A novel quality criterion for self-noise measurement of velocity-type seismometers

Abstract Accurate measurement of seismometer self-noise is crucial for instrument performance evaluation and the development of deep exploration equipment. However, the measurement accuracy is severely compromised by external background noise, time synchronization errors, and other factors. The long-standing absence of quantitative data quality assessment methods has undermined the credibility of measurement results. This study proposes the two-path coefficient of variation (TPCV) criterion, which quantifies data quality by utilizing the practical discrepancy between two theoretically equivalent computational paths in the three-channel correlation method. Simulations using synthetic data validate the statistical relationship between the TPCV criterion and self-noise measurement errors under varying conditions of synchronization error and background noise intensity. Additionally, a unified unit conversion method for calculating acceleration noise in velocity-type seismometers and a synchronization error compensation algorithm are introduced. Field experiments with a moving-coil seismometer in an urban lawn environment, affected by strong background noise, yielded TPCV values of 29.0%–65.1% (median 37.4%), exceeding the 0.30 reference threshold. The corresponding self-noise was 205–285 n g / H z @10 Hz, indicating poor data quality. In a cave environment experiment using nodal seismometers, the pre-compensation TPCV values were 26.4%–61.0% (median 29.6%), falling within the suspicious to unreliable range, with self-noise of 5–7 n g / H z @10 Hz. After synchronization compensation, TPCV values decreased to 1.2%–11.9% (median 1.7%), and self-noise recovered to 2.6–4.7 n g / H z @10 Hz, representing a 94% improvement in TPCV. This verifies the effectiveness of the synchronization compensation algorithm and the sensitivity of the TPCV criterion to time synchronization errors. This study provides a quantitative and automated quality control tool for the three-channel correlation method, recommends reference thresholds for data quality assessment, and holds promise for promoting the establishment of a more scientific and unified standard system for seismometer self-noise measurement.