A Comparison of Probabilistic Seismic Hazard Maps to Shakemap Footprints in Indonesia

Date: 4/24/2019

Time: 06:00 PM

Room: Grand Ballroom

This study provides the results of testing three probabilistic seismic hazard assessment (PSHA) PSHA maps: the Global Seismic Hazard Assessment Program (GSHAP; Giardini et al. 1999), the Global Assessment Report 2015 (GAR2015; International Strategy for Disaster Reduction 2013) and the Standar Nasional Indonesia (SNI2017; Irsyam et al. 2017). The observations are earthquakes occurred in Indonesia between May 1968 and May 2018 and modelled by the USGS Shakemap program. Despite they are not as accurate and complete as continuous recorded ground motions, the 959 collected Shakemap footprints cover 7,642,261 grid points, at 1km² resolution, when only 126 seismic stations are currently operating in Indonesia.

After verifying the representativeness of Shakemap footprints for the maximum historical peak ground acceleration (PGA) between May 1968 and May 2018, this data has been extracted from Shakemap footprints for 177,674 sites with rock conditions and 232 different mainshocks, according to a new process, developed in this study. Then, PSHA maps are tested by selecting randomly one site per mainshock to meet the condition for testing of independency between PGA observations. Furthermore, to ensure that a dataset made of 232 values is large enough for statistical analyses, a new relationship is introduced between the minimum number of observations needed, the return period of the PSHA map tested and the observation period length.

The three PSHA maps are finally tested for both the whole Indonesia and only the Western part of Indonesia. Results showed that the SNI2017 map is consistent with the last 50 years seismicity over the whole country, but overestimates it when focusing on Western Indonesia. At the opposite, the GSHAP and the GAR2015 PSHA maps fit the past seismicity on Western Indonesia, but underestimate it when considering the whole country. This study concludes that using Shakemap footprints for a large country, prone to earthquake risk in its entirety, enables to differentiate PSHA maps by using past seismicity for testing.

Presenting Author: Adrien Pothon

Authors