No abstract

Summary The accurate timing of seismological data is crucial for most quantitative examinations in seismology. We present evidence that traveltime data from many stations contain systematic variations in timing which can be identified by checking the median of station delay times as a function of time. This function is expected to be constant but many deviations are found. Several hundred stations that report arrival times to the ISC have been examined. The median station delay times of almost 8 per cent of these stations show changes of more than 1 s and thus exceed the structural signal in the data. Temporal variations of 0.5–1 s are common. Changes in the distribution of observed earthquakes and other possible explanations of such variations have been tested and fail to explain most of the observations. Therefore, the bulk of the observed changes must be caused by flaws in the timing of the data or by biased picking of arrival times. For instance, at one station with a strong annual variation of noise level, the arrival times are on average picked several tenths of a second later during months with a high noise level. Because of their systematic nature, these errors will not necessarily cancel out by using the large number of traveltimes in the ISC Bulletin and may therefore introduce a bias in many investigations. If the observed timing variations are due to the recording equipment at the stations, the errors will be present in the digital waveform data as well. Tomographic studies could potentially be affected, but in particular studies of temporal variations of Earth structure based on traveltime data, e.g. inner core rotation, need to be looked at with caution as results might be influenced by station effects. The exact nature of the bias is study‐dependent and needs thorough investigation in each individual case.

S U M M A RYThe ISC Bulletin is the most widely used data set for traveltime tomography since it comprises the largest collection of arrival times. The large number of stations and events gives the best ray path coverage of the Earth's mantle currently available, but the data also contain a considerable amount of noise. The e¡ects of noise are supposedly reduced by averaging delay times for similar paths, choosing a suitable parametrization and damping the tomographic inversions. In this article we focus on two types of errors in body wave arrival times and estimate their e¡ects on global tomographic models using synthetic tests. The ¢rst type of error stems from the ¢nite reading precision of arrival times and is equivalent to a round-o¡ error. This yields a random contribution to delay times. We show that the reading precision inferred here often does not coincide with that reported in the Bulletin. The in£uence on tomography of a reading precision of 1 s or better is almost insigni¢cant since (1) its variance is very small compared to the total variance of ISC delay times and (2) less than 5 per cent of that variance maps into the tomographic model. A few stations report some arrival times with an indicated reading precision of 0.1 s that are in reality only picked to the closest 10 s or 1 min. This results in a S/N ratio much lower than 1 and these data should be removed. The second type of error causes systematic variations of delay times as a function of time. A large amount of this error maps into the tomographic model, but, luckily, the size of the error in the input data is one order of magnitude smaller than the standard deviation of the ISC delay times. A test reveals that the rms amplitude due to these systematic errors is between 4.2 per cent (0^35 km depth) and 14.4 per cent (1800^2000 km depth) of the model rms amplitude. This blurs the tomographic model to some degree but does not change the overall amplitude or shape of seismic anomalies.