Temporary broadband sensor deployments have traditionally been predominantly emplaced using shallow vaults that require more materials, personnel, and time than direct burial. However, new developments in seismometer and seismograph technology are increasingly facilitating systems that can be directly buried in earth or snow without vault enclosures. We analyze data from two identical shallow vaults installed adjacent to two identical direct burial sites in soft floodplain alluvial and shallow water table conditions near Socorro, New Mexico. Data recorded from these four sensors over eight months in 2012 were assessed to determine if the emplacement type had a significant and systematic effect on data quality. We used metrics derived from power spectral density analysis to examine temporal trends in noise (instrument, installation, and site noise), signal‐to‐noise ratio for teleseismic and local earthquakes, and magnitude‐squared coherence of both noise and earthquake signal recordings. We found that noise on the vault sensors is higher during the transition from spring to summer than for the direct burial sensors. This difference is especially evident on the horizontal components at long periods between 20 and 170 s, with an average of 5.3 dB more noise on the vaults than the direct burials from April to June, indicating enhanced tilt susceptibility for vault emplaced sensors. However, most variability in data quality is comparable between sensors with differing emplacement methods and between sensors with the same emplacement method in this four‐station experiment. We conclude that the direct burial emplaced sensors at this test site performed as well as the vault emplaced sensors and that direct burial is preferable when considering data quality and ease of installation.
Online Material: Figures of daily median noise levels within the microseism band, monthly probability density functions for the magnitude‐squared coherence, and coherence self‐noise analysis results.