<div dir="ltr"><div><div><div><div>Baptiste,<br><br></div>Thank you for the input. <br></div><div>Let me see if I get the intent correctly.. <br><br>The intent here is to describe the case where you are observing a specific feature or aspect of some object?<br></div></div> eg: The atmosphere of Jupiter (or the 'Red Spot')<br></div> eg: Erebor Mons on Saturn's moon Titan<br><br><div><div><div><div>I'm looking at this to compare with what you've described<br></div> <a href="http://planetarynames.wr.usgs.gov/SearchResults?target=TITAN&featureType=Mons,%20montes">http://planetarynames.wr.usgs.gov/SearchResults?target=TITAN&featureType=Mons,%20montes</a><br><br>So, you want the primary Target.name to be the planet/object itself so that searches for the parent object include these refined observations.<br>And, you need to add a secondary set of fields to describe the 'resolved' or more specific feature including:<br></div><div> + its name <br></div><div> + its type (class)<br></div><div> + its location on the parent object<br><br></div><div>BTW: It seems to me that it would be useful to include the BaseTarget.position to provide the location of the parent object (Titan) at the time of the observation in a standard reference frame.<br><br></div><div>I would suggest that if the 'objectClass' and 'elementClass' values are to be defined by a Planetary nomenclature list, that the name of the class should be more specific/linked to that domain (PlanetaryTarget).. not sure if that works well for your 'sunspot' example.<br><br></div><div>Are you thinking that this could/should be added to the DatasetDM? I don't see any real problem with that, but if not, it could easily be defined in EPNCore as an extension of DatasetDM.BaseTarget and fit in the overall model domain just fine.<br><br></div><div>Mark<br><br></div></div></div></div>