2012-11-26 working group VegBIEN VegCSV¶
To do¶
- snapshots of DB
- criteria of when to make new snapshot
- VegBIEN data dictionary with definition of each table
- schema overview narrative
- 3-10 pages or short?
- species summaries
- cover class data
- how to combine?
- analytical DB in bien_web
- parse plot size out of plot method desc
- -1 is not an elevation
- add units to slope: % vs. degrees
- example TurboVeg dataset
- how provider runs import of new data
- empty VegBank input datasource
- Recce by strata
- solicit contributions to VegBIEN
Validations¶
- user validations from each data provider
- compare # CTFS, REMIB records to input rows
- avg # of stems per plant in 10 Ha plots
- change validations to run against analytical DB
- compare # plots in input data vs. dest DB
- each color-coded table group needs validations
- a dozen standard aggregations, correlations on data
- datasources to validate: CVS, FIA, GBIF, NCU, NVS, NY, TurboVeg, VegBank
From Susan:
Also see attached Potential validations on BIEN data 1 Dec.docx
- Potential validations on BIEN data
- Develop a series of tests
- Number of plots in source= number of plots in BIEN3
- Cross tab of source X country
- Map of all occurrence data; map of occurrence data by source
- Box plots of values for all quantitive variables to examine extremes [e.g. stem density, dbh, basal area, species richness others]
- Duplicate taxon rows within a plot or subplot (or tier) for aggregate taxon records.
- Check for incorrect values for some fields (i.e. slope should never be negative or > 90; aspect should never be negative or >360; altitude should never exceed ????)
- Duplicate tags within a single plot
- Can we pull out repeated measures of individuals? Need to identify which source data sets have provided repeat measures of individuals?
Schema changes¶
rename soilobs>soilsample- add coords to soilsample
- rename plantobservation to individualobservation
rename concept_source_id>concept_reference_id-- make taxonverbatim<->taxondetermination many-to-many
- this will add support for species aggregates and hybrids
- use array of fkeys to allow scoping the taxondetermination by the corresponding taxonverbatim?
- or make taxonverbatim<->taxonlabel many-to-many?
- add traits table
- row: trait name + trait value + taxon/stem
- add TNRS-parsed annotations
Info¶
- payoff to data provider is TNRS, GNRS validations
BIEN2¶
- traits are on nimoy in bien2.TraitObservation
- need permission to distribute traits from the following sources: Glopnet, LEDA
VegBIEN¶
VegX¶
- VegX = Vegetation exchange schema, not Vegetation XML schema
- has data dictionary, with definitions sourced to DwC, etc.
NVS¶
- NVS is pronounced "nivs"
- is separated out by projects
- provides multi-sheet spreadsheet
- a few data tables + lookup tables
- e.g. stem diameter data would have sites, stems, metadata
- has 150 tables
- most are materialized views
- some are redundant