Hi Tony,The problem you face is that the days1900 function uses the default calendar, not the NOLEAP one.
The climatological_axes dataset contains a monthly NOLEAP climatological axis called MONTH_NOLEAP, so you can just use that.
yes? use climatological_axes*** NOTE: regarding /home/users/tmap/ferret/linux/go/climatological_axes.cdf ... *** NOTE: Climatological axes SEASONAL_REG, MONTH_REG, MONTH_IRREG, MONTH_GREGORIAN, MONTH_NOLEAP, MONTH_360_DAY defined
yes? list t[gt=MONTH_NOLEAP], tboxlo[gt=MONTH_NOLEAP], tboxhi[gt=MONTH_NOLEAP]
TIME: 01-JAN 00:00 to 01-JAN 00:00 NOLEAP
Column 1: T is T (axis MONTH_NOLEAP)
Column 2: TBOXLO is TBOXLO (axis MONTH_NOLEAP)
Column 3: TBOXHI is TBOXHI (axis MONTH_NOLEAP)
T TBOXLO TBOXHI
16-JAN 12 / 1: 15.5 0.0 31.0
15-FEB 00 / 2: 45.0 31.0 59.0
16-MAR 12 / 3: 74.5 59.0 90.0
16-APR 00 / 4: 105.0 90.0 120.0
16-MAY 12 / 5: 135.5 120.0 151.0
16-JUN 00 / 6: 166.0 151.0 181.0
16-JUL 12 / 7: 196.5 181.0 212.0
16-AUG 12 / 8: 227.5 212.0 243.0
16-SEP 00 / 9: 258.0 243.0 273.0
16-OCT 12 / 10: 288.5 273.0 304.0
16-NOV 00 / 11: 319.0 304.0 334.0
16-DEC 12 / 12: 349.5 334.0 365.0
On 7/27/2011 6:45 PM, Jian Ma wrote:
Hi All, I am using the function days1900 to calculate monthly mean from NOLEAP daily mean. There is an error in time axis definition. Days1900 are giving Julian days from 1900. If I use the following definition, the time is wrong, due to different day count between the calendars. DEFINE AXIS/UNITS=days/T0=1-jan-1900/EDGES/cal=noleap truemonth = DAYS1900(year,month,1) Anyone has a good method to calculate NOLEAP days from 1900? Or do I have to do a NOLEAP -> Julian transformation Many thanks, Tony