Convert downloaded ERA5 files to a daily meteorology data frame

Convenience wrapper that runs extract_era5_hourly_met() over the ERA5 files in path (netCDF, or GRIB as written by download_era5_cds()) and aggregates the model-ready hourly frame to daily with met_to_daily(): precipitation and snowfall are summed, everything else is averaged, and - unless minmax = FALSE - daily minimum and maximum air and dewpoint temperature are added.

File discovery (by name pattern), the per-file reader backend, spatial sampling (method), de-accumulation of the flux variables, unit conversion and the UTC -> tz shift are all handled by extract_era5_hourly_met(); see there for the details and for the arguments passed through ....

Usage

convert_era5_netcdf(
  path,
  format = c("AEME", "LER", "raw"),
  tz = "UTC",
  minmax = TRUE,
  min_frac = 0.5,
  site = NULL,
  outfile = NULL,
  ...
)

Arguments

path

directory holding the ERA5 files (netCDF or GRIB).

format

"AEME" (default, MET_* names, a Date column), "LER" (LakeEnsemblR names) or "raw" (ERA5 short names); "LER" and "raw" return a datetime column.

tz

output / aggregation time zone: passed to extract_era5_hourly_met() and used to assign calendar days. Default "UTC".

minmax

add daily min / max columns for air and dewpoint temperature (MET_airmin / MET_airmax / MET_dewmin / MET_dewmax in AEME naming). Default TRUE.

min_frac

drop days with fewer than this fraction of the expected hourly records, passed to met_to_daily(). Default 0.5; set 0 to keep every day.

site

optional site tag; when given (and pattern is left at its default) only files whose name contains the tag are used - a shortcut for pattern = paste0("*{variable}*", site, "*").

outfile

optional path; if given the daily frame is written there with utils::write.csv() (row.names = FALSE, dates as "%Y-%m-%d").

...

further arguments for extract_era5_hourly_met(): lon, lat, geom, years, months, variables, method, precip_units, pressure_units, pattern, max_dist_km, area_crs, fill_gaps, verbose.

Value

a daily data frame: the time column first, then one column per variable in the chosen format, plus the temperature min / max columns when minmax = TRUE. The lon / lat / tz / method attributes from extract_era5_hourly_met() are carried through.

See also

extract_era5_hourly_met() for the hourly frame and the full argument list, met_to_daily() for the aggregation, standardise_met() which recognises the "raw" ERA5 short names directly.

Examples

if (FALSE) { # \dontrun{
lon <- 98.67591; lat <- 2.637047            # Lake Toba, Indonesia
met <- convert_era5_netcdf(
  path = "data/test", lon = lon, lat = lat, years = 2024,
  variables = "2m_temperature")
} # }