GRAVITY ANOMALY SEPARATION USING A COMBINATION OF GAUSSIAN AND BIDIMENSIONAL EMPIRICAL MODE DECOMPOSITION (BEMD) METHODS

• Main Author: Sahrul
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/77534
• Institution: Institut Teknologi Bandung

The separation of regional and residual components in the analysis of geological features through gravity methods is a widely employed technique by geoscientists worldwide. Although gravity anomaly separation methods have been continuously improving, no single method can be considered an absolute solution for obtaining the most accurate information about geological features. Therefore, there is always scope for refinement and further enhancement. The obstacle experienced by interpreters in interpreting geophysical data, especially gravity methods, is noise contamination in gravity data caused by geological features or errors in data acquisition. Therefore, in this research conducted various methods to separate gravity anomalies, such as the Gaussian filter, Bidimensional Empirical Mode Decomposition (BEMD), and a combination of Gaussian and BEMD techniques. The findings show that the integration of the Gaussian and BEMD methods proved to be more efficient in characterizing the near-surface gravity anomaly features in synthetic gravity data models containing noise, this conclusion is supported by the observed patterns and relatively lower Root Mean Square Error (RMSE) values, one example in the prismatic synthetic model I has an RMSE value of 0.1672, lower by 0.0237 and 0.4311 compared to the results of the Gaussian and BEMD methods respectively at 10% noise percentage noise, as well as with higher noise percentages of 20% and 30%. The next step is the application of the methods to the CBA data of the Kadidia region, which is a geothermal potential area in Central Sulawesi, Indonesia. The results obtained indicate that the integration of the Gaussian and BEMD methods is better at depicting the structure with the NW-SE trend and other structures with the NE-SW trend. Additionally, the obtained results also align with the position of the hot water manifestations situated at the boundary of low anomaly contrast and high anomaly contrast. As a result, this research offers enhanced opportunities for interpreting gravity anomalies more effectively. The procedure adopted in this study was developed utilizing the MATLAB programming language.