cesmDA: Climate data assimilation with CESM
In garciapintado/rdafCESM: Setup and Assimilation of CESM

Description Usage Arguments Details Value Author(s) References

This is the main function acting as interface between CESM and the assimilation via the package rDAF. Input here is very specific to CESM, and the user is referrred to CESM 1.2.2 documentation for those item specifically marked as CESM standard names.

The "inst" folder provides the folder "scripts" with the file cesm_setup.R, along with the folder "example_input". For a specific simulation, the file main.R is the starting point, showing an example of this package use.

cesmDA(envcase, DOUT_S_ROOT, comps, tavgs, tlimyStr, timenowStr,
       ilStr = NULL, ana, MC, prm, useGA, useGAy, yls, theta, MCgpar,
       analysis_scn = "a0", GET_ANALYSIS = TRUE, MEMORY_SAVE = TRUE,
       SAVE_NETCDF = TRUE, dsnrest = NULL, POP.zlrange=1)

`envcase`	[>=3] list (CCSMROOT , CASE[m], CASEROOT[m])
`DOUT_S_ROOT`	[m] CHARACTER; path to model data storage for IO. See CESM documentation
`comps`	[>=1] list with named elements. Names are model component to be analysed: in: c("atm", "ice", "lnd", "ocn", "rof"). Then, each element contains a [>=1] CHARACTER vector of variables to be analysed, matching variables archived in history files
`tavgs`	label to tag seasonal climate averages, in ("JFM","JAS","ann")
`tlimyStr`	[2] CHARACTER, close,open interval [tlimyStr) representing the time-averaged footprint of the observations. The format is AAAA-MM-DD HH:MM:ss
`timenowStr`	[1] CHARACTER, current timestamp in the above format
`ilStr`	CHARACTER, as two-digit loop index. For example '01' for an initial loop in an iterated scheme
`ana`	LIST, analysis variables, as defined in the package rDAF
`MC`	Data frame, as defined in the package rDAF
`prm`	LIST, assimilation global parameters, as defined in the package rDAF
`useGA`	LOGICAL whether to use Gaussian anamorphosis in analysed variables for Ensemble schemes
`useGAy`	LOGICAL, whether to also transform observations
`yls`	LIST, observations, as defined in the package rDAF
`theta`	LIST, see the cesm_setup.R in 'inst' folder in the package for an example
`MCgpar`	REAL, as defined in the package rDAF
`analysis_scn`	CHARACTER, label to identify the output files in the current assimilation scenario
`GET_ANALYSIS`	LOGICAL, whether to actually do the assimilation or only the data preparation and preliminar sensitivity analysis
`MEMORY_SAVE`	LOGICAL, whether to analysis spatial data layer by layer. For a few layer, this may be slower that analysing all layers at once, but commonly this should be set to TRUE so there is no limit in the number of variables and layer analysed regarding the saturation of memory in the R workspace
`SAVE_NETCDF`	LOGICAL, whether in addition to the R binary files, the netCDF files are to be rewritten. This is normally FALSE for single-step analysis, but TRUE for sequences of Data Assimilation Windows (DAWs)
`dsnrest`	CHARACTER, path to store update control variable files for restarting the simulation. See the example in the 'inst' folder
`POP.zlrange`	INTEGER, range of ocean layers to analyse, with the index 1 being the surface layer

The best way to understand how to use (adapt, extent) this function is via the example provided in the 'inst' folder in this package. Users are encouraged to contribute to the testing and development via the Github repository.

The function writes updated variables of CESM as system files or as netCDF files. In adition, an updated matrix of the control variables is returned by the function to the calling environment.

Javier Garcia-Pintado

Garcia-Pintado and Paul, 2018, GMD.

garciapintado/rdafCESM documentation built on July 18, 2019, 4:41 p.m.