Converting NetCDF data

library(priogridR)

pg_from_netcdf() accepts a NetCDF path or an already opened terra::SpatRaster. It can select a NetCDF subdataset, choose layers, recover raster time metadata, replace sentinel values, and then use the same alignment workflow as pg_from_raster().

The underlying raster transformation requires a priogrid working directory. Configure a persistent project location before running the conversion examples:

priogrid::pg_set_rawfolder("data/priogrid")

The conversion chunks are not evaluated while the vignette is built because the working directory is a persistent, user-specific setting.

An in-memory example

Using a SpatRaster makes this example independent of platform-specific NetCDF drivers while exercising the complete conversion path.

config <- priogrid::pg_config(
  nrow = 18,
  ncol = 36,
  temporal_resolution = "1 year",
  start_date = as.Date("2020-12-31"),
  end_date = as.Date("2021-12-31"),
  verbose = FALSE
)

precipitation <- terra::rast(
  nrows = 18,
  ncols = 36,
  nlyrs = 2,
  xmin = -180,
  xmax = 180,
  ymin = -90,
  ymax = 90,
  crs = "EPSG:4326"
)
terra::values(precipitation) <- cbind(
  rep(10, terra::ncell(precipitation)),
  rep(12, terra::ncell(precipitation))
)

Supply explicit dates when the object does not contain reliable time metadata:

precipitation_pg <- pg_from_netcdf(
  x = precipitation,
  variable = "precipitation",
  aggregation = "mean",
  time = c(2020, 2021),
  config = config
)

head(precipitation_pg)

Reading a NetCDF file

When a file contains one main variable, pass its path directly:

precipitation_pg <- pg_from_netcdf(
  x = "data/raw/precipitation.nc",
  variable = "precipitation",
  aggregation = "mean",
  config = config
)

For files with multiple subdatasets, identify the required one with source_variable:

temperature_pg <- pg_from_netcdf(
  x = "data/raw/climate.nc",
  source_variable = "temperature",
  variable = "temperature",
  aggregation = "mean",
  layer = 1:12,
  config = config
)

Inspect available subdatasets with terra::sds() or the file metadata before choosing source_variable.

Time, layers, and missing values

If valid time metadata are present, the function uses them automatically. Otherwise, supply one time value per selected layer. The layer argument can select layers by position or name before time validation.

Use na_values for encoded missing values:

precipitation_pg <- pg_from_netcdf(
  x = "data/raw/precipitation.nc",
  variable = "precipitation",
  aggregation = "mean",
  na_values = c(-9999, 9.96921e+36),
  config = config
)

Choose aggregation and disaggregation rules according to whether the variable is continuous, cumulative, or categorical. See vignette("pg_from_raster", package = "priogridR") for detailed guidance.