In recent years, distributed acoustic sensing （DAS） has experienced rapid development and extensive applications in seismology and various other fields, owing to its advantages of low cost, high spatial resolution, and real-time monitoring capabilities.As a critical sensing component in DAS observational systems, optical fibers can directly affect the quality of acquired data and, consequently, the results. Meanwhile, the physical properties of optical fibers are also closely related to the cost and difficulty of field deployment. However, a quantitative guideline for selecting suitable optical fibers for geophysical surveys is still lacking. In this study, to fill this gap, we first propose a comprehensive set of criteria based on the key attributes of seismic data to evaluate the quality of DAS records quantitatively, including relative sensitivity, signal-to-noise ratio, fidelity compared to data recorded by nodal stations, and semblance across multiple channels. Then, we conduct rigorous evaluations of DAS data recorded by two diverse seismic observational systems: A DAS cable deployed in a surface trench and a DAS cable deployed in a vertical borehole.

In the former experiment, a trench approximately 20−30 cm deep and 90 m long was excavated on a lawn adjacent to two main urban roads. Five regular optical communication cables were deployed in this trench to ensure identical coupling conditions. The five optical cables were then fused end-to-end into a single continuous cable, which was connected to a DAS interrogator for data collection. Along the trench, 21 short-period nodal seismometers were deployed at 4-m intervals to assess the fidelity of the DAS records. Active signals from hammering were generated, and spatiotemporal 2D cross-correlation template matching was used to accurately determine the location of the DAS channels. Then, the signals recorded by different optical cables were compared trace-by-trace. The analyzed data include active signals from hammering, surface waves generated by traffic, and cross-correlation functions of ambient noise. For the active-source signals, we calculated the relative sensitivity, signal-to-noise ratio in both time and frequency domains, and data fidelity. Since DAS records strains or strain rates rather than particle velocities （as seismometers do）, unit conversion is required before calculating fidelity. For the traffic-generated surface waves, fidelity and semblance were computed. The fidelity of the surface waves was notably higher than that of the signals generated by hammering. For the cross-correlation functions from seismic ambient noise, only the signal-to-noise ratio was compared. Systematic evaluations with various sources revealed distinct differences between the evaluated cables, with the overall performance of the optical cables in vibration reception ranked as follows: YZ, GYFTY, GYTA, ADSS, and GJFJV.

The best-performing YZ cable retained for the borehole experiment, in which three additional armored optical cables （YPTPU, SCJKBH, and SCTX3Y） were also selected. These four cables were installed in a 200-meter-deep borehole. Below 80 m depth, expansive clay pellets were used as infill, while silica sand was used above this depth. Given the limited space in the borehole, which made fusing challenging, each cable was connected to an equivalent interrogator for data acquisition. Several active sources were used near the well to generate signals. Due to significant differences in coupling between the shallow and deep sections of the borehole, only data recorded by channels shallower than 80 m were considered for evaluation. The relative sensitivity, signal-to-noise ratio, and semblance of data from the four cables were evaluatedand considerable differences are observed. Overall, the YPTPU cable is found to have the best performance for DAS data collection in the borehole.

The results of this study indicate that distinct variations in data quality can be found among different optical cables, particularly within the borehole applications. Therefore, proper selection of cables is essential for collecting high-quality DAS data, which is crucial for monitoring weak signals and achieving reliable subsurface imaging.