Indra Rivaldi Siregar
Geophysical Engineering | Geoscience Computing | Data Enthusiast
Geophysical Engineering | Geoscience Computing | Data Enthusiast
A fresh graduate of Geophysical Engineering with a GPA of 3.79. Highly interested in data science and geoscience computing.
Possesses experience of more than 3.5 years using Python with a variety of useful basic program outputs for seismological processing, RF processing, and geoscience mapping.
Has 2 national achievements in data science, both involving the implementation of machine learning in geoscience. A person with good communication and leadership skills.
My thesis discussed receiver function analysis to compute the crustal thickness and Vp/Vs ratio beneath a three-component station using Python. BMKG and InaTEWS invited me to participate in Receiver Function Processing and Analysis Training as a trainer or speaker. My python based program includes: retrieving the data from the data center website, filtering data, deconvolution, removing outliers, stacking data, Zhu-Kanamori method, and mapping. These scripts supported by Obspy and RF packages.
See all scripts
The shear wave velocity is one of the most important parameters in oil and gas exploration for determining the physical properties of rocks (Vs). However, due to budget constraints, this data is not always available during the investigation stage. Machine learning can be utilized as an alternate way for estimating shear wave velocity. The KNN, ANN, and SVR algorithms were used to create the machine learning model, with hyperparameter tuning used to discover the ideal hyperparameters for each algorithm.
See all scripts
As a speaker at Python mapping training, I demonstrated how to create a variety of maps, including basic maps, projection maps, station maps, earthquake maps, topography maps, cross-sections, contour maps, block mean maps, and others. All maps created could use both offline and online data.
See all scripts
Some parameters used to build the machine learning model were slope, NDVI, elevation, land cover, and curvature.
See all scripts See publication
These scripts are utilized to calculate the a-value and b-value (seismological terms) to characterize the subsurface. In this case, we used for identifying fracture characteristics of stimulation process in The FORGE, Utah, U.S. The publication from this project is still under review.
See all scripts
The machine learning model can be used to predict the lithology of rocks. Several features from well log data, such as GR, RHOB, Vp, and depth, can be used to predict lithology. In this case, after the low-pass filtering process, the generated model was optimal with a high accuracy score.
See all scripts