Perform some checks on the inventory of CMIP5 files
Source:R/check_inventory_cmip5.R
check_inventory_cmip5.RdSome simple checks for multiple time frequencies, ensembles, and completeness of simulation periods. These checks are meant as guides only, since one might not wish multiple elements of the above for climate model ensemble assessments.
Arguments
- data_inventory
A data.table as resulting from
get_inventory.- check_hist
Boolean, if
TRUE, tests that each rcp* has a corresponding historical run.
Value
An object of class "eurocordexr_inv_check_cmip5" (an overloaded list) with results from the checks. Has a special print method, which shows a verbose summary of the results.
Details
The checks are
for multiple time frequency (day, month, ...)
for multiple ensembles (r1i1p1, r2i1p1, ...)
for complete periods of simulations: here complete means at least 1860-2005 for historical and 2006-2099 for rcp*
that each rcp* has a corresponding historical run (optional, off by default; otherwise problematic with merged hist and rcp runs)
Examples
# some empty example files
fn_zip <- system.file("extdata", "inv-test-files-cmip5.zip", package = "eurocordexr")
tmpdir <- tempdir()
unzip(fn_zip, exdir = tmpdir)
dat_inv <- get_inventory_cmip5(fs::path(tmpdir, "testdata-cmip5"))
check_inventory_cmip5(dat_inv)
#> Checks performed:
#> ------------------------------------------------------
#> ------------------------------------------------------
#> Multiple time frequencies detected: Amon day
#> ------------------------------------------------------
#> ------------------------------------------------------
#> Multiple ensembles in 3 cases:
#> variable gcm experiment timefreq N ensembles
#> <char> <char> <char> <char> <int> <char>
#> 1: pr ACCESS1-0 historical Amon 3 r1i1p1, r2i1p1, r3i1p1
#> 2: tas ACCESS1-0 historical Amon 3 r1i1p1, r2i1p1, r3i1p1
#> 3: tas ACCESS1-3 historical Amon 3 r1i1p1, r2i1p1, r3i1p1
#> ------------------------------------------------------
#> ------------------------------------------------------
#> Following model runs do not have complete periods:
#> variable timefreq gcm experiment ensemble nn_files date_start
#> <char> <char> <char> <char> <char> <int> <Date>
#> 1: pr Amon CMCC-CESM historical r1i1p1 2 1855-01-01
#> 2: tas Amon CMCC-CMS rcp85 r1i1p1 2 2070-01-01
#> 3: tasmax day HadGEM2-ES rcp85 r1i1p1 2 2279-12-01
#> 4: tasmin day HadGEM2-ES rcp26 r1i1p1 11 2005-12-01
#> date_end total_simulation_years period_contiguous
#> <Date> <int> <lgcl>
#> 1: 1879-12-01 10 FALSE
#> 2: 2100-12-01 21 FALSE
#> 3: 2299-12-30 12 FALSE
#> 4: 2139-11-30 112 FALSE
#> ------------------------------------------------------
#> ------------------------------------------------------
#> Finished checks.
#> ------------------------------------------------------
#> ------------------------------------------------------