Scalable Seismological Pipeline for Assessment, Review, and Quality (SSPARQ) of Seismic Data

Abstract Ensuring the long-term quality of seismic data from large networks remains a critical challenge in seismology, especially as the volume of available waveforms continues to grow. Sensor misorientation, timing instability, and relative sensitivity anomalies can introduce significant biases in quantitative analyses. To address this need, we present Scalable Seismological Pipeline for Assessment, Review, and Quality (SSPARQ), an open-source Python-based tool designed to automate the assessment of station performance based on P-wave characteristics. SSPARQ minimizes P-wave energy on the transverse component to estimate azimuthal misorientation and incorporates additional diagnostics to evaluate clock stability and intercomponent relative sensitivity consistency. The methodology requires only short waveform windows around the P-wave arrival and relies on basic criteria including signal-to-noise ratio, intercomponent energy ratios, and cross-correlation between radial and vertical components. A density-based clustering algorithm, DBSCAN, is employed to resolve time-dependent orientation trends. SSPARQ is scalable, supports parallel processing, and is well suited for retrospective evaluation of stations over multiyear deployments. We demonstrate the effectiveness of SSPARQ using three representative seismic stations: MVCO (United States National Seismic Network–USNSN), WCI (Global Seismograph Network–GSN), and KOWA (GSN). At MVCO, we detect a +20° sensor misorientation consistent with prior studies and metadata, which was corrected during maintenance in 2019. At WCI, we identify a ∼9-month period of instrumental anomalies in 2014–2015. At KOWA, although orientation remains stable, we observe progressive clock degradation beginning in late 2012, followed by a separate interval of anomalous sensitivity behavior. These case studies confirm SSPARQ’s capability to recover subtle instrumental issues using waveform-only data, without requiring metadata or synthetic modeling. SSPARQ is intended for seismologists seeking a lightweight yet robust solution for seismic data quality assessment, especially in contexts involving temporary networks, heterogeneous instrumentation, or archived data from multiple data centers.