The time flag of hourly output from CESM seems strange

[xiaoxiaokuishu]

Hi, all.
I make a simulation of FHIST with hourly output (2013-2014), but I when cdo sinfo to check the data, the timestamp

2013-01-01 00:00:00, 2013-01-01 01:00:00 ........
2014-12-31 23:00:00 2015-01-01 00:00:00

Here, the 2015-01-01 00:00:00 is the with the wrong timestamp as my simulation is end at 2014.

 

[mlevy]

CESM output typically sets the time variable to the end of the averaging interval, so if your last time step takes the model from 2014-12-31 23:00:00 to 2015-01-01 0:00:00 then the time associated with that output will be 2015-01-01 0:00:00.

 

[xiaoxiaokuishu]

CESM output typically sets the time variable to the end of the averaging interval, so if your last time step takes the model from 2014-12-31 23:00:00 to 2015-01-01 0:00:00 then the time associated with that output will be 2015-01-01 0:00:00.

oh, thanks for the response, but for the postprocess, for example cdo, there may be sth wrong since it take the 2014-12-31-23:00 to 2015. Besides, the 2013-01-01 00:00:00 mean the time between 2012-12-31 23:00:00 to 2013-01-01 0:00:00? Is there any proper tools to deal with this matter?

 

[strandwg]

oh, thanks for the response, but for the postprocess, for example cdo, there may be sth wrong since it take the 2014-12-31-23:00 to 2015. Besides, the 2013-01-01 00:00:00 mean the time between 2012-12-31 23:00:00 to 2013-01-01 0:00:00? Is there any proper tools to deal with this matter?

Please take a look at the "time_bnds" variable. For time-averaged fields, it contains the time endpoints of the averaging period.