SMAP Level 3 Sea Surface Salinity

SMAP Level 3 Sea Surface Salinity

Postby dale.robinson » Fri Nov 18, 2016 11:43 am

Hi,

Re:
1. RSS SMAP Level 3 Sea Surface Salinity Standard Mapped Image 8-Day Running Mean V2.0
http://thredds.jpl.nasa.gov/thredds/cat ... INGMEAN_V2

2. RSS SMAP Level 3 Sea Surface Salinity Standard Mapped Image Monthly V2.0
http://thredds.jpl.nasa.gov/thredds/cat ... MONTHLY_V2

According to the titles in the THREDDS catalog, these are aggregated datasets. Looking at the metadata I see that there are three coordinate variables defined (lat, lon, time). Also, for the sss_smap variable I see that the coordinates attribute says "time lat lon", but that the actual dimensions associated with sss_smap are just lat and lon [e.g. float sss_smap(lat=720, lon=1440)].

Maybe I am wrong about this but, to aggregate these files don't you need to include time in the sss_smap dimensions?
e.g. a) float sss_smap(time =1, lat=720, lon=1440); for a single time step
b) float sss_smap(time =20, lat=720, lon=1440); for a thredds temporal subset that includes 20 time steps

Thanks.

some metadata from netCDF file below:
float sss_smap(lat=720, lon=1440);
:long_name = "SMAP sea surface salinity";
:standard_name = "sea_surface_salinity";
:units = "1e-3";
:valid_min = 0.0f; // float
:valid_max = 45.0f; // float
:_FillValue = -9999.0f; // float
:scale_factor = 1.0f; // float
:add_offset = 0.0f; // float
:coordinates = "time lat lon ";
:_ChunkSizes = 720, 1440; // int
dale.robinson
 
Posts: 6
Joined: Fri Nov 18, 2016 8:11 am

Re: SMAP Level 3 Sea Surface Salinity

Postby vtsontos » Fri Nov 18, 2016 5:23 pm

Thanks Dale.
that is a good catch, and you are correct.
I have confirmed that temporal aggregation is an issue right now (11/18/2016) for THREDDS subsetting queries involving multiple days of data.
Spatial subsetting is not impacted and works fine.
We will figure out a work around using NCML, and will let you know once the fix is implemented.
Regards,
Vardis
vtsontos
 
Posts: 29
Joined: Wed Sep 03, 2014 10:43 am

Re: SMAP Level 3 Sea Surface Salinity

Postby dale.robinson » Mon Mar 06, 2017 3:10 pm

Hi Vardis,

I'm just checking in to see if this issue with SMAP temporal aggregation was resolved.

Thanks,
Dale
dale.robinson
 
Posts: 6
Joined: Fri Nov 18, 2016 8:11 am

Re: SMAP Level 3 Sea Surface Salinity

Postby vtsontos » Mon Mar 06, 2017 3:46 pm

Hi Dale.
yes, an NCML workaround to the metadata time aggregation issue that you identified has been implemented. So you should be able to subset these the L3 SMAP-SSS 8day-running and Monthly data via THREDDS in both space and time normally now at:

http://thredds.jpl.nasa.gov/thredds/cat ... talog.html

I just tried the following NCSS request, and it returned the time aggregated spatial SSS subset for the 8-day running product consistent with the query inputs.

http://thredds.jpl.nasa.gov/thredds/ncs ... atLon=true

But let me know if you encounter any issues.

Thanks again for pointing out the original issue.

Regards,
Vardis
vtsontos
 
Posts: 29
Joined: Wed Sep 03, 2014 10:43 am

Re: SMAP Level 3 Sea Surface Salinity

Postby dale.robinson » Wed Mar 15, 2017 10:27 am

Thanks Vardis,

I have one more question. Do you have plans to aggregate this dataset? By aggregate I mean grouping multiple data files into a single, logical dataset, e.g. group monthly files into yearly aggregates or group 8-day running average files into monthly aggregates?

It would help us out a lot if that happened, so I'm just asking and crossing my fingers.

Thanks,
Dale
dale.robinson
 
Posts: 6
Joined: Fri Nov 18, 2016 8:11 am

Re: SMAP Level 3 Sea Surface Salinity

Postby vtsontos » Wed Mar 15, 2017 5:06 pm

Hi Dale,
Thanks for your follow-up inquiry regarding higher order aggregation of the Smap-SSS L3 products.
I am not sure I totally understand the request. Do you mean providing a time averaged product based on some kind of temporal aggregation interval, say monthly for the 8day running data and annually for the monthly dataset?
If so, then I am not sure if the need for a monthly mean aggregation since there is a native monthly product already available. Also technically, such compositing at non-native ly available intervals really needs to be done based on the L2 data which accounts for potential different sample sizes for space-time bins. That is not a trivial task for us to undertake, but given a strong justification and clearer understanding of the need you expressed, then it is something we can convey to the data provider. If they produce the new product, then we would archive and distribute it.

Hope this helps.
Best.
Vardis
vtsontos
 
Posts: 29
Joined: Wed Sep 03, 2014 10:43 am

Re: SMAP Level 3 Sea Surface Salinity

Postby dale.robinson » Thu Mar 16, 2017 11:03 am

Hi Vardis,

Sorry if my question wasn't stated clearly. We would be interested in all of the present data files for SMAP being aggregated into a single, logical dataset (not monthly or yearly in the examples in my last email). This is done with the aggregation NcML element, so there are no additional data files being generated. It is just an additional way of indexing the data.

I have included the links below on doing this in TDS and Hyrax.

http://www.unidata.ucar.edu/software/th ... ation.html
http://docs.opendap.org/index.php/BES_- ... _Tutorials

The PODAAC threads seems to be down now, when it comes up I can point you to some datasets on the PODAAC that have been aggregated.

Thanks.

-Dale
dale.robinson
 
Posts: 6
Joined: Fri Nov 18, 2016 8:11 am


Return to Data Access

cron