DATA PROCESSING SCHEME FOR DATA WITH PERIODIC AND SPORADIC NOISE AND ITS APPLICATION ON ELECTROMAGNETIC

• Main Author: FITRIA DEWI MARIETA (NIM: 20216005), WINA
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/31636
• Institution: Institut Teknologi Bandung

<p align="justify">The recorded data in acquisition process of geophysical methode is not purely from the response of the subsurface material, but there is also noise from other sources that lead to decreased quality of data. Noise in data can be divided into periodic and sporadic noise. In this research, there are several steps to suppress both types of interference by using a data processing scheme consisting of three stages: prestack, selective stacking, and poststack that have been modified. Based on data processing performed on synthetic and magnetoteluric (MT) data, periodic noise can be handled using recursive filters, and sporadic noise can be handled in the selective stacking stage. There are four different approaches of selective stacking that used. The first approach is an average process that produces an RMS error with the amount of 2.15, the second approach is median which gives an RMS error with the amount of 0.57, the third approach is symmetric rejection that produces an RMS error with the amount of 0.53, and the last method is M estimation that produces an RMS error with the smallest value with the amount 0.49. The application of selective stacking in the MT data shown that the average estimate yields a smoother apparent resistivity and phase curve than the estimated impedance or data processing using the software.<p align="justify"> <br />