HP98 DataSet
For more details see: Tim
Holland & Roger Powell - An internally consistent thermodynamic dataset for
phases of petrological interest.
Journal of Metamorphic Geology
16, 309 (1998).
back to main page
TABLE 7b: The experimental data used in setting up the least
squares problem for solving for the enthalpies of formation of the
phases in LSQDS, and the results of the least squares analysis. The
experimental data are grouped according to chemical subsystems:
(click on system below to go to relevant page for details)
CONTENTSThe first columns give the experimental data, in the form of
the P, T, x(CO2) and ln K of the experiments, followed by the
corresponding enthalpies of reaction calculated using the entropies,
volumes etc in Table 5. LSQDS converts the enthalpies of reaction
into an enthalpy of reaction bracket given in the first line of the
summary. The least squares enthalpy of reaction (cH) is given in the
second line of the summary, with its calculated uncertainty. The calc
column gives the calculated T, P or ln K of the experimental
conditions using the least squares results, followed by its
uncertainty. The miss column gives the amount that the calculated T,
P or ln K is outside the experimental bracket (if at all). In the
summary information for each reaction, the code on the first line
relates to the calculation of the enthalpy of reaction bracket (see
Powell and Holland, 1993); the number at the end of the second line
(if present) is a factor used in weighting experimental results in
the least squares analysis; the comment making up the third line
gives an overall appraisal of the relationship between the enthalpy
of reaction bracket and the calculated enthalpy of reaction; uH on
the fourth line is a measure of the uncertainty on each end of the
enthalpy of reaction bracket arising from experimental uncertainties
(in P, T etc), d/s is a measure of bracket overlap normalised to the
uncertainty on bracket ends as discussed in Powell and Holland
(1993), and h is the hat matrix diagonal.