Opened 11 years ago

Closed 11 years ago

#284 closed defect (fixed)

gml:origin and gml:offsetVectors are wrong for non-2D coverages

Reported by: Piero Campalani Owned by: Piero Campalani
Priority: major Milestone: 8.4
Component: petascope Version: 8.3
Keywords: origin offsetvector dimensionality Cc: abeccati
Complexity: Medium

Description

In WCS DescribeCoverage and GetCoverage responses, although gml:Envelope and gml:GridEnvelope are correct, the gml:origin and gml:offsetVector elements (RectifiedGridCoverages) have wrong dimensionality.

E.g., for a 3D coverage (eobstest) they are 2D:

<domainSet>
  <RectifiedGrid dimension="3" gml:id="eobstest-grid">
    <limits>
      <GridEnvelope>
        <low>0 0 0</low>
        <high>5 100 231</high>
      </GridEnvelope>
    </limits>
    <axisLabels>t x y</axisLabels>
    <gml:origin>
      <gml:Point gml:id="eobstest-origin" srsName="http://www.opengis.net/def/crs/EPSG/0/4326">
        <gml:pos>25.0 -40.0</gml:pos>
      </gml:Point>
    </gml:origin>
    <gml:offsetVector srsName="http://www.opengis.net/def/crs/EPSG/0/4326">0 0.5</gml:offsetVector>
    <gml:offsetVector srsName="http://www.opengis.net/def/crs/EPSG/0/4326">0.5 0</gml:offsetVector>
  </RectifiedGrid>
</domainSet>

For a 1D coverage it they are not reported:

<domainSet>
  <RectifiedGrid dimension="1" gml:id="OneD-grid">
    <limits>
      <GridEnvelope>
        <low>0</low>
        <high>10</high>
      </GridEnvelope>
    </limits>
    <axisLabels>x</axisLabels>
  </RectifiedGrid>
</domainSet>

Change History (3)

comment:1 by Piero Campalani, 11 years ago

In this case, being origin and offsetVectors mandatory elements in a RectifiedGrid, the proposed solution is to list the proper values with proper dimensionality, while setting a compound CRS name in the form:

http://www.opengis.net/def/crs-compound?
    1=%crs1%&
    2=%crs2%
    ...
    N=%crsN%

comment:2 by abeccati, 11 years ago

Status: newassigned

Then the use of such crs values should clearly be marked as "Experimental" until a corresponding standard is approved. Indeed I see no other way of providing complete and consistent output for 1 and 3+ dimensions.
This is also backward compatible behavior for 2D coverages I think

PS: accept or reassign tickets to yourself when working on them so we know "new" and "reopened" ones are looking for care.

comment:3 by Piero Campalani, 11 years ago

Resolution: fixed
Status: assignedclosed

Ticket addressed in 5bd2f3.

Note: See TracTickets for help on using tickets.