VA SIGMET with no volcano coordinates

[ernstdevreede]

From Annex 3 it appears that in VA SIGMET's the volcano's coordinates are optional for TAC.

[VA ERUPTION]
[MT nnnnnnnnnn]
[PSN Nnn[nn] or Snn[nn]
Ennn[nn]
VA CLD

In IWXXM (>=3.0) however these coordinates are mandatory in the EruptingVolcano type, so it's not possible to generate a legal IWXXM without coordinates.

Does this mean that the volcano coordinates are mandatory in both TAC and IWXXM?

[mgoberfield]

Hello,

In cases when the source volcano coordinates are not known, you can set element metce:position attribute 'nilReason' to the appropriate value.

For example:

<iwxxm:eruptingVolcano>
    <metce:EruptingVolcano gml:id="uuid.f59689d4-2376-475c-921f-91bea3f63cfb">
        <metce:name>UNKNOWN</metce:name>
        <metce:position nilReason="http://codes.wmo.int/common/nil/unknown"/>
        <metce:eruptionDate>2021-11-01T12:00:00Z</metce:eruptionDate>
    </metce:EruptingVolcano>
</iwxxm:eruptingVolcano>

Respectfully,

mark

[ernstdevreede]

Heelo Mark,

Thanks for your response.
I tried this earlier, but ran into a validation error.

My SIGMET (IWXXM3.0) had:
   <iwxxm:eruptingVolcano>
      <metce:Volcano gml:id="uuid.90e40563-23a0-49f3-a2c2-c115e2fed05d">
         <metce:name>GRIMSVOTN</metce:name>
         <metce:position nilReason="http://codes.wmo.int/common/nil/unknown" xsi:nil="true"/>
      </metce:Volcano>
   </iwxxm:eruptingVolcano>

This results in 2 validation errors:

iss:XML Schema validation error at line 141, column 94:cvc-elt.3.1: Attribute 'http://www.w3.org/2001/XMLSchema-instance,nil' must not appear on element 'metce:position', because the {nillable} property of 'metce:position' is false.===org.xml.sax.SAXParseException; lineNumber: 141; columnNumber: 94; cvc-elt.3.1: Attribute 'http://www.w3.org/2001/XMLSchema-instance,nil' must not appear on element 'metce:position', because the {nillable} property of 'metce:position' is false.
iss:Failed Schematron assertation: VolcanicAshSIGMET.VolcanicAshSIGMET-6: If a geometry is defined for iwxxm:eruptingVolcano//metce:position by providing @srsName, @srsDimension must equal to 2 and @aixsLabels must be non-empty===null

I beieve that if there is a nilReason, there also has to be the xsi:nil="true" ?
Is that not correct?

[mgoberfield]

Hello,

You are correct: 'xsi:nil=true' attribute should be there, but the METCE schema for EruptingVolcano/position element does not include the nillable='true' attribute, so the use of xsi:nil is flagged as you've pointed out.

For now, remove the xsi:nil attribute, then it should validate (it does for me in OxygenXML).

I've been working this issue in the IWXXM schemas. The task team will look into the possibility of fixing the METCE schema.

Respectfully,

mark

[ernstdevreede]

I tried this, but still run into the same
VolcanicAshsigmet:VolcanicAshSigmet-6 validation problem with

   <iwxxm:eruptingVolcano>
      <metce:Volcano gml:id="uuid.90e40563-23a0-49f3-a2c2-c115e2fed05d">
         <metce:name>GRIMSVOTN</metce:name>
         <metce:position nilReason="http://codes.wmo.int/common/nil/unknown"/>
      </metce:Volcano>
   </iwxxm:eruptingVolcano>

[ernstdevreede]

To me it looks like the validation rule VolcanicAshsigmet:VolcanicAshSigmet-6 makes it impossible to generate a Volcanic Ash SIGMET without a metce:position having all the required elements.
Nilreason with xsi:nil adheres to the xsd schema for metce, but does not validate with the Schematron rules.

