Giters

inbo / data-publication

🔓 Open biodiversity data publication by the INBO

Home Page:https://ipt.inbo.be

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Updates for inboveg-niche-vlaanderen-events SQL

peterdesmet opened this issue · comments

commented

Note: I verified version 0.7. See also specification.

Events

  • datasetID: populate with DOI once known
  • stateProvince: translate to English: [Antwerp, East Flanders, Flemish Brabant, Limburg, West Flanders]
  • stateProvince: remove XY Onbekend
  • municipality: remove XY Onbekend
  • county: remove
  • localities: if we cannot find coordinates for the 9 records, we have to leave some columns empty
  • locationAccordingTo: Watina (verified with Els) *changed to MILKLIM-areas
  • verbatimEventDate: remove (only single day dates)
  • samplingEffort: add after samplingProtocol and populate with {mossesIdentified: true}. Let me know if this is the case for all events...

Measurements

  • measurementID: decide on 8 or 10 numbers at the end (currently different between H and P). The occurrences use 8.
  • measurementType: use one of: [shrubsCoverage, treesCoverage, herbsCoverage, mossesCoverage, totalCoverage, totalCoverage]: use terms from http://rs.gbif.org/sandbox/extension/releve_2016-05-10.xml
  • measurementMethod: [coverage estimate in 1%-scale, height estimate in cm] (verified with Els) terms splitted according to measure and used ´coverage estimate in 1%-scaleandestimate`
  • measurementRemarks: remove, the fact that mosses were assessed is a property of event, best in samplingEffort

Occurrences

  • id: the first part is missing: IV2013092411562912 instead of INBO:INBOVEG:IV2013092411562912. I don't understand why it passed by the IPT.
  • occurrenceID: currently in format INBO:NICHE:00069737, shouldn't we have INBO:INBOVEG:OCC:00069737?
  • scientificName: I think we can leave "groep" in the name, maybe translate to "group"
  • taxonRank: translate:
Gen: genus
Spp: species
SppGrp: speciesAggregate
SppHyb: species
SubSpp: subspecies
Var: variety
(blank): depends on the data, see below
  • For taxonRank = (blank), correct in the database or do a case. I notice those don't have a taxonID either.
Cladonia coniocraea: species
Pseudoscleropodium purum: species
Sambucus nigra cv laciniata: ??? ask Els
Sphagnum recurvum var. brevifolium: variety
Sphagnum recurvum var. recurvum: variety
  • verbatimTaxonRank: remove
  • organismQuantityType: use Londo scale (1976)
  • eventID: remove, id is sufficient. wontfix
  • scientificName: Valeriana repens has 2 taxonIDs: INBSYS0000008260 & INBSYS0000008262
    need to be checked

    Relevé

  • mossesIdentified: use True instead of yes
  • treeLayerHeightInMeters & shrubLayerHeightInMeters: there is no data, was this never assessed? If so, should we have those fields?
  •  coverScale: remove
commented

localities: if we cannot find coordinates for the 9 records, we have to leave some columns empty

--> we can always find them, but for the moment we leave them empty (as they should be updated in the original list (:RS2012080211350639))

For these records we do not have a coördinate for the moment.
IV2013112510310640
IV2013112515474481
IV2013102210343020
IV2013102210461283
IV2013091014433694
IV2013091014550635
IV2013091015073587
IV2013100310221822
IV2013100311385184

commented
  • measurementMethod: [coverage estimate in 1%-scale, height estimate in cm] (verified with Els)

--> its a union query, so we have half: coverage estimate in 1%-scale or Percentage Pct-1-100 to decide and half estimate or height estimate in cm (cm is redundant, also to decide)

commented

eventID: remove, id is sufficient. --> it's called eventID also as a term.. I would leave it like this

commented

Valeriana repens Host
Valeriana officinalis subsp. repens (Host) O. Bolòs et Vigo

Moet gechecked worden...

commented

coverScale was removed in favour of using samplingProtocol in the core data. See gbif/rs.gbif.org@254aea0#commitcomment-17417620. We have currently have:

  • samplingProtocol: vegetation plot with Londo scale (1976) or (even we don't use relevé: vegetation plot with Londo scale (1976) with mosses identified`
  • coverScale: percentage Pct-1-100

Maybe we should combine everything in samplingProtocol as vegetation plot with 1976 Londo scale (mosses identified, 1%-scale percentages) (to verify with Els)

commented

Regarding the 9 without coordinates: those will just fail the text for now. We can fix this later.

commented

Version 0.8 only contains 24 occurrences. To check!

Most other things should be covered once we can run specification tests.