Displacement Seismic Amplitude Ratio (DSAR)

Python package for computing Displacement Seismic Amplitude Ratio (DSAR) from miniSEED seismic data stored in SeisComP Data Structure (SDS) archives. DSAR is used in volcanology as a long-term precursor for detecting changes in seismic attenuation related to volcanic unrest.

Requirements

• Python ≥ 3.10

Installation

Using uv (recommended)

uv is the recommended environment and package manager for this project.

# Install uv (if not already installed)
pip install uv

# Create a virtual environment and install dsar
uv venv
uv pip install dsar

To run scripts inside the uv-managed environment:

uv run python main.py

Using pip

pip install dsar

---

How to Use

1. Prepare your SDS data directory

input_dir must follow the SeisComP Data Structure (SDS) layout:

{input_dir}/
└── {year}/
    └── {network}/
        └── {station}/
            └── {channel}.D/
                └── {network}.{station}.{location}.{channel}.D.{year}.{julian_day}

Example file path:

D:\Data\OJN\2025\VG\OJN\EHZ.D\VG.OJN.00.EHZ.D.2025.001

The NSLC identifier ({network}.{station}.{location}.{channel}) is used throughout for naming output files and directories.

---

2. Calculate DSAR

Import

from dsar import DSAR, PlotDsar

Initialize

dsar = DSAR(
    input_dir="D:\\Data\\OJN",
    start_date="2025-01-01",
    end_date="2025-01-08",
    station="OJN",
    channel="EHZ",
    network="VG",
    location="00",
    resample="10min",   # optional, default "10min"
    verbose=True,       # optional, default False
    debug=False,        # optional, default False
)

Parameter	Type	Default	Description
input_dir	str	required	Path to the SDS root directory
start_date	str	required	Start date in YYYY-MM-DD format
end_date	str	required	End date in YYYY-MM-DD format
station	str	required	Station code (e.g. "OJN")
channel	str	required	Channel code (e.g. "EHZ")
network	str	required	Network code (e.g. "VG")
location	str	required	Location code (e.g. "00")
resample	str	"10min"	Pandas offset alias for the resampling interval
output_dir	str	None	Custom output directory; defaults to <cwd>/output/dsar
verbose	bool	False	Print detailed stream information
debug	bool	False	Print debug-level path and trace information

Set custom frequency bands (optional)

The default bands are LF [0.1, 4.5, 8.0] Hz and HF [0.1, 8.0, 16.0] Hz. Each band is defined by three frequencies [high_pass, bandpass_low, bandpass_high].

dsar.first_bands(name="LF", first_freq=0.1, second_freq=4.5, third_freq=8.0)
dsar.second_bands(name="HF", first_freq=0.1, second_freq=8.0, third_freq=16.0)

first_bands is the numerator and second_bands is the denominator of the DSAR ratio. Both methods return self, so they can be chained:

dsar.first_bands("LF", 0.1, 4.5, 8.0).second_bands("HF", 0.1, 8.0, 16.0)

Run

dsar.run()

run() iterates day by day across the date range, loads each day's miniSEED data from the SDS archive, processes both frequency bands, computes the DSAR ratio, and saves the result as a daily CSV.

Output files:

output/dsar/{NSLC}/{resample}/{NSLC}_{YYYY-MM-DD}.csv

Each CSV contains:

Column	Description
datetime	Timestamp (index)
LF	Low-frequency band amplitude
HF	High-frequency band amplitude
DSAR_{resample}	Raw DSAR ratio (LF / HF)
DSAR_6h_median	6-hour centered rolling median
DSAR_24h_median	24-hour centered rolling median

---

3. Plot DSAR

Initialize

plot = PlotDsar(
    start_date="2025-01-01",
    end_date="2025-01-08",
    station="OJN",
    channel="EHZ",
    network="VG",
    location="00",
    resample="10min",   # must match the resample used in DSAR
)

Parameter	Type	Default	Description
start_date	str	required	Start date in YYYY-MM-DD format
end_date	str	required	End date in YYYY-MM-DD format
station	str	required	Station code
channel	str	required	Channel code
network	str	"VG"	Network code
location	str	"00"	Location code
resample	str	"10min"	Must match the interval used when running DSAR
dsar_dir	str	None	Custom DSAR CSV directory; defaults to <cwd>/output/dsar
figures_dir	str	None	Custom figures directory; defaults to <cwd>/output/figures/dsar

Get the combined DataFrame

df = plot.df

plot.df reads all daily CSVs for the NSLC, concatenates them, removes duplicates, and saves a combined CSV:

output/dsar/{NSLC}/combined_{resample}_{NSLC}.csv

You can use df directly for further analysis:

print(df.head())
print(df.columns.tolist())

Plot

fig = plot.plot(
    interval_day=7,     # x-axis tick interval in days
    y_min=85,           # optional y-axis minimum
    y_max=225,          # optional y-axis maximum
    save=True,          # save figure to disk
    file_type="jpg",    # output format: "png", "jpg", etc.
)

Parameter	Type	Default	Description
interval_day	int	3	X-axis major tick interval in days
title	str	None	Custom plot title; defaults to "DSAR - {NSLC}"
y_min	float	None	Y-axis minimum (auto-scaled if not set)
y_max	float	None	Y-axis maximum (auto-scaled if not set)
save	bool	True	Save the figure to disk
file_type	str	"png"	Output file format

Output figure:

output/figures/dsar/{NSLC}/{NSLC}_{resample}_{start_date}-{end_date}.{file_type}

---

4. Add eruption markers (optional)

Use the utility functions to annotate eruption events on an existing axes:

from dsar.utilities import plot_eruptions

fig = plot.plot(save=False)
ax = fig.axes[0]

plot_eruptions(
    ax,
    axvspans=[["2025-01-03", "2025-01-05"]],   # continuous eruption intervals
    axvlines=["2025-01-07"],                     # discrete eruption events
)
fig.savefig("output.png", dpi=300)

---

5. Load a combined CSV directly

If you already have a combined CSV file and just want to load it:

from dsar.utilities import get_combined_csv

df = get_combined_csv(
    directory="output/dsar",
    station="VG.OJN.00.EHZ",
    resample="10min",
)

---

Complete example

from dsar import DSAR, PlotDsar

# --- Calculate ---
dsar = DSAR(
    input_dir="D:\\Data\\OJN",
    start_date="2025-01-01",
    end_date="2025-01-08",
    station="OJN",
    channel="EHZ",
    network="VG",
    location="00",
    verbose=True,
)
dsar.first_bands(name="LF", first_freq=0.1, second_freq=4.5, third_freq=8.0)
dsar.second_bands(name="HF", first_freq=0.1, second_freq=8.0, third_freq=16.0)
dsar.run()

# --- Plot ---
plot = PlotDsar(
    start_date="2025-01-01",
    end_date="2025-01-08",
    station="OJN",
    channel="EHZ",
    network="VG",
    location="00",
)
plot.plot(interval_day=2, y_min=50, y_max=300, save=True, file_type="png")

---

References

Caudron, C., et al., 2019, Change in seismic attenuation as a long-term precursor of gas-driven eruptions: Geology, https://doi.org/10.1130/G46107.1

Chardot, L., Jolly, A. D., Kennedy, B. M., Fournier, N., & Sherburn, S. (2015). Using volcanic tremor for eruption forecasting at White Island volcano (Whakaari), New Zealand. Journal of Volcanology and Geothermal Research, 302, 11–23. https://doi.org/10.1016/j.jvolgeores.2015.06.001