Return to Science and Society
Return to Starting Point
For the past 40 years or more an error has been appearing on the charts produced by major mapping organizations of the United States. The cartographers have been routinely marking a position in each hemisphere that they identify as the "Magnetic Pole" (e.g. the maps "Illustrated World Atlas Set: Atlas of Continents", Rand McNally & Company, pg 47, 1996; and "The Earth's Fractured Surface", National Geographic Society, April 1995). The problem is that the word "magnetic" has become a generic term and the two words together can include several types of poles with greatly differing positions depending upon their derivation. The most used pole position is certainly not that indicated on the charts. Nor is the generally understood position, actually the one that the cartographers have marked. The non-specific naming that is presently used by our nation's map-making industry is a not a trivial mistake. For the many scientists working in other than a geomagnetism specialty and for the ordinary citizens, the present magnetic pole marking only leads to a confusion as to what is being located at the given positions. The major global cartographers need to become aware of the modern science of geomagnetism and either make their maps more meaningfully or remove their dubious "Magnetic Pole" locations all together.
A brief "magnetic pole" tutorial is needed to show why the confusion arises; let me start with a discussion of field modelling. Geomagnetic field models are prepared by a variety of organizations and countries. For a particular analysis year (epoch) a field fitting procedure called "Spherical Harmonic Analysis" (SHA) allows the model contributions from above (external) and below (internal) the spherical surface of the analysis to be separately represented by a double series of Legendre polynomial coefficients, identified by increasing "degree" and "order" numbers. These series are truncated at some limiting degree and order either because the data errors restrict significance levels or because the smaller amplitude terms of higher degree and order number are thought to be due to local Earth crustal effects and induced currents. When the model is from an Earth-centered analysis, and only the internal parts are considered, the table of representative numbers are called "Gauss Coefficients" of a geomagnetic model field, to honor the Goettingen (Germany) scientist who first developed the main field modelling technique.
In our country there are field models prepared by NASA, by the U.S.Navy, and by the U.S. Geological Survey; all differing somewhat in the manner of processing and in the selection of data for the model fitting (data from surface observatories, from special surveys over land and sea, and from satellites). After an exacting study of the various national field models, a select committee (Working Group 8, Division V) of the International Association of Geomagnetism and Aeronomy agrees upon the best compromise Gauss coefficients for the recorded measurements of the magnetic field about the Earth. That committee (current chairman: Dr. Charles Barton, AGSO, GPO Box 378, Canberra, ACT 2601, Australia) fixes what is called the "International Geomagnetic Reference Field" (IGRF) for every five-year epoch since 1900. An estimate of the change-per-year for each coefficient allows projection of the model to later years. If any IGRF model is subsequently adjusted by agreement, the replacement is called the "Definitive Geomagnetic Reference Field" (DGRF). The present models truncate at degree and order 10 (see table published in J. Geomag. Geoelectr.,47, 1257-1261, 1995). This truncation limits the fitting to the longer wavelengths (about 36 ) and, together with the restriction to select only the internal field terms, is responsible for some of the difference between the locally measured fields and those fields computed from the model. It is important to remember that at a particular location, fields reconstructed from world models depend on the global spherical harmonic wave fitting to all the observations, throughout the world, that went into the model.
The first three numbers (Gauss coefficients) in the column of values depicting, for each epoch, the IGRF (or DGRF) model represent, on average, about 90% of the field contribution (although this percentage varies considerably with location). These three values are all that is required to describe an Earth-centered dipole. It has been found quite useful to establish a coordinate system using such a centered dipole field; geophysicists call this the "Geomagnetic Coordinate System." Positions, symmetric in the two hemisphere, where the dipole would intersect the Earth's surface are called the "Geocentric Dipole Poles" or more simply the "Geomagnetic Poles." For mid 1996 these positions in the north and south polar region are 79.3 North, 71.5 West and 79.3 South, 108.5 East respectively. The geomagnetic coordinate system is completely defined by one pole location and a selection of a prime meridian of geomagnetic longitude (by agreement this is the meridian that intersects the geographic South Pole). Geomagnetic coordinates are used universally to organize a great many types of geophysical data, particularly those for solar-terrestrial and magnetospheric scientists. For that reason the Geomagnetic Pole locations should have a prominent appearance on charts. The pole positions and geomagnetic coordinates of a location can be obtained from the GMCORD program available on World Wide Web ( www.ngdc.noaa.gov/seg/potfld/geomag.html ). The Geomagnetic Pole positions are far from where the commercial cartographers mark their so called "Magnetic Poles."
