In August 1981 the International Association of Geomagnetism and Aeronomy revised the International Geomagnetic Reference Field (IGRF). It is the second revision since the inception of the IGRF in 1968. The revision extends the earlier series of IGRF models from 1980 to 1985, introduces a new series of definitive models for 1965-1975, and defines a provisional reference field for 1975-1980. The revision consists of: (1) a model of the main geomagnetic field at 1980.0, not continuous with the earlier series of IGRF models, together with a forecast model of the secular variation of the main field during 1980-1985; (2) definitive models of the main field at 1965.0, 1970.0, and 1975.0, with linear interpolation of the model coefficients specified for intervening dates; and (3) a provisional reference field for 1975-1980, defined as the linear interpolation of the 1975 and 1980 main-field models. The new models are series of solid spherical harmonics up to and including the tenth degree and order for the main-field models, and up to and including the eighth degree and order for the secular variation model. The models were derived from three sets of proposed models by taking weighted means. The weights were chosen according to the apparent accuracy of the proposed models. A brief history of the IGRF, a review of basic formulas, and a set of world contour maps of the geomagnetic elements based on the IGRF 1980 model are included.

抄録全体を表示

Models of the main (core) magnetic field are limited in accuracy by the nature of the data. Anomaly fields from the crust are a noise source in these calculations which restrict the accuracy to which the main field coefficients can be determined. For meaningful coefficients the degree and order of the model is limited to about eight for observatory and other surface data and to about 13 for satellite data. Utilizing satellite and surface data together has permitted the incorporation of a solution for the anomaly field at each observatory. These fields vary from a few tens of nT (nanotesla) to several thousand nT. The residuals of the observatory measurements to such models is no longer dominated by the anomaly fields and so is commensurate with the actual measurement accuracy (about 5 to 20nT, depending on the observatory). Incorporation of the anomaly estimation has made possible the inclusion of stable time derivatives of the spherical harmonic coefficients up to the third derivative. Using these new techniques a spherical harmonic model, designated GSFC (9/80), was derived with degree and order 13 in its constant and first time derivative terms, six in its second time derivative terms, and four in its third derivative terms. The data used in deriving the model included (1) 15, 206 Magsat observations, (2) 71, 000 POGO satellite observations, (3) measurements from 148 magnetic observatories, (4) 300 measurements from filtered marine magnetic data, and (5) approximately 600 measurements from 150 selected repeat stations. The time span of the data was 1960 through 1980. RMS residuals of the model (including anomaly values) to the data are: Magsat scalar data, 10nT; Magsat vector component data, about 7nT; POGO data, 7nT; observatory X component, 38nT; observatory Y component, 19nT; and observatory Z component, 17nT.

抄録全体を表示

In accordance with the rules adopted at the IAGA meeting in Canberra in 1979, we proposed for adoption as the new International Geomagnetic Reference Field a series of main-field models to degree and order ten for 1965, 1970, 1975, and 1980, and a secular variation model for the interval 1980 to 1985. These models are briefly described.

抄録全体を表示

A set of spherical harmonic models describing the Earth's main magnetic field from 1965 to 1985 has been developed and is proposed as the next revision of the International Geomagnetic Reference Field (IGRF). A tenth degree and order spherical harmonic model of the main field was derived from Magsat data. A series of eighth degree and order spherical harmonic models of the secular variation of the main field was derived from magnetic observatory annual mean values. Models of the main field at 1965, 1970, 1975, and 1980 were obtained by extrapolating the main-field model using the secular variation models.

抄録全体を表示

Three candidate models for the International Geomagnetic Reference Field (IGRF) for the years 1965, 1970, 1975, and 1980 are evaluated using magnetic observatory and satellite data. The models were submitted by the USGS, the IGS, and NASA. Comparison based on annual means data shows the differences between models to be small, although the NASA model appears slightly more accurate in 1965 and 1975 and the IGS model in 1970. Evaluation with MAGSAT scalar and vector satellite data at 1980 shows the USGS and NASA models to be of comparable accuracy, while the IGS model displays larger residuals. The NASA model best represents the scalar POGO data at 1965 and 1970. Contoured differences between the models in components X, Y, Z, and F at the earth's surface at 1965, 1970, 1975 and 1980 are displayed and discussed.

抄録全体を表示

The IGRF candidate models for the main field 1980 and the secular variation 1980-85 are compared with observations. The IGS and USGS secular variation models agree better with the global set of secular variation estimates than does the NASA model. The NASA and USGS models represent more closely than does the IGS model a spatially limited set of 1980 data.

抄録全体を表示

For the second revision of the International Geomagnetic Reference Field (IGRF), the U. S. National Aeronautics and Space Administration (NASA), the U. K. Institute of Geological Sciences (IGS), and the U. S. Geological Survey (USGS) submitted proposed models of the Earth's main magnetic field at 1965.0, 1970.0, 1975.0, and 1980.0, and its secular variation during 1980-1985. We assessed the proposed models by comparing them with annual mean values from worldwide magnetic observatories, data for 1978-1980 from 63 U. S. magnetic repeat stations, and rates-of-change values for worldwide magnetic observatories for 1965-1985 that were derived from straight lines fitted to annual means for five-year intervals. We also mutually compared the 1980 models.
For the observatory data, the IGS 1980 model yields an absolute mean residual in X that is more than twice as great as that for any other model for any year, and that model yields an RMS residual in Z of 246nT, compared with less than 193nT for the other two 1980 models. For the repeat station data, the IGS 1980 model yields an absolute mean residual in F of 113nT, compared with about 30nT for the other two. The secular variation of Z for 1975-1980 deduced from the IGS models, and for 1980-1985 from the NASA secular variation model, yield RMS residuals of 32 and 21nT/yr, respectively, when compared with the observatory rates-of-change values. The NASA and USGS 1980 models, both based on MAGSAT data, differ by no more than about 70nT, but the IGS 1980 model differs from the other two by more than 600nT in some places. This study suggests that the second revision of the IGRF should consist of means of the proposed models, with the exceptions of the IGS 1980 and the NASA 1980-1985 models.

抄録全体を表示

Comparisons of proposed and adopted geomagnetic reference field models are made against North American test data for the period 1965 to 1985. These test data (X, Y, Z, and F) include Canadian map and aeromagnetic data and USGS land-based data. Model values are also tested against predicted component values for 1982 and 1984 derived from 27 magnetic observatories on or near the North American continent.
These tests indicate that all candidate models provide an excellent fit to North American test data at epochs 1965, 1970, and 1975. Only the USGS and NASA models fit these data satisfactorily at 1980. The best secular variation models for 1980-1985 appear to be the IGS and USGS models.
Finally, our tests indicate that the recently adopted series of reference field models provides a reasonable and satisfactory fit to North American test data.

抄録全体を表示

To compare three sets of models submitted by IGS, USGS, and NASA as proposed revisions of the International Geomagnetic Reference Field, the root mean square amplitudes over the Earth's surface were computed for the main field (MF) and secular variation (SV) models and for differences between them. The results obtained for the total field and for the nth-degree harmonics showed that the IGS 1975-80SV model, the IGS 1980MF model, and the NASA 1980-85SV model differ significantly from the other corresponding models. It is proposed that IGRF main field models for epochs 1965, 1970, and 1975 be derived from means of all three candidate models; for 1980, the mean of the USGS and NASA models would be used. The secular variation model for 1980-85 would be taken as a mean of the submitted USGS and IGS SV models.

抄録全体を表示