Notice of Stage II/IV Grid Changes

The following changes were made to the NCEP Stage II and Stage IV precipitation analyses on 11 May 2004:

Stage II:

Grid number changed from 240 to 255. The 240 is the national HRAP grid, and while the Stage II is on the actual HRAP grid, the Grid Definition Section we provide in the grib files uses grid co-ordinates that are 1/2 grid off-set from what's defined in the Grib manual, because the numbers in the Grib manual are provided by hydrologists, who define the co-ordinates of a grid box as the 'lower-left' corner of the grid box, while we (meteorologists) define it as the center of the grid box (see http://wwwt.emc.ncep.noaa.gov/mmb/ylin/pcpanl/QandA/#GRIDINFO for further details). It is better to name our grid '255' for 'user-defined', so people using the standard GRIB library/GRIB manuals would not be using the hydrologists' definition of grid co-ordinations.

Stage IV:

Grid number changed from 240 to 255 (see explanation for Stage II above). Longitude of (1,1) point corrected. The (1,1) point should be (23.117N, 119.023W), but was mistakenly defined as (23.117N, 119.017W) from 1 Jan 2002-10 May 2004 (great circle distance between the two is about 0.615km).

Link to Sample Analyses and Plots:

Provides sample before-and-after-change sample Stage 4 and Stage 2 (ST2ml) analyses and plots. Note that the old and new Stage II plots are identical (only the grid number is changed); the old and new Stage IV plots are slightly different, but since the shift is less than 1km, the difference is imperceptible.

Note: during 1 Jan 2002 - 12Z 18 Feb 2004, the Stage II and Stage IV plots in the archive and in the side-by-side daily loops appear to be on slightly different domains, as shown by the red frames in this example. This was actually due to different input parameters used by Stage II and Stage IV plots:

1 -1 39.65 97.30 1170 900 0 99999.99 ST2 Multi-sensor

1 -1 39.60 97.00 1150 920 0 99999.99 Stage IV (III MOS)

The 3rd-6th parameters in the two input lines above are the latitude/longitude of the center of the plot, number of grid points in the x and y directions. These differences led to the apparent grid discrepancy in the Stage II and Stage IV grid domains (the actual difference caused by the incorrectly specified Stage IV grid was too small to be visually perceptible).

Starting 13Z 18 Feb 2004, the Stage IV plotting line above was changed to

1 -1 39.65 97.30 1170 900 0 99999.99 Stage IV (III MOS)

to be consistent with the Stage II plots.