Add sun and moon data

This article shows how to add sun and moon data using the suncalc package. We show how to add sunrise and sunset times to split the data between day and night, and how to add the moon phase to assign days around full and new moon. This can be useful for example to investigate whether birds behave differently during day and night or during different moon phases, which are closely linked to the strength of the tides.

Load packages and data

# packages
library(tools4watlas)
library(suncalc)

# load example data
data <- data_example

Add sunrise and sunset data

We first need to add the coordinates in WGS 84 (EPSG:4326) to be able to calculate sunrise and sunset times. Then, we use the getSunlightTimes() function to calculate sunrise and sunset times for each position. Finally, we assign day and night based on whether the timestamp of a position falls between sunrise and sunset. The function also offers many more options, see getSunlightTimes() documentation for more details.

# add transformed coordinates in EPSG:4326 (WGS 84)
data <- atl_transform_dt(data, to = sf::st_crs(4326))

# add sunrise/sunset times
data[, c("sunrise", "sunset") := getSunlightTimes(
  data = data.table(date = as.Date(datetime), lat = y_4326, lon = x_4326),
  keep = c("sunrise", "sunset"),
  tz   = "UTC"
)[, c("sunrise", "sunset")]]

# assign day and night
data[, day_night := fifelse(
  datetime >= sunrise & datetime < sunset, "day", "night"
)]

# show head of the table
head(data[, .(tag, datetime, sunrise, sunset, day_night)]) |>
  knitr::kable(digits = 2)

tag	datetime	sunrise	sunset	day_night
3027	2023-09-23 03:13:25	2023-09-23 05:26:50	2023-09-23 17:38:51	night
3027	2023-09-23 03:13:28	2023-09-23 05:26:50	2023-09-23 17:38:51	night
3027	2023-09-23 03:19:49	2023-09-23 05:26:50	2023-09-23 17:38:51	night
3027	2023-09-23 03:19:52	2023-09-23 05:26:50	2023-09-23 17:38:51	night
3027	2023-09-23 03:19:55	2023-09-23 05:26:50	2023-09-23 17:38:51	night
3027	2023-09-23 03:19:58	2023-09-23 05:26:50	2023-09-23 17:38:51	night

Add moon phase data

We use the getMoonIllumination() function to calculate the moon phase for each timestamp. The moon phase is a value between 0 and 1, where 0 = new moon, 0.25 = first quarter, 0.5 = full moon, and 0.75 = last quarter. We then calculate the time to the next full moon and new moon in days, which can be useful to assign days around full and new moon.

# add moon phase per datetime
# 0 = new, 0.25 = first quarter, 0.5 = full, 0.75 = last quarter
data[, c("moon_phase") := getMoonIllumination(
  date = datetime,
  keep = c("phase")
)[, c("phase")]]

# add time to full moon (phase = 0.5) and new moon (phase = 0)
# Moon cycle = 29.53 days, negative = before full or new moon
data[, time2fullmoon := (moon_phase - 0.5) * 29.53]
data[, time2newmoon := fifelse(
  moon_phase <= 0.5,
  moon_phase * 29.53,
  (moon_phase - 1) * 29.53
)]

# assign 3 days around full and new moon
data[, near_lunar_extreme := abs(time2fullmoon) <= 3 | abs(time2newmoon) <= 3]

# show head of the table
head(data[, .(
  tag, datetime, moon_phase, time2fullmoon, time2newmoon, near_lunar_extreme
)]) |>
  knitr::kable(digits = 2)

tag	datetime	moon_phase	time2fullmoon	time2newmoon	near_lunar_extreme
3027	2023-09-23 03:13:25	0.27	-6.88	7.88	FALSE
3027	2023-09-23 03:13:28	0.27	-6.88	7.88	FALSE
3027	2023-09-23 03:19:49	0.27	-6.88	7.89	FALSE
3027	2023-09-23 03:19:52	0.27	-6.88	7.89	FALSE
3027	2023-09-23 03:19:55	0.27	-6.88	7.89	FALSE
3027	2023-09-23 03:19:58	0.27	-6.88	7.89	FALSE