[mgoberfield]

Hello,

You are missing the <eruptionDate> element, but I see your point w.r.t. the schematron rule.

Thank you for the bug report. We will try to get the METCE schema fixed.

I am a bit puzzled because I was able to validate this nilReason case against versions 3.0 and 2021-2 of the schematron rules without errors using both CRUX and oxygenXML applications. I am not a schematron expert so I will bring this case up with him at our next meeting to discuss these results.

Respectfully,

mark

[ernstdevreede]

Hi Mark,

My question might not have been clear enough. I'm discussing the structure of iwxxm:eruptingVolcano in VA SIGMET's, not the volcano description in VA advisories.

I don't see the eruptionDate at all in the eruptingVolcano element of the SIGMET xsd.

I must say the VolcanicAshSIGMET and the VolcanicAsh Advisory are a bit confusing in specifying the volcano information:
VolcanicAshSigmet:

<iwxxm:eruptingVolcano>
  <metce:Volcano>
    <metce:name>...</metce:name>
    <metce:position>...</metce:position>
  </metce:Volcano>
</iwxxm:eruptingVolcano>

whereas Volcanic Ash Advisory has:

<iwxxm:volcano>
  <metce:EruptingVolcano>
    <metce:name>...
    <metce:position>....
    <metce:eruptionDate>...
  </metce:EruptingVolcano>
</iwxxm:volcano>

So the SIGMET has no eruptionDate, but can't have an empty position element (with a nilReason)

Perhaps you didn't get any validation errors, because you were testing Volcanic Ash advisories?

[mgoberfield]

Hello,

Thank you for the note. For VolcanicAshSIGMET the <eruptingVolcano> element can be either of two types:

The difference between the two types is the addition of the eruption date in <metce:EruptingVolcano>, and I failed to catch your use of <metce:Volcano>, so thank you for the correction.

However, in either case, I still don't get the schematron error if nilReason is used in the <metce:position> element.

Respectfully,

mark

[ernstdevreede]

I definitely do get these Schematron errors (also in crux based Schematron validators), so not only is it impossible to generate these SIGMET IWXXM's but if I could it would probably never be decoded correctly because it will fail validation on input too.

For now I think the only solution to go on is that we'll have to assume that all VA SIGMETs carry the volcano coordinates; if not we don't generate IWXXM. Translation from TAC would also fail if the coordinates are missing.

I'm wondering if others may have thought up a better work-around.

@mgoberfield Thanks a lot for your answers

Regards,

Ernst de Vreede

[blchoy]

Thanks @mgoberfield for following up on this matter. I have no problem validating with both schemas and schematron for the sample VA SIGMET modified to have no erupting volcano location (see https://gist.github.com/blchoy/cfa2d8bb95e53c14e55a9e0b971ac7db).

Checking the schematron rule VolcanicAshSIGMET.VolcanicAshSIGMET-6 it should not fire when the XML instance has no including elements under <metce:position>. Could you confirm this rule was triggered by the XML instance fragment you mentioned?

[ernstdevreede]

I'm pretty sure the same construct doesn't validate in an IWXXM file with IWXXM3.0 rules.
It doesn't validate in my Java code with the VolcanicAshSIGMET.VolcanicAshSIGMET-6 error.
An example of a not validating IWXXM: https://gist.github.com/ernstdevreede/dd5c3afd7c140f4286b14430671e8c6d

Does this SIGMET validate on your side?

[blchoy]

Now I know what happened. Mark and I may have overlooked that you are making reference to IWXXM 3.0 (the next version IWXXM 2021-2 will be out within 2 weeks and my example is on this version). For VolcanicAshSIGMET.VolcanicAshSIGMET-6 in IWXXM 3.0 there was a known bug (see https://github.com/wmo-im/iwxxm/wiki/Confirmed-Issues-That-May-Affect-Operational-Versions) which had been fixed in IWXXM 3.0-dev. Unfortunately this version has never been made available officially on schemas.wmo.int. To get a copy of the fixed iwxxm.sch you can visit https://github.com/wmo-im/iwxxm/releases/tag/v3.0-dev.