R code for SPI calculations from monthly gridded or point rainfall data

This item contains three R scripts: two with functions for SPI calculations, the other demonstrating the procedure used to compute and store SPI values for GPCC and CRU data over South-Western Southern Africa.

The script station_SPI.R contains a function used to compute SPI at a list of point locations, SPI_grid.R contains functions to read in a gridded netCDF monthly rainfall file, compute SPI series at all grid points and write the result to RDS and netCDF files on disk. The code in run_SPI.R uses the functions in SPI_grid.R to compute SPI values at various timescales.

When a function fun from a script file has been sourced, documentation for it can be viewed using: docstring(fun)

Suggested directory structure:

scripts should be run from ${root}/R_scripts as the working directory

|-- data
||- CRU_404_05deg_SWSA_16-28E_-35--23N.nc
|`- gpcc_V2020_SWSA_025deg_14-26E_-35.5--26N.nc
|-- figures
|-- R_scripts
||- run_SPI.R
||- SPI_grid.R
|`- station_SPI.R

Required R pacakges:
SCI
tidyverse

data.table

metR

docstring

magrittr

ncdf4

GPCC data may be obtained from: http://dx.doi.org/10.5676/DWD_GPCC/FD_M_V2020_025

CRU data are available from: https://crudata.uea.ac.uk/cru/data/hrg/

Using cdo (https://code.mpimet.mpg.de/projects/cdo)
the GPCC data (for example, if the original download is named gpcc.nc) can be processed to the required domain using: cdo sellonlatbox,14,26,-35.5,-26 gpcc.nc gpcc_V2020_SWSA_025deg_14-26E_-35.5--26N