-scaling¶

Introduction¶

In grib, grid values are usually stored as

  grid_value = (ref_value + I*2**bin_scaling)) * 10**decimal_scaling.
    I is an positive integer
    ref_value = 32-bit IEEE floating point number
    binary_scaling = -127..127
    decimal_scaling = -127..127
    number_of_bits = the smallest N such that I < 2**N
      note: WMO does not place a maximum value of the number of bits
            however, implementations do have limits.  The limits will
            vary by software package and procedure.

This is not an absolute as the grid point values can be stored as a spectral coefficients, IEEE floating point values and other formats. The -scaling option prints the binary and scaling factors, the reference value and the number of bits used for the integer I.

Usage¶

-scaling

Example¶

$ wgrib2 gep19.t00z.pgrb2af180 -scaling -for 13:14 -v -packing
13:473374:scaling ref=2.32109e+06 dec_scale=-3 bin_scale=4 nbits=16:packing=grid point data - jpeg2000 compression,j val=(2.32109e+06+i*2^4)*10^-3, i=0..65535 (#bits=16)
14:528630:scaling ref=2215 dec_scale=-1 bin_scale=0 nbits=10:packing=grid point data - jpeg2000 compression,j val=(2215+i*2^0)*10^-1, i=0..1023 (#bits=10)

Description: inv scaling for packing (old format)

Docs derived from https://www.cpc.ncep.noaa.gov/products/wesley/wgrib2/scaling.html