Skip to contents

Calculate weather coefficients

Usage

get_weather_coefs(
  raw_bom_file,
  rolling_window = 60,
  meta_data,
  i_year = 2020,
  rainfall_na_fill = "mean"
)

Arguments

raw_bom_file

character, file path to raw bureau of Meteorology txt file

rolling_window

integer, number of days to summarise over a rolling window

meta_data

data.table Bureau of Meteorology, meta-data

i_year

integer, the year the coeffients are likely to be imputed. Defaults to 2020

rainfall_na_fill

numeric proportion, likihood of rain to fill NA values Defaults to rainfall_na_fill = "mean", which takes the overall mean proportion.

Value

data.table, of coefficients to estimate weather during a rainfall event. If no rainfall data is recorded in the raw weather NULL is returned without warning. wd_rw, mean wind direction from rolling window; wd_sd_rw, standard deviation of wind direction from rolling window; ws_rw, mean wind speed from rolling window; ws_sd_rw, standard deviation of speed from rolling window; rain_freq, historical probability of rainfall on this day based on a rolling window.

A data.table of coefficients to estimate weather during a rainfall event. If no rainfall data is recorded in the raw weather NULL is returned without warning. station_name - Weather station name; lat - latitude; lon - longitude; state - political juristiction or state; yearday - integer, day of the year, see data.table::yday(); temp - numeric, mean temperature; rh - numeric, mean temperature; wd_rd - numeric, mean wind direction from rolling window; wd_sd_rd - numeric, standard deviation of wind direction from rolling window; ws_rd - numeric, mean wind speed from rolling window; ws_sd_rd - numeric, standard deviation of speed from rolling window; rain_freq - numeric, proportional chance of rainfall on this day 0 - 1 based on a rolling window.

Details

get_weather_coefs uses historical bom rainfall data to determine the probability of rainfall on each day of the year. It also summarises the mean temperatures, wind speed and direction at the time of the rainfall. Formally called impute_rainywind

Examples

if (FALSE) { # \dontrun{
library(data.table)
meta_dat <- fread("cache/bom_stations.csv")
imp_dat <-
   get_weather_coefs(raw_bom_file = "./data/Weather_data/HD01D_Data_090182_999999999959761.txt",
                    rolling_window = 60,
                    meta_data = meta_dat)} # }