Identification of PGA Outliers Using the IQR Method to Assess Potential Local Amplification
Jurnal Pendidikan Fisika dan Teknologi · 10.11.2026 · 12 (2) · 613-622
DOI: 10.29303/jpft.v12i2.13119 ↗Kurz gefasst
Fachartikel in der Zeitschrift Jurnal Pendidikan Fisika dan Teknologi, online erschienen am 10.11.2026. Verfasst von Annisa Aulia Putri und 2 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY-SA 4.0).
Abstract (Originalsprache)
Peak Ground Acceleration (PGA) is an important parameter in earthquake hazard analysis. However, its values may deviate from general trends due to local site conditions. This study aims to analyze the distribution of PGA recorded by BMKG stations in Northern Sumatra during the 2017–2025 period, detect statistical outliers in relative PGA residuals using the Interquartile Range (IQR) method and boxplots, and identify preliminary indications of potential local amplification based on Vs30 values obtained from USGS maps. The analysis used 5,360 records from 501 events recorded by at least three stations, following data cleaning and event selection from an initial dataset of 5,567 records. Horizontal amplitudes were calculated from the east–west and north–south components, transformed using the natural logarithm, and then compared with the median ln(PGA) of each event as a relative reference to obtain relative residuals. The initial PGA distribution was highly right-skewed (skewness = 19.71; kurtosis = 686.97) and became closer to a log-normal distribution after transformation. The IQR analysis yielded Q1 = 0.4782, Q3 = 2.1105, IQR = 1.6323, and an upper bound of 4.5590, identifying 639 records (11.92%) as outliers across 69 stations. ONSM, SKSN, LTSM, PSSN, and SNBI exhibited the highest outlier frequencies. Nine of the ten stations with the highest outlier frequencies belonged to NEHRP site classes C and D, indicating that stiff to soft soil conditions may contribute to elevated PGA responses. However, the relationship between relative residuals and Vs30 was not entirely consistent, suggesting that Vs30 alone is insufficient to explain the observed variations. Relative residuals do not explicitly account for the effects of source-to-station distance, wave propagation paths, or earthquake source characteristics. Therefore, the identified outliers are interpreted as potential anomalies in PGA response rather than confirmed evidence of local site amplification. Further research is recommended using attenuation models or distance and magnitude normalization, along with field-based geophysical investigations to confirm the underlying causes.
Zitieren
APA
Putri, A. A., Lubis, L. H., & Ong, R. (2026). Identification of PGA Outliers Using the IQR Method to Assess Potential Local Amplification. Jurnal Pendidikan Fisika dan Teknologi, 12(2), 613-622. https://doi.org/10.29303/jpft.v12i2.13119
Vancouver
Putri AA, Lubis LH, Ong R. Identification of PGA Outliers Using the IQR Method to Assess Potential Local Amplification. Jurnal Pendidikan Fisika dan Teknologi. 2026;12(2):613-622. doi:10.29303/jpft.v12i2.13119
BibTeX
@article{Putri2026,
author = {Putri, Annisa Aulia and Lubis, Lailatul Husna and Ong, Russell},
title = {Identification of PGA Outliers Using the IQR Method to Assess Potential Local Amplification},
journal = {Jurnal Pendidikan Fisika dan Teknologi},
year = {2026},
volume = {12},
number = {2},
pages = {613--622},
doi = {10.29303/jpft.v12i2.13119}
}
EUVIMED