Example Ar···SO₂

• The experimental lines are in the file arso2.lin.
• As can be seen in arso2.par and ar-so2.par, there are 15 independent parameters a 85 lines in the data set. Watson's A reduction was used in all fits, as indicated by the a in option line 3. This line indicates also that the spin-multiplicity is 1, that there are 2 ("vibrational") states, that K_a ranges from 0 to 8, that all interactions are to be considered, and that the a-axis is the axis for statistical weight.
  Two choices are available to implement the alternating spin-statistics of the two substates, one with one option line (line 3; arso2.par) and one with two option lines (lines 3 and 4; ar-so2.par). The version with one option line has VSYM = 10, WTPL = 1, and WTMN = 0. This causes even rotational states to have a weight of 1 and odd states to have a weight of 0 for ("vibrational") state 0, while the weights of even and odd states are exchanged for ("vibrational") state 1.
  NOTE: Employing the same coding, the role of even and odd rotational states may have been exchanged in older versions of the program !
  In the version with two option lines, line 3 is almost identical to the one with just one option line. However, VSYM = –1 explicitely indicates that there are additional option lines to the default defined in line 3. Line 4 defines the deviation from the default: ("vibrational") state 1 has the spin-statistical weight exchanged. VSYM = 0 in line 4 indicates that this line is the last option line.
  Different rotational constants have been used for state 0 and 1 along with common centrifugal distortion constants. It should be kept in mind that normally the differences in spectroscopic constants among the two (or more) states will decrease as the order of the constants increases. Of course, exception will exist, but it is quite likely that a different ansatz exists that avoids increasing differences in spectroscopic constants with increasing order of the constant.
  F_ac (code 410001) is the second order Coriolis coupling constant (or axis rotation term) of b-type symmetry. DE (code 11) is the energy difference between state 1 and state 0.
• The respective fit files are arso2.fit and ar-so2.fit, which yield identical results. The quality of the fit and the residuals between observed frequencies and those calculated from the spectroscopic constants are very similar to those in the IAM fit. Among the variety of reasons why the data have not been reproduced within experimental accuracy two of the likelier ones are: more parameters are needed to fit the data than can be safely determined from the input data; the estimation of experimental uncertainties is a bit optimistic for some lines.
• A different way to fit the two sets of rotational constants is to fit to a common set of rotational constants (those of state 0) and to the difference between state 1 and state 0. The respective parameter and fit files are delta.par and delta.fit.
• Another option is to fit to the average rotational constants and to half (!) the difference: mw.par and mw.fit are the corresponding parameter and fit files.
• The file ar-so2.int contains the information for the calculation of the rotational spectrum at 1 K. The a-dipole component within the states is slightly larger tha 0.25 D while the c-dipole component between the states is almost 1.5 D.
• The file ar-so2.out includes values for the partition function at selected temperatures. More states (higher F_max in the int file and higher K_max in the var file) have to be considered for higher temperatures ! See for example 20K.out ! (This file was generated with an old program version.)
• A prediction of the rotational spectrum at 1K is presented in ar-so2.cat. Note: extrapolations to higher J and K_a should be viewed very cautiously !
• Energies and mixing coefficients are given in ar-so2.egy.