Finding data

Overview

Data discovery and retrieval is the first step in any evaluation process; ESMValTool uses a semi-automated data finding mechanism with inputs from both the user configuration file and the recipe file: this means that the user will have to provide the tool with a set of parameters related to the data needed and once these parameters have been provided, the tool will automatically find the right data. We will detail below the data finding and retrieval process and the input the user needs to specify, giving examples on how to use the data finding routine under different scenarios.

CMIP data - CMOR Data Reference Syntax (DRS) and the ESGF

CMIP data is widely available via the Earth System Grid Federation (ESGF) and is accessible to users either via download from the ESGF portal or through the ESGF data nodes hosted by large computing facilities (like CEDA-Jasmin, DKRZ, etc). This data adheres to, among other standards, the DRS and Controlled Vocabulary standard for naming files and structured paths; the DRS ensures that files and paths to them are named according to a standardized convention. Examples of this convention, also used by ESMValTool for file discovery and data retrieval, include:

  • CMIP6 file: [variable_short_name]_[mip]_[dataset_name]_[experiment]_[ensemble]_[grid]_[start-date]-[end-date].nc

  • CMIP5 file: [variable_short_name]_[mip]_[dataset_name]_[experiment]_[ensemble]_[start-date]-[end-date].nc

  • OBS file: [project]_[dataset_name]_[type]_[version]_[mip]_[short_name]_[start-date]-[end-date].nc

Similar standards exist for the standard paths (input directories); for the ESGF data nodes, these paths differ slightly, for example:

  • CMIP6 path for BADC: ROOT-BADC/[institute]/[dataset_name]/[experiment]/[ensemble]/[mip]/ [variable_short_name]/[grid];

  • CMIP6 path for ETHZ: ROOT-ETHZ/[experiment]/[mip]/[variable_short_name]/[dataset_name]/[ensemble]/[grid]

From the ESMValTool user perspective the number of data input parameters is optimized to allow for ease of use. We detail this procedure in the next section.

Data retrieval

Data retrieval in ESMValTool has two main aspects from the user’s point of view:

  • data can be found by the tool, subject to availability on disk;

  • it is the user’s responsibility to set the correct data retrieval parameters;

The first point is self-explanatory: if the user runs the tool on a machine that has access to a data repository or multiple data repositories, then ESMValTool will look for and find the available data requested by the user.

The second point underlines the fact that the user has full control over what type and the amount of data is needed for the analyses. Setting the data retrieval parameters is explained below.

Setting the correct root paths

The first step towards providing ESMValTool the correct set of parameters for data retrieval is setting the root paths to the data. This is done in the user configuration file config-user.yml. The two sections where the user will set the paths are rootpath and drs. rootpath contains pointers to CMIP, OBS, default and RAWOBS root paths; drs sets the type of directory structure the root paths are structured by. It is important to first discuss the drs parameter: as we’ve seen in the previous section, the DRS as a standard is used for both file naming conventions and for directory structures.

Explaining config-user/drs: CMIP5: or config-user/drs: CMIP6:

Whereas ESMValTool will always use the CMOR standard for file naming (please refer above), by setting the drs parameter the user tells the tool what type of root paths they need the data from, e.g.:

drs:
  CMIP6: BADC

will tell the tool that the user needs data from a repository structured according to the BADC DRS structure, i.e.:

ROOT/[institute]/[dataset_name]/[experiment]/[ensemble]/[mip]/[variable_short_name]/[grid];

setting the ROOT parameter is explained below. This is a strictly-structured repository tree and if there are any sort of irregularities (e.g. there is no [mip] directory) the data will not be found! BADC can be replaced with DKRZ or ETHZ depending on the existing ROOT directory structure. The snippet

drs:
  CMIP6: default

is another way to retrieve data from a ROOT directory that has no DRS-like structure; default indicates that the data lies in a directory that contains all the files without any structure.

Note

When using CMIP6: default or CMIP5: default it is important to remember that all the needed files must be in the same top-level directory set by default (see below how to set default).

Explaining config-user/rootpath:

rootpath identifies the root directory for different data types (ROOT as we used it above):

  • CMIP e.g. CMIP5 or CMIP6: this is the root path(s) to where the CMIP files are stored; it can be a single path or a list of paths; it can point to an ESGF node or it can point to a user private repository. Example for a CMIP5 root path pointing to the ESGF node on CEDA-Jasmin (formerly known as BADC):

    CMIP5: /badc/cmip5/data/cmip5/output1
    

    Example for a CMIP6 root path pointing to the ESGF node on CEDA-Jasmin:

    CMIP6: /badc/cmip6/data/CMIP6/CMIP
    

    Example for a mix of CMIP6 root path pointing to the ESGF node on CEDA-Jasmin and a user-specific data repository for extra data:

    CMIP6: [/badc/cmip6/data/CMIP6/CMIP, /home/users/johndoe/cmip_data]
    
  • OBS: this is the root path(s) to where the observational datasets are stored; again, this could be a single path or a list of paths, just like for CMIP data. Example for the OBS path for a large cache of observation datasets on CEDA-Jasmin:

    OBS: /gws/nopw/j04/esmeval/obsdata-v2
    
  • default: this is the root path(s) to where files are stored without any DRS-like directory structure; in a nutshell, this is a single directory that should contain all the files needed by the run, without any sub-directory structure.

