Is there a standard, coded table of physical units?

I am a programmer in the medical/biosignals area, and I want to represent physical units in a database table. Ideally, I would like to have a code that uniquely identifies a given physical quantity.

As an example, ISO defines strings to represent languages, such as "en-US", or "pt-BR". I am looking for similar standard (ideally from ISO or similar) that identifies electric potential, force, angle, acceleration, etc.

• Have you considered using dimensionnal analysis ? Every physical unit can be expressed as a product of meter, kilogram, second, mole, Kelvin, Ampère, and candela. Dec 22 '16 at 14:01
• You might want to look at udunits which defines a lot of units tools and has a big database of standard names for them. It's not a formal standard: for that you want SI or something as someone else mentioned. Unfortunately lots of fields hVe their own idiosyncratic units (inverse femtobarns, Sverdrups).
– user107153
Dec 22 '16 at 14:14
• I agree with @tfb that udunits is as close as one can get to a standard, coded table of physical units. It's a group effort by Unidata, which in turn is a part of UCAR (the University Corporation for Atmospheric Research). The latest and greatest is on github, github.com/Unidata/UDUNITS-2 . Their database is in the form of five XML files, one for prefixes and the other four for various classes of units (base, derived, accepted, and common). Dec 22 '16 at 16:10
• Note well: This database is not and cannot be complete. An example: Astronomers (particularly radio astronomers) oftentimes use arcseconds, abbreviated as "as". For example, a mas is a milliarcsecond (a handy unit in high precision radio astronomy). The abbreviation "as" unfortunately introduces ambiguities: "as" matches both arcsecond and attosecond, and "das" matches both deciarcsecond and decasecond. (No sane person would use either deciarcseconds or decaseconds, but the ambiguity exists.) Conflicts between specialized units used in other sciences and the more standard units also exist. Dec 22 '16 at 16:20
• @DavidHammen: Yes, in the context of udunits, 'Sv' is particularly relevant of course! Where I work it means sverdrups, not sieverts, which are extremely different units.
– user107153
Dec 22 '16 at 17:19

2 Answers

That would be the International System of Units by the Bureau International des Poids et Mesures, as seen here.

• Nice document. For non-French readers, the English translation starts at page 99. Dec 22 '16 at 14:09
• Good! I took a look, but didn't find a field that could be used as an "unique identifier" (at least in ASCII). For example, "speed" is language-dependent (vitesse, velocidade, etc), and symbols include some unicode chars (like "ohm"). What do you think about it? Dec 22 '16 at 14:11
• If it's the dimension you want, and not the unit, the standard symbols for it are L for length, M for mass, T for time, I for electric current, $\Theta$ for thermodynamic temperature, N for amount of substance and J for luminous intensity. All possible dimensions can be written as a combination of those dimensions with various powers. Dec 22 '16 at 14:14

As far as I know, no standard provides a list of unique identifiers for each quantity. However, a fairly complete list of quantities with recommended names and symbols is provided by IUPAC in the so called Green Book.

Elecrical quantities are also defined by IEC in its online Electropedia, which provides also translations of terms in different languages.

However, take into account that there are variations in terminology between fields and that the recommended names are not always used.