Indonesia Conference Directory

Noise that recorded at magnetoelluric data acquisition make the recorded data quality not good enough so the information that obtained after data processing might be not correct or not suit with the subsurface condition. Several characters that we can see at the noisy magnetotelluric data are the spiky shaped and non-stationarity time-series curve. This non-stationarity character can-t be handled by Fourier Transformation process. In this research used Empirical Mode Decomposition (EMD) method in original of Huang as filtering method in order to overcome the non-stationarity. This method decomposed the signal into group of oscillation mode called Intrinsic Mode Decomposition (IMF). One of the best IMF chosen as the filtering result by spectrum analysis in frequency domain. From the filtering result obtained that the time-series curve getting smoother and the non-stationarity character compressed. From this research concluded that EMD filtering can be implemented at magnetotelluric data processing and can emphasized the effect caused by noise.