On met.no climate observation data is stored as a single time step per file with a 6 hour resolution. These individual files can be retrieved efficiently by first subsetting to a smaller area, then downloading the data and merging it into a single CFVariable:
library(ncdfCF)
# Open the two files
ds <- open_ncdf("https://thredds.met.no/thredds/dodsC/nora3/1961/08/01/00/fc1961080100_003_sfx.nc")
ds06 <- open_ncdf("https://thredds.met.no/thredds/dodsC/nora3/1961/08/01/06/fc1961080106_003_sfx.nc")
# Open TS variables - no data is being read
ts <- ds[["TS"]]
ts06 <- ds06[["TS"]]
ts
#> <Variable> TS
#> Long name: Surface temperature
#>
#> Values: (not loaded)
#>
#> Coordinate reference system:
#> name grid_mapping
#> projection_lambert lambert_conformal_conic
#>
#> Axes:
#> axis name long_name length values unit
#> X x x-coordinate in Cartesian system 889 [778360.875 ... 3442360.75] m
#> Y y y-coordinate in Cartesian system 1489 [-1270477 ... 3193523] m
#> grib1_vLevel1052 grib1-level 105 1 [0]
#> T time 1-U [1961-08-01T03:00:00] seconds since 1970-01-01 00:00:00 +00:00
#>
#> Auxiliary longitude-latitude grid:
#> axis name extent unit
#> X longitude [-30.168 ... 85.793] degrees_east
#> Y latitude [44.025 ... 84.057] degrees_north
#>
#> Attributes:
#> name type length value
#> units NC_CHAR 1 K
#> standard_name NC_CHAR 15 air_temperature
#> long_name NC_CHAR 19 Surface temperature
#> grid_mapping NC_CHAR 18 projection_lambert
#> coordinates NC_CHAR 18 longitude latitude
#> _ChunkSizes NC_INT 4 1, 1, 1489, 889
# Subset to a smaller area of 200 x 400km - still no data being read
tssmall <- ts$subset(x = c(2400000, 2600000), y = c(0, 400000))
ts06small <- ts06$subset(x = c(2400000, 2600000), y = c(0, 400000))
tssmall
#> <Variable> TS
#> Long name: Surface temperature
#>
#> Values: (not loaded)
#>
#> Coordinate reference system:
#> name grid_mapping
#> projection_lambert lambert_conformal_conic
#>
#> Axes:
#> axis name long_name length values unit
#> X x x-coordinate in Cartesian system 67 [2401360.75 ... 2599360.75] m
#> Y y y-coordinate in Cartesian system 133 [1523.02124 ... 397523.03125] m
#> grib1_vLevel1052 grib1-level 105 1 [0]
#> T time 1-U [1961-08-01T03:00:00] seconds since 1970-01-01 00:00:00 +00:00
#>
#> Attributes:
#> name type length value
#> units NC_CHAR 1 K
#> standard_name NC_CHAR 15 air_temperature
#> long_name NC_CHAR 19 Surface temperature
#> grid_mapping NC_CHAR 18 projection_lambert
#> coordinates NC_CHAR 18 longitude latitude
#> _ChunkSizes NC_INT 4 1, 1, 1489, 889
# Merge the CFVariables along the time dimension - this forces data to be read
# from the server. Note that the "time" axis is now of length 2.
tssmall$append(ts06small, "time")
#> Warning in min(x): no non-missing arguments to min; returning Inf
#> Warning in max(x): no non-missing arguments to max; returning -Inf
tssmall
#> <Variable> TS
#> Long name: Surface temperature
#>
#> Values: [272.76 ... 298.79] K
#> NA: 5704 (32.0%)
#>
#> Coordinate reference system:
#> name grid_mapping
#> projection_lambert lambert_conformal_conic
#>
#> Axes:
#> axis name long_name length values unit
#> X x x-coordinate in Cartesian system 67 [2401360.75 ... 2599360.75] m
#> Y y y-coordinate in Cartesian system 133 [1523.02124 ... 397523.03125] m
#> grib1_vLevel1052 grib1-level 105 1 [0]
#> T time 2 [1961-08-01T03:00:00 ... 1961-08-01T09:00:00] seconds since 1970-01-01 00:00:00 +00:00
#>
#> Attributes:
#> name type length value
#> units NC_CHAR 1 K
#> standard_name NC_CHAR 15 air_temperature
#> long_name NC_CHAR 19 Surface temperature
#> grid_mapping NC_CHAR 18 projection_lambert
#> coordinates NC_CHAR 18 longitude latitude
#> _ChunkSizes NC_INT 4 1, 1, 1489, 889
#> actual_range NC_DOUBLE 2 272.76, 298.79
On met.no climate observation data is stored as a single time step per file with a 6 hour resolution. These individual files can be retrieved efficiently by first subsetting to a smaller area, then downloading the data and merging it into a single
CFVariable: