Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 325, 745-754 (1997)

Previous Section Next Section Title Page Table of Contents

7. Conclusion

In the afore going section we have shown that there exists some discrepancy between the HST H-Ly [FORMULA] glow data and the theoretical approach. The widths of the theoretical spectra calculated with our radiation transport model are slightly smaller than the measured HST spectra. This problem is mainly caused by not taking into account the actual aperture of HST GHRS instrument and by using a simplified method in correcting for the Doppler shift due to the earth's motion in the ecliptic plane e.g. not taking into account sun-barycenter motion, earth-moon motion (see Sect. 3). Taking these effects into account needs much more work on the data what is not worth the effort because the uncertainties (noise) in the present HST-data are of comparable order.

The discrepancy, however, between the interplanetary hydrogen inflow velocity, determined with the HST spectra and the theoretical spectra, is not explained by an uncertainty in the data. The only uncertainty in deriving the hydrogen inflow velocity from the HST data is caused by a simplified calculation of the earth's velocity component along the detectors' line of sight (Sect. 3.1.2). This amounts to the order of 0.025 km/s. As we see in Fig. 7 a red-shift of the calculated upwind spectra by [FORMULA] 5 km/s, much higher than the admittable uncertainty of 0.025 km/s, decreases the difference between data and theoretical spectra. The HST upwind spectra would imply a hydrogen inflow velocity of [FORMULA] 21 km/s which is in a clear contradiction to inflow velocity measurements of the LISM helium by the ULYSSES GAS experiment (Witte et al. 1993). The ULYSSES GAS experiment indicates a LISM helium inflow velocity of 26 km/s and, assuming a dynamical equilibrium in the LISM plasma far away from the sun, the same inflow velocity of 26 km/s should be valid for LISM hydrogen. Also the theoretical description of the crosswind spectrum, best fitted with no red-shift at all, becomes worse for lowering the hydrogen inflow velocity.

Several authors (e.g. Lallement et al. 1993) tried to explain this velocity discrepancy with the hydrogen deceleration at the LISM interface. In fact close to the shock region at about 80 - 100 AU in the upwind direction the density models (e.g. Osterbart & Fahr 1992; Baranov & Malama 1993) predict such an effect for hydrogen (see Fig. 3). But taking this interface effect into account in a kinetic form, as is done by the density models used here, in the vicinity of the sun, where the majority of the H-Ly [FORMULA] sources seen by the HST GHRS instrument are located, no deceleration is left. The deceleration results from the fact of a different charge-exchange influence to the different parts of the H-velocity distribution function. It should not be identified with the action of a force. So only decreasing the hydrogen inflow velocity at infinity, would improve the theoretical description of the HST data, but this implies, as mentioned before, the HST data are in contradiction to the ULYSSES GAS experiment. Also the discrepancy of the HST crosswind spectrum is not resolved, since the best theoretical description is given by the unshifted case, and the theoretical spectrum becomes worse for lowering the hydrogen inflow velocity at infinity. Also the crosswind spectrum is affected by such a shift since the projection of the relative velocity of the earth-bound HST to the line of sight counts for the spectral shift.

We have mentioned the remarkable fact that Doppler red-shifts of the calculated upwind spectra by equivalent velocities of about -5 km/s noticeably improve the fits to the observed HST-spectra while with this procedure practically no improvement is achieved for the crosswind spectrum. As we could clearly rule out, a change of the actual ionisation rate within supportable limits is no remedy for this flaw in the theoretical representation either. Due to a change in the differential extinction of the hydrogen velocity distribution function a change in the effective bulk velocity will occur. However, the admitted magnitude of this change is much less than 5 km/s. Understanding this improvement in the upwind spectral fits as a serious hint to a needed correction in the modelled hydrogen dynamics one would have to ask how a decelerated hydrogen flow could be achieved in the upwind direction at solar distances that contribute to the HST-spectra. Since the distances of HST-relevant Ly [FORMULA] scattering sources are between 1 and 5 AU one should find an explanation for a decelerated flow at these distances. Though all interface models presented in the literature up to now can predict interface-induced hydrogen deceleration of the order of 5 km/s at large distances ([FORMULA] 60 AU), no deceleration are pointed out by these models for much smaller distances. Here one could only have a hydrogen flow decelerated by 5 km/s if it were already decelerated by this amount far ahead of the interface region, meaning that LISM hydrogen and helium should be dynamically de-coupled. Since this conclusion because of many physical reasons is hard to accept, we here thought of two alternative reasons why upwind hydrogen could appear decelerated at regions close to the sun (1 to 5 AU).

a) This could be due to anomalously large values of the solar Ly [FORMULA] radiation pressure like given by [FORMULA] (see Fig. 8) (unlikely as mentioned before) or even:

b) This could be since the "expected" upwind direction for hydrogen is not coincident with the LISM helium inflow direction but is tilted with respect to that by an angle [FORMULA] such that the projected Doppler velocity thereby becomes smaller by about 5 km/s. With simple algebra one calculates that a tilt angle: [FORMULA] = arccos (21/26) = [FORMULA]: would actually cover the needs. The following physical condition can be envisioned: Assume the solar system to be moving with 26 km/s into a direction characterized by the unit vector [FORMULA]. The LISM magnetic field [FORMULA] may be inclined to this direction by an angle [FORMULA] given by: [FORMULA]. Then a squeezed interface structure will be established as the result of magnetohydrodynamic stress forces (see Fahr et al. 1988). The resulting interface has a symmetry axis which is tilted by [FORMULA] with respect to [FORMULA]. While helium not coupling to this squeezed interface will still enter the solar system from the direction [FORMULA], hydrogen has to attain the imprint of this interface. Thus, if LISM hydrogen entering from the left side of [FORMULA] is more extinguished than that entering from the right side, this evidently then leads to a tilt of the hydrogen inflow direction towards the right side. The needed tilt of the interface axis with respect to the helium inflow could easily occur if magnetic fields of the order of a few µGauss are present in the LISM which are tilted by some angle with respect to the plasma bulk flow or LISM helium flow vector. The exact conditions for such tilted magnetic interfaces are analyzed in a paper by Fahr et al. (1988). The needed tilt angle of [FORMULA] = [FORMULA] which may be indicated from the above spectral fit procedures can then be interpreted in terms of LISM magnetic field inclinations and magnitudes needed. Recently Ratkiewicz et al. (1996) have published results on 2-D MHD simulation of the heliospheric interface configuration which even allow quantitative conclusions with this respect.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1997

Online publication: April 28, 1998