Evaluation over Australia of Candidate Models for IGRF90

• Published in: Journal of geomagnetism and geoelectricity Vol. 44; no. 9; pp. 843 - 854
• Main Authors: BARTON, C. E., WINCH, D. E., MCEWIN, A. J., HITCHMAN, A. P., HOPGOOD, P. A.
• Format: Journal Article
• Language: English
• Published: Tokyo Society of Geomagnetism and Earth, Planetary and Space Sciences 01.01.1992; Terra
• Subjects: Earth sciences; Earth, ocean, space; Exact sciences and technology; Geophysics: general, magnetic, electric and thermic methods and properties; Internal geophysics; Australia; Australasia; Reference model; magnetic field; Geomagnetic field; Updating; accuracy; secular variations; spherical models; revision
• ISSN: 0022-1392; 2185-5765
• DOI: 10.5636/jgg.44.843

The five candidate field models for IGRF90, and the actual IGRF90 model adopted, are compared with AGRF90, the regional field model for Australia. Similarly, the three candidate secular variation models for 1990 to 1995 are compared with observatory and magnetic repeat station data for the region. All the candidate models, and the adopted IGRF90, provide an acceptable fit to the Australian model and observational data. That said, the US90 (U. S. Geological Survey) field model for 1990 gives slightly better agreement with AGRF90 than do the other models; the US90 secular variation model clearly gives the closest fit to the secular variation estimates derived from direct observations; and the BN90 (UK/US) secular variation model is the least consistent with the other candidate models and the observations. The south magnetic (dip) pole positions given by the candidate models, and the adopted IGRF models, are closely clustered together for each epoch. With the exception of the UK/US models, the poles all lie within a radius of about 3.5km for each of the epochs 1985, 1990 and 1995. Dip-poles of the UK/US models for 1985, 1990 and 1995 lie about 11 to 14km north of the others. Dip-poles of the adopted DGRF85 and IGRF90 models form a uniform, smooth continuation of the locus of dip-poles of earlier IGRF models. We conclude that the US90 main-field and secular variation models give the closest fit to the regional field model and data, and that the model adopted by IAGA Working Group V-8 for IGRF90 is a good compromise for the Australian region and in the vicinity of the south magnetic pole.