If the full 10 degree-and-order terms of the IGRF are used to determine where, in the polar regions, the field seems to be most vertical, then this place is called the "IGRF Model Magnetic Dip Pole" position or more simply the "Model Dip Pole". The "dip" in the name arises because, since the early days of geomagnetism, a special "dip magnetometer" has been available (essentially a long magnetized needle suspended about a horizontal axis through its mid position) to determine the angle at which the Earth's field pointed into the Earth. But this Model Dip Pole position depends upon which model is used for the Gauss coefficients of the spherical harmonic analysis obtained from the best fitting from all the observatories. The 1995 IGRF Model Dip Pole north and south positions for mid 1996 are located at 79.0 N, 105.1 W and -64.7 S, 138.6 E respectively. Such positions can be more than 30 in longitude from the geomagnetic poles! For North American scientists the IGRF model field values can be obtained from the US Geological Survey (J. Quinn, USGS Mailstop 968, Box 25046, Denver, CO 80225).
Another "magnetic" pole position can be found. Instead of calculating it from the IGRF model, a ground survey of the relevant polar regions can be carried out to find, by actual measurement, where the field is most fully vertical (90 ) in dip. Such a survey takes into account all the locally present fields (external and internal) including the special field effects (induced currents and permanent magnetization) unique to the geology in the region of the interest. The measurement isn't concerned with the contributions from other world observatories (as is the IGRF Model Dip Pole). The determined position is the true "Magnetic Dip Pole" and would be an unique location for a chart because it is where a person, travelling in the region could observe a vertical field. (It is the place that Mr. Average would expect his polar region chart to have a corresponding mark). This "Magnetic Dip Pole" is not at the same location as the other two "magnetic" type poles mentioned above; it can be several degrees or more away from the IGRF model dip pole; the difference changes with time. Magnetic Dip Poles are of importance for ionospheric physicists because radiowave behavior in the region responds to the nearby magnetic fields. The locations of Magnetic Dip Poles have been exactly determined only rarely over the years.
A further mapping confusion occurs because the Earth's main field is constantly changing in size and position. The source of field is traced to flow patterns generated in the deep Earth's liquid outer-core and core-mantle boundary processes. The pole positions at the Earth's surface are moving westward (secular variation) several tenths of a degree per year (at speeds depending on such things as the number of harmonics included in the analysis). That is why the models are recomputed about every five years. So, to have significant meaning, each pole position also needs to be identified with the model epoch or measurement date. The elusive "Magnetic Poles" marked on the commercial charts never appear with a date. The user is left to hope that the map publication date gives the appropriate dipole year.
Under the generic term "Magnetic Poles" there are many pole positions that have more exact names appropriate to the type of analyses used in their definition or unique to a special scientific discipline. In addition to those mentioned above there are "Eccentric Axis Dipole Poles" and "Eccentric Axis Dipole Dip Poles" (obtained from a modelling that allows a repositioned SHA analysis center to maximize the dipole components), as well as "Virtual Geomagnetic Poles" (obtained from field directions measured in continental rock samples). However, the three sets of poles described in detail above are probably the most prominent in geophysics. Which of these three is the "Magnetic Pole" on the Rand McNally and National Geographic charts? Possibly, the cartographers may use the IGRF Model Dip Pole to locate where they would like to mark a "Magnetic" pole. However, it is difficult to tell what is indicated under their generic term. Even if it is a Model Dip Pole, there should be some indication of what model and which epoch date.
I sincerely hope that the industry response to this EOS article about their magnetic pole designation will inspire the cartographers become enlightened on the subject of magnetic poles (c.f. Fraser-Smith , 1987) and provide a more accurate marking. However, the reaction may be "no one else has ever complained" or "every other world chart manufacturer is doing the same" or "Dr. XXX assures us that our identification of 'Magnetic Pole' is quite correct and to change it after such long use would only cause confusion". Any of that logic might also be applied to the many pre-Columbian, flat-world maps that were produced into the 15th century.
References
Fraser-Smith, A.C., Centered and eccentric
geomagnetic dipoles and their poles , 1600-1985, Rev.
Geophys., 25, 1-16, 1987.