  • RAWOBS: this is the root path(s) to where the raw observational data files are stored; this is used by cmorize_obs.

Dataset definitions in recipe

Once the correct paths have been established, ESMValTool collects the information on the specific datasets that are needed for the analysis. This information, together with the CMOR convention for naming files (see CMOR-DRS) will allow the tool to search and find the right files. The specific datasets are listed in any recipe, under either the datasets and/or additional_datasets sections, e.g.

datasets:
  - {dataset: HadGEM2-CC,  project: CMIP5, exp: historical, ensemble: r1i1p1, start_year: 2001, end_year: 2004}
  - {dataset: UKESM1-0-LL, project: CMIP6, exp: historical, ensemble: r1i1p1f2, grid: gn, start_year: 2004,  end_year: 2014}

_data_finder will use this information to find data for all the variables specified in diagnostics/variables.

Recap and example

Let us look at a practical example for a recap of the information above: suppose you are using a config-user.yml that has the following entries for data finding:

rootpath:  # running on CEDA-Jasmin
  CMIP6: /badc/cmip6/data/CMIP6/CMIP
drs:
  CMIP6: BADC  # since you are on CEDA-Jasmin

and the dataset you need is specified in your recipe.yml as:

- {dataset: UKESM1-0-LL, project: CMIP6, mip: Amon, exp: historical, grid: gn, ensemble: r1i1p1f2, start_year: 2004,  end_year: 2014}

for a variable, e.g.:

diagnostics:
  some_diagnostic:
    description: some_description
    variables:
      ta:
        preprocessor: some_preprocessor

The tool will then use the root path /badc/cmip6/data/CMIP6/CMIP and the dataset information and will assemble the full DRS path using information from CMOR-DRS and establish the path to the files as:

/badc/cmip6/data/CMIP6/CMIP/MOHC/UKESM1-0-LL/historical/r1i1p1f2/Amon

then look for variable ta and specifically the latest version of the data file:

/badc/cmip6/data/CMIP6/CMIP/MOHC/UKESM1-0-LL/historical/r1i1p1f2/Amon/ta/gn/latest/

and finally, using the file naming definition from CMOR-DRS find the file:

/badc/cmip6/data/CMIP6/CMIP/MOHC/UKESM1-0-LL/historical/r1i1p1f2/Amon/ta/gn/latest/ta_Amon_UKESM1-0-LL_historical_r1i1p1f2_gn_195001-201412.nc

Observational data

Observational data is retrieved in the same manner as CMIP data, for example using the OBS root path set to:

OBS: /gws/nopw/j04/esmeval/obsdata-v2

and the dataset:

- {dataset: ERA-Interim,  project: OBS,  type: reanaly,  version: 1,  start_year: 2014,  end_year: 2015,  tier: 3}

in recipe.yml in datasets or additional_datasets, the rules set in CMOR-DRS are used again and the file will be automatically found:

/gws/nopw/j04/esmeval/obsdata-v2/Tier3/ERA-Interim/OBS_ERA-Interim_reanaly_1_Amon_ta_201401-201412.nc

Since observational data are organized in Tiers depending on their level of public availability, the default directory must be structured accordingly with sub-directories TierX (Tier1, Tier2 or Tier3), even when drs: default.

Data loading

Data loading is done using the data load functionality of iris; we will not go into too much detail about this since we can point the user to the specific functionality here but we will underline that the initial loading is done by adhering to the CF Conventions that iris operates by as well (see CF Conventions Document and the search page for CF standard names).

Data concatenation from multiple sources

Oftentimes data retrieving results in assembling a continuous data stream from multiple files or even, multiple experiments. The internal mechanism through which the assembly is done is via cube concatenation. One peculiarity of iris concatenation (see iris cube concatenation) is that it doesn’t allow for concatenating time-overlapping cubes; this case is rather frequent with data from models overlapping in time, and is accounted for by a function that performs a flexible concatenation between two cubes, depending on the particular setup:

  • cubes overlap in time: resulting cube is made up of the overlapping data plus left and right hand sides on each side of the overlapping data; note that in the case of the cubes coming from different experiments the resulting concatenated cube will have composite data made up from multiple experiments: assume [cube1: exp1, cube2: exp2] and cube1 starts before cube2, and cube2 finishes after cube1, then the concatenated cube will be made up of cube2: exp2 plus the section of cube1: exp1 that contains data not provided in cube2: exp2;

  • cubes don’t overlap in time: data from the two cubes is bolted together;

Note that two cube concatenation is the base operation of an iterative process of reducing multiple cubes from multiple data segments via cube concatenation ie if there is no time-overlapping data, the cubes concatenation is performed in one step.