WMS outputs from OSMC (and other in-situ datasets)

[karlmsmith]

Reported by steven.c.hankin on 22 Apr 2008 23:03 UTC We need to be able to generate WMS outputs from the OSMC. This is not yet an urgent matter. But it is a good idea to have our plans for it a bit firmer, so that we can discuss it with Derrick and Ted if questions about the GIS compatibility of OSMC arise. The issues raised through exploring this task should wherever possible be handled in a manner that generalizes to other in-situ LAS servers (like the carbon server).

Jing: a key question: Is there a general approach for how one should handle relational constraints in WMS? For example, how would one request a WMS image map showing population density constrained to include only individuals with more than 15 years of education (where 15 is a parameter that can be specified in the request). It might be worth visiting sites like 52North (http://52north.org/) to see if there is some general Java code for WMS for in-situ queries that we might borrow.

IFF there is a general way to encode constraints in WMS:[[BR]] Then it seems like LAS should provide one WMS layer per configured variable. The LAS time and depth constraints could be handled as WMS constraints. Similarly the LAS custom Data Constraints (e.g. color_by = country, age, or platformType) will become WMS constraints.

IFF there is NOT a general way to encode constraints in WMS:[[BR]] Then we'll need to decide upon ad hoc layer definitions. For example, each configured variable in LAS might generate 3 layers -- one each for the last day, last week, last month. Or whatever seems appropriate case by case ...

Note that we will also want to add an LAS transparency Option to the OSMC (and other in-situ) maps.

Migrated-From: http://dunkel.pmel.noaa.gov/trac/las/ticket/504

[karlmsmith]

Comment by steven.c.hankin on 22 Apr 2008 23:07 UTC By the way -- in discussions of WFS I have heard mention of an "OGC filter specification". This is a generalized means to convey relational constraints. It might be applicable to WMS, too. However it is regarded (I think) as terribly complicated, which maybe dooms it never to get widely used.

[karlmsmith]

Comment by jing.y.li on 23 Apr 2008 18:53 UTC There is no such general way for relational constraints in WMS. Roughly speaking, WMS has two sets of parameters, standard parameters (BBOX, time dimension, elevation dimension, etc) and vendor-specific parameters. I think these constraints should go to the vendor-sepcific parameters.

The WMS sepcification for vendor-specific parameters is very short. If a WMS server includes vendor-specific parameters, it has to advertise them in the Capabilities XML and define them within an internal DTD of that XML document. A WMS server shall produce a valid result even if vendor-specific parameters are missing or malformed (i.e., the server shall supply default values for these parameters).

[karlmsmith]

Modified by @kevin-obrien on 5 Jan 2011 23:54 UTC