Sigmet and Volcanic Ash Advisories - filling volcano element.

[blchoy]

Discussed in #337

^{Originally posted by Mattonius5 September 24, 2024}
Hello,
This affects All IWXXM versions, concerning VAA/Sigmet element "volcano/erupting volcano", especially "metce:eruptionDate", Given the VAA example below:

<iwxxm:volcano>
    <metce:EruptingVolcano gml:id="uuid.1b9ee023-c87d-4fa0-bf9b-bb1c7215286c">
        <metce:name>KARYMSKY 1000-13</metce:name>
        <metce:position>
            <gml:Point gml:id="uuid.687c6e95-460a-4397-a602-ea8483acc3fc" srsDimension="2" axisLabels="Lat Long" srsName="http://www.opengis.net/def/crs/EPSG/0/4326">
                <gml:pos>54.03 159.27</gml:pos>
            </gml:Point>
        </metce:position>
        <metce:eruptionDate>2008-09-23T01:30:00Z</metce:eruptionDate>
    </metce:EruptingVolcano>
</iwxxm:volcano>

Question 1, Volcanic Ash Advisory:

Element "iwxxm:volcano" requires "metce:EruptingVolcano" which then requires "name", "position" and "eruptionDate".
My understanding is, that "eruptionDate" should be extracted from mandatory TAC element "Eruption details". Annex 3 allows for two options "Free text up to 64 characters" and "UNKNOWN".
I've collected extensive collection of Volcanic Ash Advisories in TAC form from all around the world. Around 60% of them do NOT contain eruption date in any form inside the "Eruption details" element. In these instances, eruption date cannot be extracted, thus I cannot satisfy requirements of "metce:EruptingVolcano" by filling it's "eruptionDate".

Observations:

• When "eruptionDate" is missing, produced IWXXM fails schema validation.
• If I remove iwxxm:volcano element, Schematron validation fails on "VolcanicAshAdvisory.VolcanicAshAdvisory-3"
• Having empty iwxxm:volcano/ passes validation.

What should I do in these three following cases?

1. generate translationFailed report
2. create empty element iwxxm:volcano/ (this passes schema + schematron validation), however information about volcano name and position are lost.
3. Use date from field "DTG"
4. Use date from field "OBS/EST VA DTG"
5. Something else?

Question 2 - Volcanic Sigmet

Volcanic Sigmet uses a similar element "iwxxm:eruptingVolcano" with very important distinction, it requires either element "metce:volcano" or it's substitution group "metce:eruptingVolcano". This way the name and position can be filled as part of complex type "VolcanoType" regardless presence of eruption date. If eruption date exists, "metce:eruptionVolcano" can be used (same as VAA). If eruption date does NOT exist, "metce:volcano" can be created. Thus we can pass all schema or schematron rules.

Why was not VAA modeled in the same fashion as Volcanic Sigmet in regards to "volcano" element?

Questions 3 - Volcanic Sigmet

Considering the example for VAA on top, Sigmet looks almost identically, instead of iwxxm:volcano, there is iwxxm:eruptingVolcano, the rest is as described in question 2.
Annex 3 Table A6-1A, element Phenomenon for VA has following structure:

• [VA ERUPTION] (optional part)
• [MT volcano_name] (optional part)
• [PSN single_point_coordinate] (optional part)
• VA CLD (mandatory part)

What to do if Volcano name is not present?

1. Create translationFailed
2. Use "UNNAMED" (VAA allowed name)
3. Use "UNKNOWN" (VAA allowed name)

Subsequent questions
What does the "VA ERUPTION" optional element means? My deductinon form online research is following:

• If present, it indicates that Volcano just erupted, thus I can convert it using "metce:eruptingVolcano" filling "eruptionDate" from "observation date".
• If not present, eruption date is unknown, thus "eruptionDate" is absent, thus I shall use "metce:volcano" element.

APAC regional SIGMET guide states following:

1. If the erupting volcano is known and named then the following format is appropriate: VA ERUPTION MT ASHVAL PSN <lat,lon> VA CLD
2. If the location of the eruption is known but is un-named then the following format is appropriate: VA ERUPTION PSN <lat,lon> VA CLD
3. If the source of the volcanic ash is uncertain, then the following format is appropriate: VA CLD

In case 1, I would fully create element "metce:volcano" or "metce:eruptingVolcano" (depends on if any date should be added) like below:

<iwxxm:eruptingVolcano>
    <metce:Volcano>
        <metce:name>MT HEKLA</metce:name>
        <metce:position>
            <gml:Point>
                <gml:pos>63.98 -19.67</gml:pos>
            </gml:Point>
        </metce:position>
    </metce:Volcano>
</iwxxm:eruptingVolcano>

In case 2, I would add to volcano:name "UNNAMED", like below:

<iwxxm:eruptingVolcano>
    <metce:Volcano>
        <metce:name>UNNAMED</metce:name>
        <metce:position>
            <gml:Point>
                <gml:pos>63.98 -19.67</gml:pos>
            </gml:Point>
        </metce:position>
    </metce:Volcano>
</iwxxm:eruptingVolcano>

In case 3, I would NOT include iwxxm:eruptingVolcano in IWXXM as it's schema cardinality is 0-1 and no Schamatron rule enforces its presence.

