MAGNETOTELLURIC DATA PROCESSING BASED ON HILBERT – HUANG TRANSFORM

• Main Author: DWI JATMIKO (NIM : 22315005), MARENDA
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/22967
• Institution: Institut Teknologi Bandung

Magnetotelluric (MT) is one of the most common and widely passive geophysical methods used in the exploration of natural resources. MT’s acquisition data is often practically contaminated by noise caused by natural processes or local objects surrounding the measuring site. Such phenomenon will make the MT data becomes a non-stationary random and can cause the estimated transfer function to be biased, therefore it required a special method to process the MT data. In this paper, MT data processing is performed using a non-stationary signal analysis method of Hilbert - Huang transform (HHT) with the aim of obtaining apparent resistivity and phase curve. Stages of MT data processing are conducted by decomposing the MT signal into several IMF components using empirical mode decomposition (EMD). Furthermore, analysis and elimination of baseline drift noise was done before the calculation of instantaneous spectrum (IS) values. The value of IS in the time – frequency domain that has been obtained will be used to estimate the value of the MT transfer function. HHT’s method test is performed on two synthetic data, free noise data and noisy data of 10% to 20%, and also on the real data to evaluate the effectiveness of this method. The results of test in these data indicate that the estimated MT transfer function obtained by the HHT method is more stable than the FT method. The HHT method is able to overcome the problem of non – stationary characteristic of MT data which can lead to bias estimation results.