Seismoacoustics is the combined study of vibrations in the Earth and sound waves in the atmosphere, used to characterize non-earthquake geohazards such as avalanches, landslides, and volcanic eruptions.

pyTREMOR fetches live seismic waveform data from global FDSN/EarthScope servers and sonifies it — converting ground motion into spectrogram video + audio (MP4). Configured for near-real-time monitoring of active volcanic regions across Oceania, Asia, and Australia, referenced from Hobart, Tasmania.

1. FETCH    →  Downloads raw seismic waveform data from EarthScope/FDSN (last N hours, UTC)
2. FILTER   →  Applies bandpass filter (1–18 Hz for BHZ/40 sps, 1–9 Hz for GSI/20 sps), trims and processes the signal
3. SONIFY   →  Renders frequency-coloured waveform + spectrogram + audio into an MP4 saved to dataset/

• Python 3.x
• ObsPy — seismic data access and processing
• pygame — audio playback
• tqdm — progress display
• ffmpeg — video encoding (must be in system PATH)
• matplotlib, numpy (installed with ObsPy)

# Recommended (pip)
pip install obspy pygame tqdm

# Or using the included setup file
python3 setup.py install

macOS only: After installing Python from python.org, run the certificate installer once to allow urllib (used by ObsPy) to make HTTPS connections:

bash "/Applications/Python 3.12/Install Certificates.command"

Replace 3.12 with your installed Python version. Without this step, ObsPy's service discovery will silently fail and all EarthScope-routed stations will return "This station is not replying!" even though the network is reachable.

Each station produces a spectrogram video (MP4) saved to dataset/ with the filename format:

dataset/YYYY-MM-DD-HH-MM-STATION.mp4

The video renders as shown:

To clear old output files:

# Linux / macOS
rm dataset/*.mp4

# Windows (PowerShell)
Remove-Item dataset\*.mp4

All stations use IU (GSN), AU, or GE networks via EarthScope. Time windows are calculated in UTC for near-real-time accuracy. Default window: last 5 hours.

The map below shows all configured stations sorted by distance from Hobart, Tasmania, with coordinates and volcanic context. Generated by docs/generate_station_map.py.

Stations are defined in the autoconfig file. Each active station uses the format:

#LABEL{network=IU,station=SNZO,channel=BHZ,freqmin=1,freqmax=23,speed_up_factor=200,fps=1,spec_win_dur=8,db_lim=-180|-130}

To add a station: add a new #LABEL{...} line.
To disable a station: remove the # prefix.
To change the time window: edit LASTHOURS=5 at the top.

Station parameters:

# Help
python3 pyTREMOR.py --help

# Interactive mode
python3 pyTREMOR.py --cmd

# Autorun — all configured stations, last N hours
python3 pyTREMOR.py --autorun

Runs on Windows, Linux, and macOS with Python 3.x. All time handling uses UTC internally — local system timezone does not affect data retrieval.

macOS only. Python installed from python.org ships without SSL certificates for the built-in urllib library. ObsPy uses urllib for FDSN service discovery (fetching .wadl descriptors), so all HTTPS endpoints fail silently with SSL: CERTIFICATE_VERIFY_FAILED — even though curl, requests, and the browser can reach the same URLs without issue.

Fix (one-time):

bash "/Applications/Python 3.12/Install Certificates.command"

Replace 3.12 with your Python version. This symlinks certifi's CA bundle as Python's default SSL trust store.

How to verify the fix worked:

from obspy.clients.fdsn import Client
c = Client('https://service.earthscope.org', timeout=30)
print(c.services.keys())  # should print: dict_keys(['dataselect', 'event', 'station', ...])

The cron environment has a restricted PATH. Ensure your cron entry sources the full Python path and logs output:

0 7-22 * * * /bin/bash /path/to/pytremor/run_pytremor.sh >> /path/to/pytremor/logs/cron.log 2>&1

Check logs/cron.log for errors. Common causes: SSL failure (see above), FDSN service outage, or stale lockfile at /tmp/pytremor.lock.

macOS cron has several gotchas that make it harder to set up than on Linux. Follow these steps carefully.

macOS blocks cron from accessing user files by default. Without this, the script will silently do nothing.

1. Open System Settings → Privacy & Security → Full Disk Access
2. Click + and add /usr/sbin/cron
3. If /usr/sbin/cron is not visible in Finder: press ⌘ Shift G in the file picker and type /usr/sbin/

Same path: Full Disk Access → + → add your terminal app (Terminal.app or iTerm2).

crontab -e

Add this line to run pyTREMOR every hour from 7am–10pm:

0 7-22 * * * /bin/bash /Users/yourname/pytremor/run_pytremor.sh

Replace /Users/yourname/pytremor with the actual path. The run_pytremor.sh script handles logging and lockfile management — do not redirect output again in the crontab itself.

# Check cron daemon is active
sudo launchctl list | grep cron

# Watch the log in real time
tail -f /Users/yourname/pytremor/logs/cron.log

A healthy log entry looks like:

Wed May 21 08:00:01 AEST 2026: [OK] SNZO, [OK] RABL, ...

If pyTREMOR was killed mid-run, the lockfile may remain and block all future runs:

rm /tmp/pytremor.lock

Check if this is the cause: cat /tmp/pytremor.lock — if the PID inside is no longer running (ps aux | grep <PID>), the lock is stale and safe to delete.

Alternative: macOS launchd (via a .plist file in ~/Library/LaunchAgents/) is more reliable than cron on macOS — it persists across reboots and respects system sleep/wake cycles. Cron works but requires the above steps every time on a new machine.

• GeoNet (NZ) stations (e.g. WHVZ, KRVZ) are not supported — they require a separate endpoint (service.geonet.org.nz) not routed by EarthScope
• Requires a stable internet connection for FDSN data retrieval
• Very recent data (< 5 min) may not yet be available on EarthScope servers
• dataset/ output accumulates — clear regularly with rm dataset/*.mp4

pyTREMOR is released under the GPLv3.

✉️ root /at/ victormazon.com