Optimizing Interpolation Methods and Point Distances for Accurate Earthquake Hazard Mapping in Gas Pipeline Routing

Abstract

When optimizing the route for gas pipelines while taking seismic risk into account, a vital first step is assessing the seismic hazard at every point within the region that the route must be passed. It's crucial to conduct a seismic hazard analysis for each specific point. However, achieving a continuous assessment of seismic hazard for all points is practically unfeasible. Ultimately, a practical approach involves directly determining hazards in certain points, while employing interpolation for the remaining points. This approach involves deliberating the spacing of grid points for hazard determination and selecting an appropriate interpolation method for estimating hazards in the remaining points. The distance of the points and the interpolation method for each problem must be determined by trial and error. This article explores various point distances and interpolation methods in a case study context. To assess accuracy, this study explores 15 different modes of point distances and employs two distinct interpolation methods—Inverse Distance Weighting and Ordinary Kriging Interpolation models. The selection of point distances is based on a percentage of longitude and latitude. Values from 0.02 to 0.3 are considered. A baseline distance of 0.02 is selected, and other distances and interpolation methods are compared against it. Four statistical indicators—mean error, mean absolute error, root mean square error, and coefficient of determination—are considered. Results indicate that the ordinary Kriging Interpolation Method achieves higher accuracy. Considering error percentages and the apparent similarity of hazard maps, a distance of 0.14 points appears suitable, balancing computational time and required accuracy.