Skip to contents

Growing season degree days per station-year

Source: R/wet_temp_gsdd.R
wet_temp_gsdd.Rd

Growing season degree days (GSDD) from daily mean water temperature, one value per station and year, computed by gsdd::gsdd() (Coleman & Fausch 2007): the season starts in the first week whose 7-day mean rises above 5 °C and stays there, and ends in the first week whose 7-day mean falls below 4 °C, and GSDD is the sum of the daily means over it. The result is in the long format of wet_window_stats(), so it goes to cd::cd_baseline(), cd::cd_anomaly() and cd::cd_trend() as is.

Usage

wet_temp_gsdd(
  x,
  value = "t_mean_c",
  by = "station_number",
  start_date = as.Date("1972-03-01"),
  end_date = as.Date("1972-11-30"),
  ...
)

Arguments

x

Daily water temperature, as from wet_temp_daily() or wet_temp_fill(): date, the value column and the by columns, and optionally filled.

value

Name of the daily mean column.

by

Id columns that separate series.

start_date, end_date

The part of each year to consider, as in gsdd::gsdd() (the year is ignored).

...

Passed to gsdd::gsdd(), e.g. start_temp, end_temp, pick or ignore_truncation.

Value

A data frame: the by columns, variable ("gsdd"), period ("growing_season"), year, value (°C-days, NA where the season is cut), anomaly_type ("absolute"), unit ("degC_day"), and over the days from start_date to end_date that have a value: n_days (how many), frac_ice (NA: there is no ice flag), frac_provisional (share whose status is "provisional", NA without a status column) and frac_filled (share that were filled, NA without a filled column). The columns are those of wet_window_stats() plus frac_filled. One row per series and year that gsdd::gsdd() returns. The interval of a wet_temp_fill() result does not carry into GSDD: daily bounds do not add up to bounds on a sum.

Details

A season cut by a missing day gives NA, not a value that is too low: days absent from x are missing, and gsdd::gsdd() uses the longest run with no missing day. Fill the gaps first with wet_temp_fill(). frac_filled then says how much of each value is modelled.

References

Coleman, M.A., and Fausch, K.D. 2007. Cold summer temperature limits recruitment of age-0 cutthroat trout in high-elevation Colorado streams. Transactions of the American Fisheries Society 136(5): 1231-1244. doi:10.1577/T05-244.1.

Examples

if (requireNamespace("gsdd", quietly = TRUE)) {
  date <- seq(as.Date("2019-01-01"), as.Date("2020-12-31"), by = "day")
  doy <- as.integer(format(date, "%j"))
  x <- data.frame(station_number = "08AA001", date = date,
                  t_mean_c = pmax(0, 6 + 9 * sin(2 * pi * (doy - 110) / 365)))
  wet_temp_gsdd(x)

  # A week missing in July 2020 leaves that season unknown
  gap <- date >= as.Date("2020-07-01") & date <= as.Date("2020-07-07")
  wet_temp_gsdd(x[!gap, ])
}
#>   station_number variable         period year    value anomaly_type     unit
#> 1        08AA001     gsdd growing_season 2019 2270.457     absolute degC_day
#> 2        08AA001     gsdd growing_season 2020       NA     absolute degC_day
#>   n_days frac_ice frac_provisional frac_filled
#> 1    275       NA               NA          NA
#> 2    268       NA               NA          NA