Thanks for advise and guidance how should these elements be converted and please correct me where I am wrong.
Matous.

[amilan17]

https://github.com/wmo-im/iwxxm/wiki/TT-AvData-Discussion-2025-Jan-22 notes:

@blchoy to update for FT2025-2

[blchoy]

There is one thing I am not aware of.

metce:VolcanoType is an extension of gml:AbstractFeatureType, viz:

<complexType name="VolcanoType">
   <complexContent>
      <extension base="gml:AbstractFeatureType">
         <sequence>...</sequence>
      </extension>
   </complexContent>
</complexType>

and metce:EruptingVolcanoType is an extension of metce:VolcanoType:

<complexType name="EruptingVolcanoType">
   <complexContent>
      <extension base="metce:VolcanoType">
         <sequence>...</sequence>
      </extension>
   </complexContent>
</complexType

Now iwxxm:eruptingVolcano in SIGMET is defined as:

<element name="eruptingVolcano" type="metce:VolcanoPropertyType" minOccurs="0" maxOccurs="1">

But both of the following present in sigmet-VA-EGGX.xml can passed validation.

<iwxxm:eruptingVolcano>
   <metce:Volcano gml:id="uuid.99646618-9ec7-47c0-8b50-daba5b6b1da9">
      <metce:name>MT HEKLA</metce:name>
      ...
   </metce:Volcano>
</iwxxm:eruptingVolcano>

<iwxxm:eruptingVolcano>
   <metce:EruptingVolcano gml:id="uuid.1b9ee023-c87d-4fa0-bf9b-bb1c7215286c">
      <metce:name>KARYMSKY 1000-13</metce:name>
      ...
   </metce:EruptingVolcano>
</iwxxm:eruptingVolcano>

I have to say that this is not the intention.

Solution 1:

The correct way to deal with this issue is to modify metce.xsd to allow metce:eruptionDate to be nillable:

<element name="eruptionDate" type="dateTime" nillable="true">

and in case the eruption time/date is UNKNOWN in VAA:

<iwxxm:volcano>
   <metce:EruptingVolcano gml:id="uuid.1b9ee023-c87d-4fa0-bf9b-bb1c7215286c">
      <metce:name>KARYMSKY 1000-13</metce:name>
      <metce:position>
         <gml:Point gml:id="uuid.687c6e95-460a-4397-a602-ea8483acc3fc" srsDimension="2" axisLabels="Lat Long" srsName="http://www.opengis.net/def/crs/EPSG/0/4326">
            <gml:pos>54.05 159.45</gml:pos>
         </gml:Point>
      </metce:position>
      <metce:eruptionDate xsi:nil="true" nilReason="https://codes.wmo.int/common/nil/unknown"/>
   </metce:EruptingVolcano>
</iwxxm:volcano>

Solution 2

If METCE is not to be touched, then we may want to skip its use by defining a new volcano feature for use by VAA only. This should not border too much as its schema will have to be renewed with the removal of aviation colour code in the message.

Just some thoughts.

[blchoy]

Regarding Question 3 on VA SIGMET:

What to do if Volcano name is not present?

Use UNNAMED or UNKNOWN as appropriate, but the point is that iwxxm:eruptingVolcano/metce:Volcano/metce:name must present.

What does the "VA ERUPTION" optional element means?

My interpretation is that an eruption has occurred/is occurring. In VA SIGMET we do not have a separate indicator to indicate this state; the presence of iwxxm:eruptingVolcano should be adequate this state.

Now only metce:Volcano should be presented, as there is no requirement in Annex 3 to provide the eruption date/time.

In case of re-suspended volcanic ash (i.e. no erupting volcano), VA CLD will be used in the TAC message and for IWXXM one should not mention iwxxm:eruptingVolcano.

[amilan17]

https://github.com/wmo-im/iwxxm/wiki/TT-AvData-Discussion-2025-Jun-25 notes:
the team discussed and decided to add a new attribute for use in VAA

[blchoy]

Done.

[blchoy]

Implemented in IWXXM 2025-2.