search

cedardevs / onestop

OneStop is a data discovery system being built by CIRES researchers on a grant from the NOAA National Centers for Environmental Information. We welcome contributions from the community!
GNU General Public License v2.0
44 stars 21 forks source link

Bug Report: Point dataset presented incorrectly as very large bounded boxes #1548

Open tomersagi opened 1 year ago

tomersagi commented 1 year ago

Bug Description When searching for datasets in the Mediterranean, we were surprised to find this result declaratively collected from a river estuary in the USA: https://data.noaa.gov/onestop/collections/details/56c9661e-ac7b-4aba-9495-214eb0e2cb5f

The bounding box is shown to cover almost half of the world:

image

But looking at the metadata file, it was obvious that whatever process is scraping this data into OneStop is assigning 0 value to two of the bounding box points when the original dataset is marked with a point:

<gmd:geographicElement>
<gmd:EX_GeographicBoundingBox id="boundingGeographicBoundingBox">
<gmd:westBoundLongitude>
<gco:Decimal>-124.003</gco:Decimal>
</gmd:westBoundLongitude>
<gmd:eastBoundLongitude>
<gco:Decimal>0</gco:Decimal>
</gmd:eastBoundLongitude>
<gmd:southBoundLatitude>
<gco:Decimal>0</gco:Decimal>
</gmd:southBoundLatitude>
<gmd:northBoundLatitude>
<gco:Decimal>46.261667</gco:Decimal>
</gmd:northBoundLatitude>
</gmd:EX_GeographicBoundingBox>
</gmd:geographicElement>

Platform/Environment Windows / Chrome

Severity Severity: HIGH - geographic data search cannot be trusted Priority: Low - looks like the site is dead anyways

To Reproduce Steps to reproduce the behavior:

  1. Search for datasets in the Mediterranean bounding box including the Gibraltar straits
  2. Find datasets in the results that are obviously not from this area

Expected Result This dataset should be marked as a point dataset.

Actual Result The dataset has a huge bounding box