Section 4 - Code Table 15 : Type of spatial processing used to arrive at given data value from the source data

Code Meaning
0 Data is calculated directly from the source grid with no interpolation
1 Bilinear interpolation using the 4 source grid grid-point values surrounding the nominal grid-point
2 Bicubic interpolation using the 4 source grid grid-point values surrounding the nominal grid-point
3 Using the value from the source grid grid-point which is nearest to the nominal grid-point
4 Budget interpolation using the 4 source grid grid-point values surrounding the nominal grid-point
5 Spectral interpolation using the 4 source grid grid-point values surrounding the nominal grid-point
6 Neighbor-budget interpolation using the 4 source grid grid-point values surrounding the nominal grid-point
255 Missing
Notes:

( 1) This method assumes that each field really represents box averages/maxima/minima where each box extends halfway to its neighbouring grid point in each direction to represent averages/maxima/minima of values from the source grid with no interpolation.

( 2) Budget interpolation means a low-order interpolation method that quasi-conserves area averages. It would be appropriate for interpolating budget fields such as precipitation. This method assumes that the field really represents box averages/maxima/minima where each box extends halfway to its neighbouring grid point in each direction. The method actually averages bilinearly interpolated values in a square array of points distributed within each output grid box.

( 3) Performs a budget interpolation at the grid point nearest to the nominal grid point.