Hi Dawn, After a few moment's reflection I realized that I may have given you a bum steer with the advice below. The fact that the @mod regridding uses an unweighted average makes it relatively insensitive to having your time axis cells represented with great precision -- though it may be sensitive if your axis uses the 1st of the month (rather than mid-month) as the locations of the time axis points. (Some time axes do this.) To diagnose the behavior that you are seeing have a look at the @MODNGD results let v_clm = v[gt=month_reg@MODNGD] which will tell you how many time points from your native axis are used in the modulo calculation onto the climatological axis. See anything peculiar? - Steve ====================================== Hi Dawn, You have chosen to regrid to axis month_reg instead of month_irreg. This means that the calendar you are using consists of 12 equally-sized months, instead of a true Gregorian climatological calendar. That's the heart of your problem, probably. To ensure that the climatological regridding is completely accurate you have to make sure that the calendar axis of the original data exactly matches the target climatological axis. From the Ferret Users' Guide: Unlike other transformations and regridding, modulo regridding is performed as an unweighted average: each non-missing source point contributes 100% of its weight to the destination grid box within which it falls. If the source and destination axes are not properly aligned this can lead to apparent shifts in the data. For example, if a monthly time series has data points at the first of each month and a climatological axis is defined at midmonths, then unweighted modulo averaging will lead to an apparent 1/2-month shift. To avoid situations of this type simply regrid to the climatological axis using linear interpolation prior to the modulo regridding.(I believe that the above unweighted behavior was selected specifically to address the messy aspects of climatological regridding, where the existence of leap years compromises the "modulo" nature of the calendar axis.) You can check the time axis of "atl.nc" using the SHOW GRID (or SHOW AXIS) command: SHOW GRID/T v ==> Make sure that TBOXLO of each month is the 1st of the month at hour 00:00 If you try the SHOW AXIS command on the month_irreg axis you'll see how the leap-year length of February complicates climatological calendar. The length of Feb is not an integer number of days, so all months after February appear to have a 5 hour and 49 minute offset. But the TBOX values show that the length of each month is correct. yes? use climatological_axes - Steve ============================== ferret ocean wrote:
-- -- Steve Hankin, NOAA/PMEL -- Steven.C.Hankin@xxxxxxxx 7600 Sand Point Way NE, Seattle, WA 98115-0070 ph. (206) 526-6080, FAX (206) 526-6744 |