ANALYSIS SPATIAL INTERPOLATION PARAMETERS ORDINARY KRIGING AND INVERSE DISTANCE WEIGHTED (IDW) FOR GRAVITY ANOMALY

• Main Author: Ashhabul Kahfi, Nadhif
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/76747
• Institution: Institut Teknologi Bandung

Basically, geophysical data obtained from the acquisition process consists of location values of measurement points in the form of coordinates and measured values of gravity acceleration (mGal) as attributes in the interpolation process, making the geophysical data spatial data. The aim of this research is to evaluate the most used spatial interpolation methods, namely Ordinary Kriging as a stochastic interpolation method, by analyzing the use of estimation and variogram modeling parameters, and Inverse Distance Weighted (IDW) as a deterministic interpolation method, by analyzing the use of Radius and power parameters in generating the CBA Anomaly which will be separated using the Moving Average Method. The evaluation process will be based on quantitative observations through the best prediction error, correlation, and determination coefficient values generated in the Cross Validation process, as well as qualitative evaluations based on the interpolation results from the selected parameters used in various ways. The data used in this study are the results of measurements from the Field Study of Geophysical Engineering at ITB in the Karangsambung region, Kebumen Regency, Central Java on July 31 to August 3, 2019. The quantitative spatial interpolation results in estimating the CBA Anomaly map by the IDW method show that the best Radius parameter used is 600 meters and the best power parameter is 3, while using the Ordinary Kriging method shows that the best variogram model parameter is demonstrated by using the Gaussian model with a maximum lag of 600 (sill: 280, range: 15, nugget effect: 2.5, error variance: 2.5), lag tolerance 0, angle tolerance 90°, and interval between lags of 30. The prediction error values for each method are 1.10 and 1.26 (MAE), 3.95 and 5.06 (MSE), 1.99 and 2.25 (RMSE), and 1.99 and 2.25 (STDV), respectively. Furthermore, qualitatively, it can be observed from the interpolation results of the CBA Anomaly with the parameters used that the Ordinary Kriging method does not significantly impact the shape and position of the gravity anomaly body from various variogram models. On the other hand, the IDW method, specifically with a power of 1, provides a significant difference compared to other power values in the shape and position of the gravity anomaly body, therefore both of spatial interpolation methods for gravity data given results in deterministic interpolation classification caused have a minimum of probabilistic.