Convert AEME MET_* meteorology to CF / CMIP variable names and units

Renames the AEME MET_* columns (met_vars()) to the short variable names used by CMIP6 / CORDEX and the CF conventions, and converts each to its CF-canonical unit. cf_to_met() is the inverse.

Usage

met_to_cf(
  met,
  vars = NULL,
  timestep = NULL,
  time_name = "time",
  outfile = NULL
)

Arguments

met

data frame with a date/time column plus MET_* columns, or a path to a .csv / .rds file holding one.

vars

optional subset of MET_* columns to convert; default every mapped column present.

timestep

seconds represented by one row, used only for the precipitation/snowfall flux conversion. NULL (default) infers it from the median spacing of the time column; pass it explicitly for an irregular series or a single row.

time_name

name for the time column in the output. Default "time" (CF); pass "Date" to keep the AEME name.

outfile

optional path; when given the result is written with utils::write.csv(row.names = FALSE) and returned invisibly.

Value

the data frame with CF short-name columns in CF units. Attributes cf_units and cf_standard_names (named by the output columns) and timestep_s record the conversion.

Details

AEME	CF / CMIP	CF standard_name	unit
MET_tmpair	tas	air_temperature	K
MET_tmpdew	tdps	dew_point_temperature	K
MET_humrel	hurs	relative_humidity	%
MET_prsttn	ps	surface_air_pressure	Pa
MET_prmslp	psl	air_pressure_at_mean_sea_level	Pa
MET_prvapr	pvap	water_vapor_partial_pressure_in_air	Pa
MET_wndspd	sfcWind	wind_speed	m s-1
MET_wnddir	sfcWindDir	wind_from_direction	degree
MET_wnduvu	uas	eastward_wind	m s-1
MET_wnduvv	vas	northward_wind	m s-1
MET_radswd	rsds	surface_downwelling_shortwave_flux_in_air	W m-2
MET_radlwd	rlds	surface_downwelling_longwave_flux_in_air	W m-2
MET_cldcvr	clt	cloud_area_fraction	%
MET_pprain	pr	precipitation_flux	kg m-2 s-1
MET_ppsnow	prsn	snowfall_flux	kg m-2 s-1

Unit changes: air and dew-point temperature degC -> K; vapour pressure hPa -> Pa; cloud cover fraction -> percent; precipitation and snowfall "mm per timestep" -> a kg m-2 s-1 flux (dividing by the timestep in seconds - 1 mm of water over 1 m2 is 1 kg). Wind, radiation and the surface/sea-level pressures are already in CF units and pass through unchanged. Columns that are not MET_* (an id column, say) are carried through untouched.

See also

cf_to_met() for the inverse, met_vars(), standardise_met() and guess_met_vars() for bringing arbitrary names to MET_*.

Examples

met <- data.frame(
  Date = seq(as.POSIXct("2024-01-01", tz = "UTC"), by = "hour", length.out = 5),
  MET_tmpair = c(15, 15.4, 16, 16.2, 15.8),
  MET_wndspd = c(3, 4, 3.5, 5, 4.2),
  MET_pprain = c(0, 0.2, 1.1, 0, 0.4))
met_to_cf(met)
#>                  time    tas sfcWind           pr
#> 1 2024-01-01 00:00:00 288.15     3.0 0.000000e+00
#> 2 2024-01-01 01:00:00 288.55     4.0 5.555556e-05
#> 3 2024-01-01 02:00:00 289.15     3.5 3.055556e-04
#> 4 2024-01-01 03:00:00 289.35     5.0 0.000000e+00
#> 5 2024-01-01 04:00:00 288.95     4.2 1.111111e-04