|
Hi Seontae, The average that's computed using world coordinates (dates) is done as a weighted average, starting and ending exactly at the dates and times that are specified, and weighted by the parts of the grid boxes that are included in that time interval. The one where you give index values, L=636:638@ave, is the simple arithmetic mean of the three values. In the Users Guide, all of the documentation about transformations such as @AVE say, "See the General Information on transformations for important details about this transformation". The link points to this section: ../../documentation/users-guide/variables-xpressions/XPRESSIONS#GeneralInfo Here the relevant text is down a few paragraphs, "If integration or averaging limits are given as world coordinates, the grid boxes at the edges of the region specified are weighted according to the fraction of grid box that actually lies within the specified region. If the transformation limits are given as subscripts, the full box size of each grid point along the axis is used—including the first and last subscript given. The region information that is listed with the output reflects this." So when the coordinates are in the middle of each grid box, averaging 3 grid boxes is a computation of (half_box2* first_value) + (whole_box2* second_value) + (half_box3* third_value), all divided by the sum of those partial box sizes. You can see the difference in the time range listed with each output nino3[t=15-dec-2002:15-feb-2003@ave] uses this range: TIME : 15-DEC-2002 00:00 to 15-FEB-2003 00:00 (averaged) list nino3[l=636:638@ave] uses the time range to the edges of the grid boxs, without weighting. TIME : 30-NOV-2002 04:43 to 01-MAR-2003 12:11 (averaged) On 10/4/2010 3:56 PM, Seontae Kim wrote:
|