Eli,
The way CF is defined, the bounds are the recommended way to capture
that information whether or not the interval is constant. There are
enough exceptions (missing data point, etc) and variations (a few time
examples - time is at the end of the interval, intervals overlap, there
are gaps between intervals) that it seems easier to me to use the bounds
than to come up with an entire vocabulary to capture this information.
If you would like to provide guidance to your users in a more controlled
fashion than comments, you might be able to use the ACDD
<http://wiki.esipfed.org/index.php/Attribute_Convention_for_Data_Discovery_1-3>
time_coverage_resolution global attribute. (Not much help if you have
multiple variables in a file with a different interval for each.)
You are also welcome to propose attributes that could capture the
information needed to represent acquisition intervals in space and time.
Grace and peace,
Jim
On 8/13/18 12:38 PM, Ateljevich, Eli@DWR wrote:
Jim et al,
Thanks and sorry for the long delay.
For what it is worth, yours is a fine generic example. David’s
“interval” comment seems to fit the bill. The question is then whether
time_bounds is even necessary. It seems like up to the difference
between (n) and (n-1), “time_bounds” has a reasonable default, which
is the space between “time”.
Jim, the way you wrote it without “interval” illustrates the work I
have to do if the data are not marked as regularly, I would have to
interrogate every single time cell to figure it out, and even then it
isn’t clear if it is luck -- the next two hours could be different or
a piece of missing data could add ambiguity. Perhaps more importantly
for a metadata standard, it seems to be an omission of an important
piece of information about the collection and processing program. To
be honest this seems like something that should be in there even for
instantaneous data.
Thanks so much to both of you,
Eli
*From:* CF-metadata [mailto:[email protected]] *On
Behalf Of *David Hassell
*Sent:* Tuesday, July 24, 2018 8:46 AM
*To:* Jim Biard <[email protected]>
*Cc:* CF Metadata <[email protected]>
*Subject:* Re: [CF-metadata] metadata for regularly spaced data and
averages
Hello Eli,
Would the standardised "interval" cell methods keyword be of use
(http://cfconventions.org/Data/cf-conventions/cf-conventions-1.7/cf-conventions.html#recording-spacing-original-data
<https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fcfconventions.org%2FData%2Fcf-conventions%2Fcf-conventions-1.7%2Fcf-conventions.html%23recording-spacing-original-data&data=02%7C01%7C%7C94921b212562405e2f8608d5f17cab36%7Cb71d56524b834257afcd7fd177884564%7C0%7C0%7C636680440115696558&sdata=kyGARYHxVPwIN6ixZiBDA0RLfl2A2RnXgW3ouc6Kda0%3D&reserved=0>):
float data(time);
data:long_name = "The data";
data:cell_methods = "time: mean (interval: 1 hour)" ;
data:standard_name = "?????";
data:units = "????";
All the best,
David
On 23 July 2018 at 21:56, Jim Biard <[email protected]
<mailto:[email protected]>> wrote:
Eli,
I'm not quite sure what you are looking for here. There are no
other mechanisms in CF right now for describing sampling interval.
Perhaps it would help if you described in greater detail the
inputs to your process and the outputs from your process that you
wish to store in a netCDF file. Here's what I am getting from what
you wrote in your first message, with arbitrary numbers selected
to make it concrete.
Inputs:
Each hour:
* Instrument takes 100 samples
* Instrument reports the time and the average of the 100 samples
Outputs for a netCDF file:
* One month of hourly average values
* The time each hourly average was reported
* The length of the acquisition interval
If this is what you have it is straightforward to store in netCDF
using CF. Using CDL, the file structure would be:
netcdf file: sample.nc
<https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fsample.nc&data=02%7C01%7C%7C94921b212562405e2f8608d5f17cab36%7Cb71d56524b834257afcd7fd177884564%7C0%7C0%7C636680440115696558&sdata=xMOLOSOoq0xxjINQ0JW76ORPZCrQCZWeOhWpLCytit0%3D&reserved=0>
{
dimensions:
time = 100;
bnds = 2;
variables:
float time(time);
time:long_name = "reporting time";
time:units = "seconds since 1970-01-01 00:00:00";
time:standard_name = "time";
time:axis = "T";
time:calendar = "gregorian";
time:bounds = "time_bounds";
float time_bounds(time, bnds);
float data(time);
data:long_name = "The data";
data:cell_methods = "time: mean (any comment that
you find helpful)";
data:standard_name = "?????";
data:units = "????";
data:
time = 3600.0, 7200.0, ... ;
time_bounds = 0.0, 3600.0, 3600.0, 7200.0, 7200.0,
10800.0, ... ;
}
I hope this helps frame the question and/or provides a bit of the
answer.
Grace and peace,
Jim
On 7/20/18 5:42 PM, Ateljevich, Eli@DWR wrote:
I have data that are recorded hourly after sampling the
previous hour. The original sampling rate is high, but in some
cases not provided so it doesn’t tend to make it into the
metadata.
I can represent the hourly average with cell_methods and
time_bnds, but it is painful for two reasons:
1. It costs some storage overhead in simple cases because the
bounds are obvious.
