Identification of noise sources associated with human activities
Predominant anthropogenic noise sources vary in different city sectors, among workplaces (Ch 1240–1440), main roads (Ch 850–1110), a residential area (Ch 690–830) and a less populated area (Ch 110–300) (Figure 1b and 2). Characterizing seismic noise from particular sources can help us understand local social interactions with city lockdown measures. Our 5-km-long dense DAS array at 2 m spacing covers plenty of public infrastructures. Hence, we chose specific subarrays and identified noise sources – footstep signals, passing vehicles and industrial noise, by comparing seismic noise variations before and after the COVID-19 restrictions.
To analyze the impact of lockdown measures on footsteps, we selected 1-hour data (local time 10 am–11 am) from a subarray beneath a pedestrian-only path on the main campus for similar days (March 5, April 16, and June 4) (Figure 3). Intuitively, walking footstep signals showed up in the data as linear streaks with a slow moveout (1–2 m/s). On March 5, during the regular semester, the DAS array picked up many walking signals (plenty of data streaks in Figure 3a). Contrarily on April 16, after the stay-at-home order was issued, only a few signals are detected on this path (Figure 3b). In Phase Green (June 4), the footstep signals are almost not recovered despite the easing of some restriction measures. This invariability is confirmed by the average spectrum plot in Figure 3d, showing the absence of peaks at 2 Hz and 4 Hz in both April 16 and June 4 curves, which are considered as the footstep signals (Zhu et al., 2021).
We next analyzed traffic noise recordings (Figures 3e-h) from a subarray beneath Curtin Road, the main road on campus. There is a significant decrease in passing vehicles when comparing data between March 5 and April 16. This is due to the shutdown of the university preventing people from traveling to campus. The bus service was also reduced. On June 4, more linear signals indicate more passing vehicles. The decrease-increase traffic noise pattern is obviously different from the loss-to-flat pattern of pedestrian movement. This trend is also confirmed in the frequency spectrum (Figure 3h): a significant drop of the power spectra between 10 and 50 Hz about 20 dB, then an increase by 5 dB. We interpreted 10–50 Hz as the frequency band of passing vehicles.
In addition, we identified higher frequency noise associated with construction activities. On the western campus, a new parking garage and utility upgrades near the fibers were planned to be constructed from December 17, 2019, to April 20, 2021. Due to the suspension of the industrial activity during the stay-at-home measures, the data on April 15 show no detected events (Figure 3i). After restarting industrial activities since May 7 (Phase Yellow), we observe strong industrial noise on June 1 in Figure 3j. The noise in the frequency band of 10–30 Hz could be identified as noise from construction vehicles and the broadband impulses (10–100 Hz) between 11:09 – 11:10 am were from machinery, which produced short bursts of vibrations.