Processing of noise contaminated magnetotelluric data using digital filters based on MATLAB

Abstract

Magnetotelluric (MT) method is known as an effective geophysical method for imaging the condition of the subsurface resistivity distribution, especially in the geothermal system. However, this method is very sensitive to noise. The noise can occur due to mechanical activity such as vibration and electrical activity generated by powerline electric current near the point of measurement. To eliminate this noise is used digital filter consisting of despike filter and notch filter. Despike filter is used to eliminate irregular spike noise, while the notch filter is used to remove regular powerline noise. Furthermore, D+ can be applied to enhance the impedance data trend. Both of filters and D+ feature are built and then arranged into MTFilter program. The result of data processing using MTFilter program shows the improvement of MT data quality. Moreover, the inversion processing results of MT field data shows the resistivity distribution of structure which is more reliable with the characteristics of the geothermal system.