arthur-e / suntransit

Simple, fast approximation of sunrise, sunset time on Earth

suntransit

Simple, fast approximation of sunrise, sunset time on Earth.

Earth system models and other biogeochemical simulations often require an estimate of the day length and, therefore, of sunrise and sunset times. Highly precise solar transit information can be obtained from Python libraries like pyephem and astral, but these are general-purpose libraries and the corresponding routines are slow. If we can tolerate an error in transit times on the order of a few minutes, then we should also be able to calculate transit times much faster! That's the idea here: using a fast approximation for sunrise and sunset times. Read more about the alternatives and the performance of an earlier version of this library.

Example Use

import datetime
import numpy as np
from suntransit import sunrise_sunset

missoula = (46.8625, -114.0117)
today = datetime.date(2021, 10, 7)

# Get (sunrise, sunset) hour in UTC
sunrise_sunset(missoula, today)
# (13.733140671716853, 1.062543659843307)

# Get (sunrise, sunset) hour in local time (GMT-6:00)
(np.array(sunrise_sunset(missoula, today)) - 6) % 24
# array([ 7.73314067, 19.06254366])

And for (quasi-)vectorization:

from functools import partial

calgary = (51.0458, -114.0575)
ankara = (39.93, 32.85)

data = np.stack([missoula, calgary, ankara], axis = 0)

alt_sunrise_sunset = partial(sunrise_sunset, dt = today)
np.apply_along_axis(alt_sunrise_sunset, 1, data).round(1)
# array([[13.7,  1.1],
#        [13.8,  1. ],
#        [ 3.9, 15.4]])

Installation

It is recommended to install with pip:

pip install suntransit

The only dependency is numpy.

Documentation

Online documentation can be found here.

Testing

python tests/tests.py

References

Meeus, Jean. "Astronomical Algorithms." 1991. William-Bell Inc. Richmond, Virginia, U.S.A.

U.S. Naval Observatory. "Almanac for Computers." 1990. Reproduced by Ed Williams. https://www.edwilliams.org/sunrise_sunset_algorithm.htm