search

i-adopt / examples

This is a repository of I-ADOPT variables
https://i-adopt.github.io/examples/
0 stars 1 forks source link

total organic carbon in soil #3

Open mabablue opened 1 year ago

mabablue commented 1 year ago

PID of Variable: http://vocabs.lter-europe.net/EnvThes/30325

Variable Description:

Property (label): amount of matter

Object of Interest (label): Carbon

Matrix (label): organic matter

Context Object(s) (label): soil

Constraint(s) (label) and which Entity it constraints:

Dimension Information:

Applicable Unit(s):

Link to a Publication of the Variable or Method:

Link to Turtle File (ttl) in this Repository (optional):

gwemon commented 4 months ago

@mabablue Your property above says "amount of matter" yet it points to a concept in EnvThes that is "amount of substance". The SI units for "amont of substance" in the SI system is the mole: https://www.nist.gov/pml/owm/si-units-amount-substance. Do you expect/restrict the results to be expressed in a unit derived from moles? or could it also be expressed as a mass?

mabablue commented 3 months ago

hi Gwen, thanks that is actually a typo... it should be amount of substance. This is the corrected version: variable: total organic carbon in soil property: amount of substance hasObjectOfInterest: carbon matrix: organic matter hasContextObject: soil

The unit used in Integrated Monitoring for TOC is: mg/kg

mabablue commented 3 months ago

shouldn't we add hasApplicableUnit as extension to I-ADOPT?

gwemon commented 3 months ago

Hi Barbara, if you define a variable as having a property "amount of substance" then the result is expected to be in a unit derived from "moles", the SI unit for amount-of-substance. What you have is a mass of carbon per mass of substrate. So you could define your I-ADOPT variable as that i.e. property is mass. But also, because moles and mass of elemental composition are so easy to convert from one to another there is also the option to use a broader term for the property that could justify having units in moles or mass . This is the approach we took in the BODC PUV (that decision was taken a long long time ago!).

gwemon commented 3 months ago

shouldn't we add hasApplicableUnit as extension to I-ADOPT? I think that it is too early at this stage to decide this. There are multiple ways we could link the vrariable to its applicable units (directly with hasApplicable as you say or via the vector dimension. However before we consider this, we need a full understanding of the impact it will have. So it would be good to take the habit of considering the units when analysing/decomposing a variable. Then as we work with more examples, hopefully the best way to link the units to the variable will become more obvious.

gwemon commented 3 months ago

Our closest P01 concept to your example would be: https://vocab.nerc.ac.uk/collection/P01/current/OCCNXXXX/ i.e. concentration of organic carbon measured in sediments. In the mapping it says unit is "percent" but this is just what BODC uses. Other users must specify the units alongside the P01 code. If starting from fresh, I think that I would recommend creating a variable that is aligned to one vector dimension. i.e. in your case, [Mass] per [Mass]. In our case this is no longer possible. However how do you capture the fact that this is relative to the "mass" of soil? Alternatively you could have the property defined as "Proportion by weight" (i.e. [Mass] of Carbon in Organic Matter per [Mass] of soil etc... (see e.g. http://vocab.nerc.ac.uk/collection/S06/current/S0600067/)