In ropenscilabs/cmipr: Client for Coupled Model Intercomparison Project (CMIP) Data

cmipr introduction

cmipr - Client to work with CMIP data - downscaled climate and hydrology projections. Package lists avail. files, downloads and caches, and reads into raster objects.

Install

Dev version

devtools::install_github("ropenscilabs/cmipr")

library("cmipr")

List files

List files in their FTP directory

cmip_list_files()
#> # A tibble: 9 × 2
#>         date           file
#>       <date>          <chr>
#> 1 2014-04-02 bcsd/1950-2099
#> 2 2014-04-02      bcsd/2deg
#> 3 2015-06-02               
#> 4 2014-04-02               
#> 5 2011-08-23               
#> 6 0016-08-26               
#> 7 2014-06-20               
#> 8 2014-06-17               
#> 9 2014-04-02    bcsd/yearly

cmip_list_files('cmip5/bcsd')
#> # A tibble: 4 × 2
#>         date    file
#>       <date>   <chr>
#> 1 2014-03-28      BC
#> 2 2014-03-28    BCSD
#> 3 2014-03-28 SD_noBC
#> 4 2014-03-28  regrid

Fetch data

Define a path to a file.

key <- "bcsd/yearly/cnrm_cm3.1/cnrm_cm3.1.sresa1b.monthly.Prcp.2034.nc"

Fetch the file

(res <- cmip_fetch(key))
#> <CMIP file>
#>    File: /Users/sacmac/Library/Caches/cmipr/cnrm_cm3.1.sresa1b.monthly.Prcp.2034.nc
#>    File size: 4.93842 MB

Caching

When requesting data, we first check if you already have that data cached. You'll know when this happens as the request will finish much faster when you have data cached already.

We use the package hoardr to manage cached files.

On package load, a hoardr object is created, it's an R6 object.

cmip_cache
#> <hoard> 
#>   path: cmipr
#>   cache path: /Users/sacmac/Library/Caches/cmipr

See ?cmip_cache for all cache management help.

The cmip_cache object has variables and functions on it. For example, we can get the cache path

cmip_cache$cache_path_get()
#> [1] "/Users/sacmac/Library/Caches/cmipr"

Or list all files in the cache

cmip_cache$list()
#> [1] "/Users/sacmac/Library/Caches/cmipr/cnrm_cm3.1.sresa1b.monthly.Prcp.2034.nc"

Or delete all files (won't run this here though)

cmip_cache$delete_all()

Read data

After fetching data, you need to read the data into a RasterLayer or RasterBrick object

Get some data first

keys <- c(
  "bcsd/yearly/cnrm_cm3.1/cnrm_cm3.1.sresa1b.monthly.Prcp.2039.nc",
  "bcsd/yearly/gfdl_cm2_1.1/gfdl_cm2_1.1.sresa1b.monthly.Prcp.2033.nc"
)
res <- lapply(keys, cmip_fetch)

You can read a single file

cmip_read(res[[1]])
#> class       : RasterLayer 
#> band        : 1  (of  12  bands)
#> dimensions  : 222, 462, 102564  (nrow, ncol, ncell)
#> resolution  : 0.125, 0.125  (x, y)
#> extent      : -124.75, -67, 25.125, 52.875  (xmin, xmax, ymin, ymax)
#> coord. ref. : +proj=longlat +datum=WGS84 +ellps=WGS84 +towgs84=0,0,0 
#> data source : /Users/sacmac/Library/Caches/cmipr/cnrm_cm3.1.sresa1b.monthly.Prcp.2039.nc 
#> names       : Prcp 
#> z-value     : 2039-01-16 
#> zvar        : Prcp

many files

cmip_read(unlist(res))
#> class       : RasterBrick 
#> dimensions  : 222, 462, 102564, 24  (nrow, ncol, ncell, nlayers)
#> resolution  : 0.125, 0.125  (x, y)
#> extent      : -124.75, -67, 25.125, 52.875  (xmin, xmax, ymin, ymax)
#> coord. ref. : +proj=longlat +datum=WGS84 +ellps=WGS84 +towgs84=0,0,0 
#> data source : in memory
#> names       :       Prcp.1,       Prcp.2,       Prcp.3,       Prcp.4,       Prcp.5,       Prcp.6,       Prcp.7,       Prcp.8,       Prcp.9,      Prcp.10,      Prcp.11,      Prcp.12,      Prcp.13,      Prcp.14,      Prcp.15, ... 
#> min values  : 0.0000000000, 0.0014285714, 0.0000000000, 0.0000000000, 0.0000000000, 0.0000000000, 0.0000000000, 0.0010645161, 0.0000000000, 0.0000000000, 0.0000000000, 0.0018709678, 0.0648709685, 0.0221071430, 0.0000000000, ... 
#> max values  :    36.971066,    16.024536,    30.073355,    18.687033,    19.824322,    11.378433,     8.971000,    13.397967,    12.315033,    53.819515,    43.141335,    38.081871,    29.059355,    24.486929,    24.351355, ...

Visualize

library("raster")
plot(cmip_read(res[[1]]))

ropenscilabs/cmipr documentation built on May 18, 2022, 7:35 p.m.