Site Effects in the City of Port au Prince (Haiti) Inferred From 2010 Earthquake Aftershocks Recordings.

Monday, 15 December 2014
Sadrac ST Fleur1,2, Françoise Courboulex1, Etienne Bertrand3, Anne Deschamps1, B F Mercier De Lepinay1, Dominique Boisson2, Claude Prepetit2 and Susan Elizabeth Hough4, (1)GeoAzur, University of Nice-Sophia Antipolis, Valbonne, France, (2)URGéo, Faculté des Sciences, Université d'Etat d'Haiti, Port-au-Prince, Haiti, (3)CEREMA, Laboratoire de Nice, Nice, France, (4)USGS, Pasadena, CA, United States
The Haitian earthquake of 12 January 2010 (Mw=7) caused an unprecedented disaster in Port-au-Prince as well as in smaller cities close to the epicenter. The extent of damage appears to be initially attributed to the proximity of the earthquake in Port-au-Prince, the extreme vulnerability of many structures, and a high population density. However, the damage distribution for this earthquake suggests a general correlation of damage with small-scale topographical features and local geological structure.

The main objective of this work is to investigate site effects in the city of Port-au-Prince. It is also to better define the response of different sites to earthquakes and establish transfer functions between each site and a particular site defined as a reference site. Specific soil columns is determined in the vicinity of each station in order to carry out 1D simulations of soil response at these sites.

About 90 earthquakes (2<Mw<5) were recorded between March 2010 and February 2013 on a local network composed of nine accelerometers installed by USGS, three broad-band velocimeters installed by NRCAN and two velocimeters of the French “Sismo at school” network. We located 39 of these events using the permanent network and 43 were located by Douilly et al. (2013) using a temporary network.

The ground motion recordings at these stations were then analyzed in order to study the topographic and lithologic amplification effects observed on these sites. To quantify site effects under each station, we have used classical spectral ratio methods. In a first step, the HVSR earthquake method (Horizontal over Vertical ratio) was used to choose a reference station in Port au Prince that should be ideally a station without any site effects. We selected HCEA station as reference station. In a second step, we estimated the transfer function at each station by the SSR (Standard Spectral Ratio). Finally, these transfer functions estimated by the spectral ratios technique were compared to 1D simulations. We evidenced the existence of a possible surface wave effects in the plain of Port-au-Prince and topographical effects in the hills of Canape-vert.