"Reference procedure" has 2 1/2 meanings:

1) In a legal context, "reference methods" are those that are defined as correct. They may be very tedious, so labs can use other methods, but if there is a dispute, the result of the reference method counts. This is rather frequent in food analysis (e.g. the Kjeldahl method is the reference method for protein, although most people use the Dumas method) and also in clinical chemisty: here, organisations like the IFCC have defined "reference methods" for some parameters (e.g. enzyme activity), against which other methods (e.g. ELISA) are calibrated.

2) If one calibrates one method against results of another method, then the one against which one calibrates is the "reference method", as it is suposed to be more accurate (it may also only be the more established/older one).

1/2) The term "reference method" is also used as a meaningless, but prestige-carrying term to state that the method one uses (and that is usually developed by oneself) is very good/much better than any other method. So it's an easy way of self-promotion.

There is an interesting paper on : ABSOLUTE METHODS IN

ANALYTICAL CHEMISTRY

(Technical Report)

http://pac.iupac.org/publications/pac/pdf/1995/pdf/6711x1905.pdf

An interesting paper on uranium:

Article Presentation of Differential Laser-induced fluorimetry as a ...

There is an interesting paper on uranium:

Paper has been published:

Application of a differential technique in inductively coupled plasma emission spectrometry: Presentation of a relative reference measurement procedure for the determination of total mass fraction of uranium in mineralised rocks and similar matrices,

P. Murugesan, C. R. Khorge, A. A. Patwardhan, Manjeet Kumar, P. K. Tarafder

and D. P. S. Rathore*

Journal of Analytical Atomic Spectrometry (Impact Factor: 3.16). 09/2014; 29(10):1912-1917. DOI: 10.1039/C4JA00224E