Estimation of the systematic error of three-component geomagnetic data using the ABIC minimization method

✍ Scribed by Yuzo Marukawa; Hiroyuki Kamei

• Publisher: John Wiley and Sons
• Year: 1999
• Tongue: English
• Weight: 705 KB
• Volume: 6
• Category: Article
• ISSN: 1075-2196
• DOI: 10.1002/(sici)1099-0763(199909)6:3<135::aid-arp114>3.0.co;2-k

✦ Synopsis

Geomagnetic data are often affected by a systematic error caused by the posture error and the temperature change in an instrument, etc. The vector field measurements using the threecomponent fluxgate gradiometer are more sensitive to the posture and temperature changes than total field measurements.

An approach to estimate and remove the systematic error in the vector geomagnetic data is proposed. Firstly, the systematic error is modelled by a linear function and the magnetic anomaly caused by a subsurface structure is modelled by a cubic B-spline function. Then an equation is solved to determine the model parameters of the systematic error and magnetic anomaly. A trade-off between the misfit residual and the roughness of the magnetic anomaly is considered by use of Akaike's Bayesian information criterion (ABIC).

The method is applied to a vector geomagnetic data set that was measured using the threecomponent fluxgate gradiometer at an archaeological site in Japan. The results are compared with a conventional method. The comparison shows the effectiveness of the proposed method.