Reporting small Delta 17O values: existing definitions and concepts

The three-isotope tracer Delta(17)O is increasingly used in atmospheric chemistry and other research areas. Thanks to the development of isotope-ratio mass spectrometry (IRMS), delta(17)O and delta(18)O can be determined with a precision of a few 0.01 per thousand, and values for Delta(17)O may be calculated with similar precision. However, interpreting small and precisely determined Delta(17)O values as a deviation from an expected mass-dependent fractionation process is not straightforward. Several aspects are of high importance. In the present paper we review existing definitions, formulas and some other aspects of Delta(17)O reporting. One of the most confusing aspects is a variance of definitions and corresponding formulas. While Delta(17)O is traditionally defined to characterise a data point, i.e. Delta(17)O is considered as a deviation from an expected mass-fractionation line, the recently introduced definition (Miller MF. Geochim. Cosmochim. Acta 2002; 66: 188) characterises a fractionation line itself, in terms of its ordinate intercept. The formulas corresponding to this definition gives a characteristic for a specific process. When the 'traditionally defined' Delta(17)O is in use, an expected fractionation processes--the key point for Delta(17)O reporting--should be defined and parameterised with the same accuracy as intended for reporting Delta(17)O. When Delta(17)O is reported for a data point, not only a value for lambda but an ordinate intercept of a reference fractionation line should be given with high accuracy. We note that defining a single fractionation process is hardly possible for many natural compounds. For such compounds we propose to use a phenomenological reference line, namely an isotope composition range of natural sources. Next, aspects of Delta(17)O comparison and mass-balance calculations are considered. All the aspects considered for Delta(17)O may be relevant for others three-isotope tracers, e.g. Delta(33)S.