From the comparison of different theoretical calculations concerning expected Ly spectra shown in Fig. 5 one can easily conclude that the different ingredients entering the calculations in case 1 and case 2 can definitely influence the values of LISM parameters best fitting the actual, observational HST results. Compared to case 2 (realistic solar profile, self-absorption, active interface), in case 1 (flat solar profile, no self-absorption, no interface) one derives spectral features with smaller blue shifts for the upwind and crosswind spectra and with substantially higher spectral peak intensities though starting from identical LISM parameters. From this experience it turns out that for a reliable parameter analysis based on comparisons of data with theory one has to include effects only appropriately taken into account by case 2 of our calculations. In the following spectral analysis we therefore start out from "case 2" calculations (i.e. actual solar Ly emission profile, interplanetary self-absorption, heliospheric interface effect is taken into account) for a definite set of LISM parameters (see Tab. 1).
In Fig. 6 we show HST-GHRS spectra and theoretical spectra obtained by the case-2 calculations described in this paper for a set of LISM parameters mentioned in Tab. 1. Although for the set of adopted LISM parameters the calculated spectra fit the data fairly well, one may nevertheless notice spectral regions where data and theory clearly deviate from each other. In order to clearly identify such regions in Fig. 6 where deviations become manifest we have also plotted the differences ( - ) as function of the wavelength. ( is the best fit result to the HST data applying a Voigt-profile of 30000 K, see Clarke et al. (in prep.)). Though it can be seen that these differences are always smaller than the intensities, especially in the 94-upwind spectrum (Fig. 6a) one may notice non-negligible deviations which seem especially due to the fact that the theoretical spectrum is too much blue-shifted with respect to the data.
To improve on this fact we have tested how an artificial red-shift of the theoretical spectrum by 0.02 A (Fig. 7) (corresponding to a bulk velocity decrease by 5.0 km/s, respectively) would reduce the resulting intensity differences. It can be seen when comparing Figs. 6 and 7 that a red-shift by 5 km/s would lead to definitely better fits for the upwind spectra while larger shifts would again increase the resulting differences. For the crosswind spectrum the unshifted theoretical spectrum shows the smallest deviation between the data and the calculated spectrum i.e. a general shift does not improve the synoptic fit of all spectra. As we can show, there is much less need for a redshifting of the calculated spectra if larger values for the solar Ly radiation pressure are adopted for the period of the HST-observations. Taking a value of which could easily be justified on the basis of the SME satellite measurements (Rottman 1988) or the SOLSTICE-measurements (Rottman et al. 1994; White et al. 1994) one would be left only with a redshift need corresponding to 2.5 km/s (Fig. 8). However, high values for µ can only be expected at near solar maximum, while for the period of our HST-observations (1994, 1995) definitely lower values must be expected.
The downwind spectra (spectrum from 09.03.96) not shown here, but very similar to the spectrum from 06.03.95 Fig. 6 have a very low significance level (i.e too few statistics). Thus no reliable fits for determining width and position of the interplanetary Ly spectrum could be done for these two downwind data sets of the HST. Though in these spectra a clear signature of the Ly glow is seen, they can hardly be used for interpretation purposes (see below).
© European Southern Observatory (ESO) 1997
Online publication: April 28, 1998