Seismic waveform data are generally contaminated by noise from various sources. Suppressing this noise effectively so that the remaining signal of interest can be successfully exploited remains a fundamental problem for the seismological community. To date, the most common noise suppression methods have been based on frequency filtering. These methods, however, are less effective when the signal of interest and noise share similar frequency bands. Inspired by source separation studies in the field of music information retrieval (Jansson et al., 2017) and a recent study in seismology (Zhu et al., 2019), we implemented a seismic denoising method that uses a trained deep convolutional neural network (CNN) model to decompose an input waveform into a signal of interest and noise. In our approach, the CNN provides a signal mask and a noise mask for an input signal. The short-time Fourier transform (STFT) of the estimated signal is obtained by multiplying the signal mask with the STFT of the input signal. To build and test the denoiser, we used carefully compiled signal and noise datasets of seismograms recorded by the University of Utah Seismograph Stations network. Results of test runs involving more than 9000 constructed waveforms suggest that on average the denoiser improves the signal-to-noise ratios (SNRs) by ∼5  dB, and that most of the recovered signal waveforms have high similarity with respect to the target waveforms (average correlation coefficient of ∼0.80) and suffer little distortion. Application to real data suggests that our denoiser achieves on average a factor of up to ∼2–5 improvement in SNR over band-pass filtering and can suppress many types of noise that band-pass filtering cannot. For individual waveforms, the improvement can be as high as ∼15  dB.

No abstract

<p>Historically, location algorithms have relied on simple, one-dimensional (1D, with depth) velocity models for fast, seismic event locations. The speed of these 1D models made them the preferred type of velocity model for operational needs, mainly due to computational requirements. Higher-dimensional (2D-3D) seismic velocity models are becoming more readily available from the scientific community and can provide significantly more accurate event locations over 1D models. The computational requirements of these higher-dimensional models tend to make their operational use prohibitive. The benefit of a 1D model is that it is generally used as travel-time lookup tables, one for each seismic phase, with travel-time predictions pre-calculated for event distance and depth. This simple, lookup structure makes the travel-time computation extremely fast.</p><p>Comparing location accuracy for 2D and 3D seismic velocity models tends to be problematic because each model is usually determined using different inversion parameters and ray-tracing algorithms. Attempting to use a different ray-tracing algorithm than used to develop a model almost always results in poor travel-time prediction compared to the algorithm used when developing the model.</p><p>We will demonstrate that using an open-source framework (GeoTess, www.sandia.gov/geotess) that can easily store 3D travel-time data can overcome the ray-tracing algorithm hurdle. Travel-time lookup tables (one for each station and phase) can be generated using the exact ray-tracing algorithm that is preferred for a specified model. The lookup surfaces are generally applied as corrections to a simple 1D model and also include variations in event depth, as opposed to legacy source-specific station corrections (SSSCs), as well as estimates of path-specific travel-time uncertainty. Having a common travel-time framework used for a location algorithm allows individual 2D and 3D velocity models to be compared in a fair, consistent manner.</p>

No abstract

No abstract