Testing strong motion stations continuity of operation using random fields and intensity data

In the present work we perform a test of the continuity of operation of 76 strong motionstations selected for the validation and ranking of different Probabilistic Seismic HazardAssessment models within the framework of the project "Constraining observations intoSeismic Hazard". We exploit a web tool developed within the project to select, for eachone of the 76 stations, the earthquakes that might have affected the station but werenot recorded by it. To perform the check we produce potential "observations" at the testsites (i.e. the location of the stations that did not record the selected earthquakes),based on the modeling of spatially correlated random fields of a ground motionIntensity Measure (IM), constrained by the available instrumental and macroseisimicintensity observations. The proposed methodology allows one to produce, for eachearthquake considered, a ground motion distribution at the sites of interest that can beused to check whether the given earthquake should have triggered or not the stationsselected. Here the IM of interest is PGA, since this is the IM used to set the triggeringthreshold of the older analogue instruments. The actual threshold value used in thetest is PGA = 0.01g. More than the specific On-Off status of a station however, what isof practical relevance here is whether a potential observation is larger than themaximum acceleration available in the seismic history of the given station, since this isthe actual observation that is used in the PSHA testing procedure.