Some comparisons among geomagnetic field models, observatory data and airborne magnetometer data: Implications for broad scale anomaly studies over Canada.

Abstract

The secular change characteristics of several geomagnetic field models have been compared with the observed secular changes at sixteen magnetic observatories in Canada and nearby regions, for the period 1960 to the present. The significance of inadequacies in the secular terms of the field models when used as reference fields for the compilation of large scale magnetic surveys is discussed. Errors of the order of 50 nT/year are encountered with some models. A model derived primarily from satellite data shows the best description of the secular change in the vertical component in western Canada during the interval 1969 to 1973, which includes the epochs of several airborne surveys. The revised IGRF75 describes the secular changes more closely than does the original IGRF (International Geomagnetic Reference Field), but cannot account for recent secular accelerations. Comparisons between field models and an extensive airborne data set show good agreement in the vertical component between the airborne survey data and two satellite-based models. Correction surfaces are developed to reduce the survey data to a common epoch, by considering differences between observatory monthly means and the corresponding values predicted by the field model.

Long-wavelength magnetic anomalies over Canada, using polynomial and upward continuation techniques.

“…The effect is small, but is important for long-wavelength studies, and the corrections have been applied to the data used in the subsequent analyses. COLES (1979) showed that the mean residual relative to POGO 6/71 for the combined surveys, after adjustment, was -12 T, with an r. m. s. deviation of 143 nT.…”

“…COLES (1979) has shown that inadequacies in the secular terms of the field models are sufficient to account for many, if not all, of these discrepancies. These effects are small enough that they do nott significantly affect the maps shown in Figs.…”

“…Where the shortwavelength field is weak, such as over the southern Canada Basin, the Arctic Islands, and the Western Cordillera, the long-wavelength field tends to be negative. POGO model is approximately a "best fit" reference in this region (COLES, 1979).) Thus regions of different local anomaly amplitude characteristics can produce a long-wavelength anomaly field: in northern latitudes, regions where high positive anomalies in close proximity predominate produce long-wavelength positive anomalies and regions of subdued magnetic relief produce longwavelength negative anomalies.…”

“…1 and 5, but they become important when considering wavelengths of order hundreds to a few thousands of kilometres where amplitudes can be a few tens of nanoteslas or less. COLES (1979) has demonstrated a way of correcting for these inadequacies. For each magnetic observatory in or near the survey region, the change in average quiet day field level from the epoch of the survey to an update epoch is compared with the change in value of a field model between the two epochs.…”

“…Langel, personal communications, 1975Langel, personal communications, , 1976Langel, personal communications, , 1978. This model is referred to in this paper and in COLES (1979) as POGO 6/71; the coefficients, under the name POGO 2/72, will be given in Langel, Coles and Mayhew (in preparation). Figure 4 shows the difference field in the vertical component between the POGO model and the IGRF, over Canada and environs.…”