Insights

General Metadata Content

Much citation of involved individuals ObsProp and Methodological information, if available, only encapsulated in prose descriptions

Procedural Prose

The dataverse.cirad.fr Replication Data for: datasets tend to have an Excel tab "Read me", where the column headers of the data tabs is explained. The descriptions provided for each column tend to be composed of:

• Observable Property
• Procedure
• FoI/Sampling (e.g. "from soil samples", "from soil solution")
• UoM, always enclosed in () at the end

Examples:

• Total concentration of Zn in soil solution extracted from soil sample as decribded in the related paper (mol Zn l-1)
• Mean biomass of earthworms of each replicate at the the end of the exposure period (g)
• Inorganic phosphorus extracted with the water method as described in the related paper (mg Pi/kg)
• Mass of Zn supplied to soil by organic residue application overtime from Eq. (1) in the related paper (mg Zn kg-1)

Maybe we can learn to parse this

Experimental vs. Monitoring

Quite a lot of the datasets (those without spatial reference) are experimental, list treatments, impacts thereof.

These often don't include time, or if so relative to the treatment time.

FoI information often a mix of Site, Field, Treatment...

Vocabularies used

AGRIS/CARIS Vocab

Many datasets utilize concepts from AGRIS/CARIS, only reference to html doc, no concept links