As part of our routine review of data we collect, we select data windows around a limited number of larger earthquakes. In the past, we have simply verified that the observed arrivals were acceptably close to predicted arrival times. However, beginning in January 1998, we started to record our arrival time picks as part of our routine data review. Although we typically pick just larger teleseisms, we will usually pick whatever arrivals are visible in the time windows we review. Thus, on occasion, we are picking arrivals from local or regional events.

Our arrival time picks are made from either broadband (at 20 samples per second) or short period (at 40 samples per second) unfiltered data. We make "classic" picks in the sense that we visually review the time series and mark, to the best of our ability, the onsets of phase arrivals.

Our basic arrival time dataset provides only the station name, the channel from which we made the pick, and the arrival time. We then use a catalog of large teleseisms (the Harvard CMT catalog) to perform a rudimentary event and phase association of our arrival time picks. Both the raw and "associated" arrival time datasets are available below.

Please note that the data presented here are experimental. We can make no guarantees about the quality or completeness of these data. Our first and foremost concern is proper quality control of the waveforms--the arrival time datasets presented here are just a byproduct of that work. If you have questions or comments on this project please send mail to: bolton@usgs.gov.

"associated" arrival time picks in tabular form. The columns of this file are, respectively, station name, channel name, year, day, hour, minute, and second of the phase pick, year, day, hour, minute, second, latitude, longitude, depth (km) of the associated earthquake, and the phase association.