Addressed most of Kerry's TODO list. Details are below.

Also made some fixes to qb4st - need Rob to take a look at them to see if they are correct.

====== DONE =====

Removed :pixelPerDegree (fixes 1, 4)

Changed crs URI (fixes 2)

Created a qb4st:CRS object for the DGGS (fixes 12)

Checked qb4st for relevant updates - added CRS labels (Fixes 3)

Changed union to separate :dataImageValue and :dataPixelValue properties (fixes 6)

Removed prov:wasDerivedFrom :AGDC (fixes 20)

Changed prov:wasAssociatedWith :GeoscienceAustralia to prov:wasAttributedTo (fixes 21)

Changed names of grid squares from arbitrary "s1" and "s2" to "R000" and "R0000" (fixes 27)

===== NOT DONE COMMENTS ====

9. "QB4ST [qb4st] does not currently have a term for that, but it may in the future."

It still does not

15. Example 8 led:etmBand led:latMin led:longMax led:value are not in ANU-LED-example.ttl. They need to be (and they nee to be explained). LEd uses "bounds" instead. Also the example seems to be outofdate wrt QB4st in the ontology? Also it uses led:Pixel -- shouldn't it be a GridSquare (tile?).

For Sam

16. SSN might well change its name and prefix shortly.

For Kerry

17. Several changes to ssn in the area of example 9. Platform, observes, onPlatform, maybe hasMeasurementCapability. Should add ssn:accuracy.

For Kerry

22. Example 13: check ssn:Property for change

For Kerry

24. Example 14. timedomain is a time:Interval that appears to be very short (begins and ends at "2001-10-26T21:32:52) . However, it is normal for satellite imagery to instead be given a (real) time period due to the time taken to travel over the surface. I like that the particular :grid Square has only a time instant, but surely the timedomain could be longer?

I haven't acted on this yet. Sounds like this criticism would also apply to the :time dimension values on the actual data. The way it is now is pleasantly simple and I am reluctant to change it.