Web server is partly based on a digital compilation of oceanic isochrons of Royer et al. [1992] and the continental tectonic elements of the PALEOMAP Project [Scotese, 1990]. The North Atlantic and Arctic regions were modeled according to the rotation parameters of Rowley and Lottes [1988]. The "root" continent, Central Africa, was oriented according to a synthetic APWP in a standard paleomagnetic reference frame. Finally, the global model was integrated in some cases by detailed regional scale models of, for example, the Mediterranean, southeast Asia, and southwest Pacific.

We use a digital compilation of 223 tectonic elements for the Mesozoic and the Cenozoic. These blocks form the basic layer of any reconstruction and represent the distribution of the continental lithosphere through geologic time. The most difficult task was to obtain a reliable unstretched Continent-Ocean Boundary (COB) when a block was partly bounded by oceanic crust. This task was accomplished using gravity anomaly data [Sandwell and Smith, 1997] and, whenever possible, Moho depth and basement profiles generated using the Cornell Interactive Mapping Server (accessible by URL: http://atlas.geo.cornell.edu/ima.html).

Finally, additional paleogeographic features that can be viewed through the Web server are the age of the ocean floor and the distribution of Large Igneous Provinces (LIPs). A legend is available in a separate window. Apart from the above-mentioned digital compilation of tectonic elements of continental crust, the maps can display up to 2074 outlines of LIPs [Coffin and Eldholm, 1994] and 1031 polygons representing blocks of oceanic crust bounded by transform faults and isochrons. The LIP data set is a modified version of the digital compilation available via anonymous ftp from the University of Texas Institute for Geophysics [ftp://ftp.ig.utexas.edu/pub/LIPS/]. Most of the 1031 elements of oceanic crust were extracted from the latest release (version 1.5) of the digital ocean floor age grid [Müller et al., 1997], using the method described by Schettino [1999]. A complete display of the age of oceanic areas on both sides of a spreading center can only be obtained when the reconstruction time coincides with one of the anomaly times defined in the rotation model. These anomalies are: A5 (10.9 Ma), A6 (20.1 Ma), A13 (33.1 Ma), A18 (40.1 Ma), A21 (47.9 Ma), A25 (55.1 Ma), A31 (67.7 Ma), A34 (83.5 Ma), M0 (120.4 Ma), M4 (126.7 Ma), M10 (131.9 Ma), M16 (139.6 Ma), M21 (147.7 Ma), and M25 (154.3 Ma). Figure 3 shows a sample paleogeographic map generated by the Web server for anomaly A13. Map projection is Mollweide with center of projection at (30S,180E).

**Fig. 3.** Plate reconstruction for anomaly A13 (33.1 Ma). Map projection is an oblique Mollweide, with center of projection at 30S, 180E.

Future Plans

Future plans include a Java interface for visualizing velocity fields and flow lines, a system for selecting the display area, and the capability of selecting other reference systems (for example, hot-spot frames of reference).

References

Besse, J., and V. Courtillot, Revised and synthetic apparent polar wander paths of the African, Eurasian, North American and Indian plates, and true polar wander since 200 Ma, J. Geophys. Res., 96, 4029-4050, 1991.

Bocharova, N. Y., and C. R. Scotese, Revised global apparent polar wander paths and global mean poles, PALEOMAP Project Progress Rep. No. 56-1293, 20 pp., Department of Geology, University of Texas, Arlington, Tex., 1993.

Coffin, M. F., and O. Eldholm, Large igneous provinces: crustal structure, dimensions, and external consequences, Rev. Geophys., 32, 1-36, 1994.

Jupp, P. E., and J. T. Kent, Fitting smooth paths to spherical data, Appl. Stat., 36, 34-46, 1987.

Müller, R. D., W. R. Roest, J.-Y. Royer, L. M. Gahagan, and J. G. Sclater, Digital isochrons of the world's ocean floor, J. Geophys. Res., 102, 3211-3214, 1997.

Rowley, D. B., and A. L. Lottes, Plate-kinematic reconstructions of the North Atlantic and Arctic: Late Jurassic to Present, Tectonophys., 155, 73-120, 1988.

Royer, J. Y., R. D. Müller, L. M. Gahagan, L. A. Lawver, C. L. Mayes, D. Nürnberg, and J. G. Sclater, A global isochron chart, Univ. Texas Inst. Geophysics Tech. Rep. 117, 38 pp., University of Texas Institute for Geophysics, Austin, Tex., 1992.

Sandwell, D. T., and W. H. F. Smith, Marine gravity anomaly from Geosat and ERS 1 satellite altimetry, J. Geophys. Res., 102, 10,039-10,054, 1997.

Schettino, A., Computer aided paleogeographic reconstructions, Computers and Geosci., 24, 259-267, 1998.

Schettino, A., Polygon intersections in spherical topology: Application to Plate Tectonics, Computers and Geosci., 25, 61-69, 1999.

Schettino, A., and C. R. Scotese, A Synthetic APWP for Africa (Jurassic-Present) and Global Plate Tectonic Reconstructions, Eos Trans. AGU, 81, S180, 2000.

Scotese, C. R., Atlas of phanerozoic plate tectonic reconstructions, PALEOMAP Prog. Rep. 01-1090, 57 pp., Department of Geology, University of Texas, Arlington, Tex., 1990.