Include conventional units

[kulnor]

Under the umbrella of the CODATA DRUM project focusing on the digitization and FAIRification of the Fundamental Physical Constant (knowledge base and API), we have documented the list of all the associated units, which can be found in this Google sheet. This is based on the NIST values and files.

Columns D/E/F in that sheet (si_expression, si_id, and si_url) provide the crosswalk to the SI digital framework.

There is, however, a small knowledge gap, as some constants use the conventional electrical units, such as the Ampere-90 and Coulomb-90.
While these are not SI, it is unfortunate that we do not have a crosswalk and definition (the same is true UCUM).

Could these be considered for inclusion in the reference point as listed below?

Note that the identifiers were chosen not to clash with the SI reference point unit expression (e.g., A90 would be A with exponent 90). We used XC as the Roman numeral for 90 instead (this could be changed).

! id	unit
AXC	ampere-90
CXC	coulomb-90
FXC	farad-90
HXC	henry-90
ohmXC	ohm-90
VXC	volt-90
WXC	watt-90

[Crparedes]

Dear Pascal,
Thank you for sharing the CODATA DRUM knowledge base and API. Very interesting project.
I apologize for the late reply.

This is something we should discuss further.

My first impression is that it is challenging, since the SI Reference Point implements the knowledge model of the SI Brochure, currently (9th Ed. v3.02) constrained to SI Base units, SI units with special names, and non-SI units accepted for use in the SI. The conventional electrical units are not part of any of these groups and are defined in terms of abrogated values of constants that we have not defined explicitly in the constants.ttl graph, which only contains the current values of the seven defining constants.

Probably, could you consider doing the crosswalk in the CODATA Knowledge model by providing their exact ratios (similarly to our use of si:hasNumericFactor for si:nonSIUnits in the units.ttl graph) to relate them to the values of the SI units in the SIRP?

Note that a similar situation might be necessary in the hypothetical case that the World Meteorological Organization would like in the future to link their knowledge model in the WMO Codes Registry to the SIRP: some units SI derived unit of time derivative of pressure or the Conventional unit of time derivative of pressure would require a conversion factor expressed as a ratio.

[larsbarring]

Dear Cristhian and colleagues,

As a future extension, I think it would be highly relevant for the SI Reference Point to make historic values available back to some suitable year or version. In science and engineering, the moving target of "current values" is not the only thing that matters: reference materials, design documents, and other more or less legally binding information of varying vintage ultimately depend on the SI units and their definitions in force at the time. This matters for metrological traceability too — recovering the context of older documents and measurements requires access to the corresponding historic values.

On a related note, I am pleased that initial contacts have now been established between the SI Reference Point activities and CODATA's work on establishing a semantic web presence for the Fundamental Physical Constants.

[kulnor]

Thanks for the feedback. I certainly do understand the challenge. Most importantly, I wanted to make sure this is documented somewhere in case anyone asks. I'll reach out to UCUM and QUDT about this as well when I get the chance.

[larsbarring]

some units SI derived unit of time derivative of pressure or the Conventional unit of time derivative of pressure would require a conversion factor expressed as a ratio.

Indeed, and this is something that I am currently looking into. Once one goes beyond the list of units covered by SI, things become interesting -- and considerably more complex. Not least in this perspective it is important to keep careful track of the precision, and uncertainty, of conversion factors and constants.