2. The fact that it is regularly sampled is part of the
station design. Here, I have to infer two things: (regular
reporting interval and relative position of timestamp from
time_bnds). The first one in particular is not a comment to me.
I can certainly see the lack of generality creeping looming,
but this is such an important special case. Is there no common
shorthand or possibly something I could at least do in
**addition** to the cell_bounds. Am I missing a long
conversation on this?
Obviously my question extends to any regular sampling
geometries, I just picked a simple one in 1D.
Thanks,
Eli
_______________________________________________
CF-metadata mailing list
[email protected] <mailto:[email protected]>
http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
<https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fmailman.cgd.ucar.edu%2Fmailman%2Flistinfo%2Fcf-metadata&data=02%7C01%7C%7C94921b212562405e2f8608d5f17cab36%7Cb71d56524b834257afcd7fd177884564%7C0%7C0%7C636680440115696558&sdata=zx%2BWBcr24WJBASAKnZ4FmcKR0AgeG2JATBUIlXYgzD4%3D&reserved=0>
--
Visit us on
*Facebook
*<https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.cicsnc.org%2F&data=02%7C01%7C%7C94921b212562405e2f8608d5f17cab36%7Cb71d56524b834257afcd7fd177884564%7C0%7C0%7C636680440115696558&sdata=%2B5kd1yhZ92qNZ%2FJrxbCWZtSwWGBoXn1tI3tzlBjn7JA%3D&reserved=0>
*Jim Biard
Research Scholar
*Cooperative Institute for Climate and Satellites NC *
*North Carolina State University *
*NOAA National Centers for Environmental Information *
/formerly NOAA’s National Climatic Data Center/
151 Patton Ave, Asheville, NC 28801
e: [email protected]_
o: +1 828 271 4900
/Connect with us on Facebook for _climate_ and _ocean and
geophysics_ information, and follow us on Twitter at
_@NOAANCEIclimate_ and _@NOAANCEIocngeo_.
/<https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.cicsnc.org%2F&data=02%7C01%7C%7C94921b212562405e2f8608d5f17cab36%7Cb71d56524b834257afcd7fd177884564%7C0%7C0%7C636680440115696558&sdata=%2B5kd1yhZ92qNZ%2FJrxbCWZtSwWGBoXn1tI3tzlBjn7JA%3D&reserved=0>*
<https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.cicsnc.org%2F&data=02%7C01%7C%7C94921b212562405e2f8608d5f17cab36%7Cb71d56524b834257afcd7fd177884564%7C0%7C0%7C636680440115696558&sdata=%2B5kd1yhZ92qNZ%2FJrxbCWZtSwWGBoXn1tI3tzlBjn7JA%3D&reserved=0>
_______________________________________________
CF-metadata mailing list
[email protected]_
_http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata_<https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.cicsnc.org%2F&data=02%7C01%7C%7C94921b212562405e2f8608d5f17cab36%7Cb71d56524b834257afcd7fd177884564%7C0%7C0%7C636680440115696558&sdata=%2B5kd1yhZ92qNZ%2FJrxbCWZtSwWGBoXn1tI3tzlBjn7JA%3D&reserved=0>
--
<https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.cicsnc.org%2F&data=02%7C01%7C%7C94921b212562405e2f8608d5f17cab36%7Cb71d56524b834257afcd7fd177884564%7C0%7C0%7C636680440115696558&sdata=%2B5kd1yhZ92qNZ%2FJrxbCWZtSwWGBoXn1tI3tzlBjn7JA%3D&reserved=0>
David Hassell
National Centre for Atmospheric Science
Department of Meteorology, University of Reading,
Earley Gate, PO Box 243, Reading RG6 6BB
Tel: +44 118 3785183
_http://www.met.reading.ac.uk/_<https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.cicsnc.org%2F&data=02%7C01%7C%7C94921b212562405e2f8608d5f17cab36%7Cb71d56524b834257afcd7fd177884564%7C0%7C0%7C636680440115696558&sdata=%2B5kd1yhZ92qNZ%2FJrxbCWZtSwWGBoXn1tI3tzlBjn7JA%3D&reserved=0>
--
CICS-NC <http://www.cicsnc.org/> Visit us on
Facebook <http://www.facebook.com/cicsnc> *Jim Biard*
*Research Scholar*
Cooperative Institute for Climate and Satellites NC <http://cicsnc.org/>
North Carolina State University <http://ncsu.edu/>
NOAA National Centers for Environmental Information <http://ncdc.noaa.gov/>
/formerly NOAA’s National Climatic Data Center/
151 Patton Ave, Asheville, NC 28801
e: [email protected] <mailto:[email protected]>
o: +1 828 271 4900
/Connect with us on Facebook for climate
<https://www.facebook.com/NOAANCEIclimate> and ocean and geophysics
<https://www.facebook.com/NOAANCEIoceangeo> information, and follow us
on Twitter at @NOAANCEIclimate <https://twitter.com/NOAANCEIclimate> and
@NOAANCEIocngeo <https://twitter.com/NOAANCEIocngeo>. /
_______________________________________________
CF-metadata mailing list
[email protected]
http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
