thistle 0.3.0rc1

Satellite orbit propagation and data generation with automatic TLE switching

Thistle

Satellite orbit propagation and data generation built on SGP4 and Skyfield. Handles automatic TLE switching for long-duration propagation, and generates orbit data, ground site ranges, and satellite events from a single propagation pass.

Installation

pip install thistle

Requires Python >= 3.9.

Quick start

import numpy as np
from thistle import Propagator, read_tle, generate

tles = read_tle("satellite_data.tle")
prop = Propagator(tles, method="midpoint")

times = np.arange(
    np.datetime64("2024-01-15T00:00"),
    np.datetime64("2024-01-16T00:00"),
    np.timedelta64(60, "s"),
)

data = generate(times, prop, ["eci", "lla"])
# data["eci_x"], data["lat"], data["lon"], ...

Propagator

A single TLE degrades in accuracy as you propagate further from its epoch. For tracking a satellite over days or weeks you need multiple TLEs and a strategy for switching between them.

from thistle import Propagator, read_tle

tles = read_tle("satellite_data.tle")
prop = Propagator(tles, method="midpoint")

Switching strategies

Strategy	Description
"epoch"	Uses the most recent TLE at or before the target time
"midpoint"	Transitions at the midpoint between consecutive TLE epochs
"tca"	Transitions at the time of closest approach between neighboring TLEs

You can also pass a strategy instance directly:

from thistle import Propagator, EpochSwitchStrategy, read_tle

tles = read_tle("satellite_data.tle")
prop = Propagator(tles, method=EpochSwitchStrategy([]))

This is useful for subclassing SwitchingStrategy to implement custom switching logic.

Lookup methods

time = np.datetime64("2024-01-15T12:00:00")

satellite = prop.find_satellite(time)  # Skyfield EarthSatellite
satrec = prop.find_satrec(time)        # sgp4 Satrec
tle = prop.find_tle(time)              # (line1, line2) strings

Direct propagation

The Propagator can be used as a drop-in replacement for a Skyfield EarthSatellite:

from skyfield.api import load
ts = load.timescale()

geo = prop.at(ts.utc(2024, 1, 15, 12, 0, 0))

Data generation

generate() propagates once and extracts multiple data groups in a single pass.

from thistle import generate

data = generate(times, prop, ["eci", "lla", "keplerian", "sunlight"])

Groups

Group	Keys	Units
eci	eci_x, eci_y, eci_z, eci_vx, eci_vy, eci_vz	m, m/s
ecef	ecef_x, ecef_y, ecef_z, ecef_vx, ecef_vy, ecef_vz	m, m/s
lla	lat, lon, alt	deg, deg, m
keplerian	sma, ecc, inc, raan, aop, ta, ma, ea, arglat, tlon, mlon, lonper, mm	m, -, deg (mm: deg/day)
equinoctial	p, f, g, h, k, L	m, -, -, -, -, deg
sunlight	sun	int8 (0=umbra, 1=penumbra, 2=sunlit)
beta	beta	deg
lst	lst	hours [0, 24)
mag_enu	Be, Bn, Bu	nT
mag_total	Bt	nT
mag_ecef	Bx, By, Bz	nT

All values are returned as NumPy arrays. Angles and dimensionless quantities are float32; positions, velocities, and range data are float64.

Ground site range

Pass sites to generate() to compute slant range and range rate without a second propagation:

data = generate(times, prop, ["lla"], sites={"ksc": (28.57, -80.65)})
# data["range_ksc"]       -> slant range in meters
# data["range_rate_ksc"]  -> range rate in m/s

Sites can also be a list (keys become range_0, range_1, ...):

data = generate(times, prop, ["lla"], sites=[(28.57, -80.65), (34.05, -118.24)])

Each site tuple is (lat, lon) or (lat, lon, alt_m).

The standalone generate_range() function is also available:

from thistle import generate_range

rng = generate_range(times, prop, sites={"ksc": (28.57, -80.65)})

Doppler shift

from thistle import doppler_shift

doppler_hz = doppler_shift(data["range_rate_ksc"], freq=437e6)

Events

Find satellite events within a time window. All event functions accept either an EarthSatellite or a Propagator and return lists of dicts.

Passes

from thistle import find_passes

passes = find_passes(start, stop, prop, lat=28.57, lon=-80.65, min_elevation=10.0)
for p in passes:
    print(p["start"], p["stop"], p["peak_elevation"])

Returns dicts with keys: start, stop, peak_time, peak_elevation.

Node crossings

from thistle import find_node_crossings

crossings = find_node_crossings(start, stop, prop)
for c in crossings:
    print(c["start"], c["longitude"], c["ascending"])

Returns dicts with keys: start, stop, longitude, ascending.

Sunlit and eclipse periods

from thistle import find_sunlit_periods, find_eclipse_periods

sunlit = find_sunlit_periods(start, stop, prop)
eclipse = find_eclipse_periods(start, stop, prop)

Ascending and descending periods

from thistle import find_ascending_periods, find_descending_periods

ascending = find_ascending_periods(start, stop, prop)
descending = find_descending_periods(start, stop, prop)

All period functions return dicts with keys: start, stop.

Visibility circle

Compute the ground footprint where a satellite at a given altitude is visible above a minimum elevation angle:

from thistle import visibility_circle

lats, lons = visibility_circle(28.57, -80.65, alt=0.0, sat_alt=408_000, min_el=10.0)

Accuracy

TLE propagation

SGP4 propagation from TLEs is inherently limited to roughly 1-10 km near epoch, degrading to 10-100 km over days. For sub-kilometer accuracy, use precision ephemerides (SP3) instead of TLEs.

Time scale

Thistle treats UTC as UT1, introducing ~0.2-0.9 seconds of time offset (~1-7 m position error for LEO). This is negligible compared to TLE propagation errors and avoids the ~12x overhead of proper UTC/leap-second handling.

Coordinate system

ECI outputs use the ICRF (International Celestial Reference Frame), equivalent to J2000 for most applications.

License

MIT. See LICENSE for details.