Documentation Cluster - February 2013 telecon

2) Old Action Items

A. Ted will add a paragraph on the wiki about the sensor-instrument and channel-band issue (August 2012)

·         Ted – talked to Phil Jones at NCDC about how use term instrument and sensor and sensor vs. detectors

·         Ed – if we add detector does that mean we have a 4-level structure

·         Aleksandar – from an engineering perspective

o   A Platforms have 1 or more instruments

o   A instrument can have 1 or more sensors

o   a  sensor can only have 1 detectors – which is an electrical charge device)

·         in laymen terms – people confuse instrument and sensor

·         Ted’s new text on Standard Names for Satellite Observations (http://wiki.esipfed.org/index.php/Standard_Names_For_Satellite_Observations)

·         NASA needs the 2 separate terms (platform and instrument)

·         Steven – should we define what a sensor ‘does’ – a sensor takes one or more measurements

·         This typology is hardware centric

·         These terms are defined in Sensor ML

·         Walt – an example is an instrument with 9 cameras – each camera is a sensor

ACTION – repeat of C – Aleksandar will add sensor ML definitions

·         Sensor ML is also used by hydrologists, weather and the ESA (European Space Agency)

·         ESA – use sensor ML for lineage information

B. Ted up upload his ISO slides to the wiki & commons (Dec 2012)

·         DONE

C. Aleksandar will add the sensorML definitions for sensor, instrument, channel, and band (from August 2012)

·         Aleksandar will do this (he promised)

1)      NASA ESDSWG group HDF5/CF conventions (Ed)

·         Talked to Charley Zender who is leading the HDF5 convensions group (documenting conversion handling

·         https://wiki.earthdata.nasa.gov/display/ESDSWG/HDF5+Conventions+Working+Group

o   To access this website you need permission

o   Go to https://urs.eosdis.nasa.gov/ and register

o   Then https://wiki.earthdata.nasa.gov/ and contact Earthdata Support ([email protected]) 

o   OR email Ross Bagwell at [email protected]

·         There are 6 outcomes for the working group

1.      Document requirements for interoperable handling (e.g., HDF dimension scales) (interoperability)

2.      Identify which CF-Conventions to require in future NASA HDF datasets (Climate Forecast)

3.      Outline process to check CF-compliance of NASA HDF datasets (Climate Forecast Compliance)

4.      Outline process to integrate ISO Metadata content into public NASA HDF Datasets (ISO)

5.      Draft candidate examples of groups and hierarchical metadata conventions for CF (Examples)

6.      The Attribute Conventions for Data Discovery and HDF

o   Have only had one meeting

·         Ted – they are trying to put groups in the metadata for NetCDF, HDF5, and ISO

·         Ed – the goal is the increase interoperability in organizing satellite data in HDF5 or netCDF4 with CF

·         The problem is how define coordinate systems for field or regard CF only has 1 coordinate need more (ex. 9)

o   Walt – VR ML – has ref geometry that scale – they have already dealt with this issue

·         VRML is a graphics language

·         Virtual relative – relative objects that scale

·         Ted wants examples described in wiki

ACTION: Ed will add GRIS example

ACTION: Walt – add VR and more HDF4 examples

·         To link on wiki to Documentation add [[Category: Documentation Cluster]] on your page

·         Walt has lots of things that doesn’t work when checking CF complains on the website

o   Aleksandar suggest he post questions to the ESIP-CF list serve and they/we will try to answer it

·         The ESIP wiki is available to a broader audience than the NASA earthdata wiki

o   Want to add things to the ESIP wiki and like on the NASA wiki

3) Temporal Data (Walt)

·         TAI – not UTC

·         Is there a conversion to make TAI CF complant

·         How can TAI be represented OR do you have to change tot UTC?

·         Ted – has not heard of it

·         Aleksandar – it is related to the atomic clock – counts seconds – BUT CF does not deal with it

·         Because CF was a modeling community – they were not structures for things like TAI

ACTION – Walt will add TAI time issue to the ESIP wiki as a CF problem