Aside from avoiding the use of the word "time" for anything that is not "days since" or equivalent, I would do it as you described. Instead of time I would use "leadtime" or "L", particularly so that I could define "time" (reftime + leadtime) (with standard_name "time") and include that as well in the variable list.
On Tue, May 17, 2011 at 3:02 PM, Karl Taylor <[email protected]> wrote: > Hi Benno, > > This isn't really a comment on your email, but a question (only somewhat > related) occurred to me. When you store multiple trajectories (or > forecasts), each started from a difference reference, but sampled in the > same way (at equal intervals), how should this be done? Suppose, for > example, you want to store ozone data from 10 freely moving balloons, with > samples taken hourly following release (for 24 hours), but each balloon > released at a different time. Would you do this as follows? > > dimensions: > time = 24 ; > ref_time=10 > variables: > float O3(time,ref_time) ; > O3:long_name = "mole_fraction_of_ozone_in_air" ; > O3:units = "1e-9" ; > O3:coordinates = "lon lat z" ; > double time(time) > time:standard_name = "elapsed_time??? or relative_time???" > time:long_name = "elapsed time since the beginning of the trajectory" > time:units = "hr" > double ref_time(ref_time) ; > ref_time:standard_name = "reference_time???" ; > ref_time:long_name = "time when balloon was released and the reference > for elapsed_time (relative_time)" > ref_time:units = "days since 1970-01-01 00:00:00" ; > float lon(time,ref_time) ; > lon:standard_name = "longitude" ; > lon:units = "degrees_east" ; > float lat(time,ref_time) ; > lat:standard_name = "latitude" ; > lat:units = "degrees_north" ; > float z(time,ref_time) ; > z:standard_name = "height_above_reference_ellipsoid" ; > z:units = "km" ; > z:positive = "up" ; > > Note the possible standard_names (I think, suggested by others). > > Best regards, > Karl > > > > > On 5/17/11 8:46 AM, Benno Blumenthal wrote: > > CF has standard names forecast_reference_time, forecast_period and > time which are interrelated in a particular way. > > I have a trajectory dataset which also has reference_time, period, > and time which are interrelated in the same way, but forecast is not > an appropriate descriptor: the reference_time is the start of the > trajectory, the period is the time relative to the start_time along > the trajectory. > > I am wondering how important "forecast" is in the semantics of these > particular standard_names -- does it really have to be a forecast? > After all, these are the standard names for the time coordinates, e.g. > independent variables, while forecast is a property of the dependent > variables, i.e. how they were computed. > > Do we need more general names? Am I taking the current names too literally? > > > > _______________________________________________ > CF-metadata mailing list > [email protected] > http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata > > -- Dr. M. Benno Blumenthal [email protected] International Research Institute for climate and society The Earth Institute at Columbia University Lamont Campus, Palisades NY 10964-8000 (845) 680-4450
_______________________________________________ CF-metadata mailing list [email protected] http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
