Computing composite indices (wsim_composite)ΒΆ

The wsim_composite utility computes Composite Indices from multiple surplus and deficit input variables.

The composite indices are output to a netCDF file with the following variables:

surplus

the worst (most positive) return period of all input surplus variables

surplus_cause

a coded variable indicating which of the surplus variables was responsible for the composite surplus indicator value

deficit

the worst (most negative) return period of all input deficit variables

deficit_cause

a coded variable indicating which of the deficit variables was responsible for the composite deficit indicator value

both

the greater magnitude of surplus and deficit, whenever both of these indices have a magnitude higher than a specified threshold

Usage is as follows:

Usage: wsim_composite \
    (--surplus=<file>)... \
    (--deficit=<file>)... \
    --both_threshold=<value> \
    [--mask=<file>] \
    --output=<file>

The --surplus and --deficit arguments are used to specify variables that should be considered in computing the composite surplus and deficit indices. each argument may refer to more than one variable, and each argument may be provided multiple times. The WSIM variable definition notation is fully supported. The --both_threshold argument specifies the minimum magnitude of the composite surplus and deficit indices for a pixel to be considered to simultaneously experience surplus and deficit.

The --mask allows an optional mask to be defined, so that all composite indices have the same maximum extent even if the input datasets do not.

As a more complete example, the following command was is used to produce the WSIM composite indices for January 2017 from multiple raster inputs:

./wsim_composite.R \
  --deficit "~/freq/PETmE_freq_trgt201701.img::1@fill0@negate->Neg_PETmE" \
  --deficit "~/freq/Ws_freq_trgt201701.img::1->Ws" \
  --deficit "~/freq/Bt_RO_freq_trgt201701.img::1->Bt_RO" \
  --surplus "~/freq/Bt_RO_freq_trgt201701.img::1->Bt_RO" \
  --surplus "~/freq/RO_mm_freq_trgt201701.img::1->RO_mm" \
  --mask "~/freq/Ws_freq_trgt201701.img" \
  --both_threshold 3 \
  --output "/tmp/composite_201701.nc"

A few points are worth noting in this example:

  • The Bt_RO variable is considered in computation of both the surplus and deficit indices.

  • The PETmE variable is transformed by filling NODATA values with zero, and then negating all values.

  • The Ws variable was used as a mask for the composite indices. If this were not done, the composite deficit would be populated in all pixels, because of the transformation done to PETmE.