DBPapers
DOI: 10.5593/sgem2017/14/S05.015

ENHANCING EARTHQUAKE SHAKEMAPS FOR ROMANIA, BY USING AZIMUTHAL ANALYSIS AND CROWD-SOURCED INTENSITY DATA

D. Toma-Danila, C. O. Cioflan
Monday 11 September 2017 by Libadmin2017

References: 17th International Multidisciplinary Scientific GeoConference SGEM 2017, www.sgem.org, SGEM2017 Conference Proceedings, ISBN 978-619-7408-00-3 / ISSN 1314-2704, 29 June - 5 July, 2017, Vol. 17, Issue 14, 119-126 pp, DOI: 10.5593/sgem2017/14/S05.015

ABSTRACT

ShakeMaps are one of the most applicative products of seismological research and monitoring, becoming available minutes after earthquakes. They are maps depicting the ground shaking produced by an earthquake and the estimated intensity distribution, being extremely useful for authorities, population, private sector and researchers too. However, the outcome of the result – consisting of easy to read maps representing the potential area affected, often hides the fact that many compromise decisions and uncertainties are included in the methodology. Such decisions are related for example to the way real-data is handled (some values can be modified more or less, in order to match ground-motion prediction equations - GMPE, at rock or soil level) or the way GMPEs or interpolation methods are selected. In Romania, the ShakeMap system developed by USGS is implemented, but due to the characteristics of the Vrancea Intermediate-Depth Source (which is the most active and hazardous in the country), there is an apparent mismatch with the typical algorithm developed mainly for crustal earthquakes. That is why we chose to replicate the ShakeMap methodology partially, turn it into an open-source Matlab code and test new refinement procedures that can be of use in other cases of custom needed implementation. Among the new features that we describe in this paper and for which we present results in comparison are: i) the integration of new GMPE’s for Vrancea earthquakes; ii) a new azimuthal dependent procedure for proper GMPE selection considering real recordings; iii) a dynamic module for integrating crowd-sourced intensity data filtered adequately and iv) new types of GIS map representations. In Romania, the seismic network operated by NIEP consists of more than 140 real-time seismic stations; in the new ShakeMap development, we include new rules for discriminating between reliable and erroneously reported values of macroseismic intensities.

Keywords: earthquake; ShakeMap; intensity; Vrancea; ground motion prediction equation.