Terrain Classification of Aster gDEM for Seismic Microzonation of Port-Au Haiti, Using - and - Based Analytic Methods
Abstract
The aftermath of the M7.0 Haiti earthquake of 12 January 2010 witnessed an impressive scientific response from the international community. In addition to conventional post-earthquake investigations, there was also an unprecedented reliance on remote-sensing technologies for scientific investigation and damage assessment. These technologies include sensors from both aerial and space-borne observational platforms. As part of the Haiti earthquake response and recovery effort, we develop a seismic zonation map of Port-au-Prince based on high-resolution satellite imagery as well as data from traditional seismographic monitoring stations and geotechnical site characterizations. Our imagery consists of a global digital elevation model (gDEM) of Hispaniola derived from data recorded by NASA-JPL's Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) instrument onboard the multi-platform satellite Terra. To develop our model we also consider recorded waveforms from portable seismographic stations (Hough et al., in review) and 36 geotechnical shear-wave velocity surveys (Cox et al., in review). Following a similar approach developed by Yong et al. (2008; Bull. Seism Soc. Am.), we use both pixel- and object- based imaging analytic methods to systematically identify and extract local terrain features that are expected to amplify seismic ground motion. Using histogram-stretching techniques applied to the rDEM values, followed by multi-resolution, segmentations of the imagery into terrain types, we systematically classify the terrains of Hispaniola. By associating available Vs30 (average shear-wave velocity in the upper 30 meter depth) calculated from the MASW (Multi-channel Analysis of Surface Wave) survey method, we develop a first-order site characterization map. Our results indicate that the terrain-based Vs30 estimates are significantly associated with amplitudes recorded at station sites. We also find that the damage distribution inferred from UNOSAT (UNITAR Operational Satellite Applications Program) data matches our estimates. However, the strongest amplifications are observed at two stations on a foothill ridge, where Vs30 values indicate that amplification should be relatively lower. Hough et al. (2010, this session) conclude that the observations can be explained by topographic amplification along a steep, narrow ridge. On the basis of these preliminary results, we conclude that the terrain-based framework, which characterizes topographic amplification as well as sediment-induced amplification, is needed to develop a microzonation map for Port-au-Prince.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2010
- Bibcode:
- 2010AGUFM.U13A0012Y
- Keywords:
-
- 7203 SEISMOLOGY / Body waves;
- 7212 SEISMOLOGY / Earthquake ground motions and engineering seismology;
- 7255 SEISMOLOGY / Surface waves and free oscillations;
- 7294 SEISMOLOGY / Seismic instruments and networks