6 Possible dichotomy in the surface composition of comet Hale-Bopp

The relative fraction of Mg poor silicate grains needed to be increased gradually with heliocentric distance/solar phase angle in order to match the polarization colour at phase angles larger than $30 ^\circ$. Presence of the component Aol with size distribution b in the ensemble causes an increased absorption (Dorschner et al. 1995) and hence an increase in positive polarization at 0.3650 $\mu$m compared to $0.4845~\mu {\rm m}$ and longer wavelengths (Fig. 4). The same effect can be mimicked if the organic part of the grain resembles the very red material found in the Kuiper-belt objects (Tegler & Romanishin 2000) instead of the composition of organic refractory and amorphous carbon of equal mass assumed in the present work (Sect. 4.2). Organic refractory (Li & Greenberg 1997) has a gradual increase in k with decreasing wavelength in the optical region, while for amorphous carbon (Rouleau & Martin 1991) k although large, decreases with decreasing wavelength. As very little is known about the nature of cometary organics, any change in the assumed composition of the organic component may also alter the shape of the polarization curve and hence the derived compositions. Caution is also needed in interpreting the required change in the composition of Aol (or grains with enhanced blue absorption) with heliocentric distance/phase angles for a good fit, as a real effect. This unreliability arises due to the use of Mie scattering and Effective Medium Theory in our model which has been shown to be inadequate in reproducing the polarization vs. phase curves for porous grains of non Rayleigh inclusions (Wolf et al. 1998). As naturally occurring composite grains are more likely to have non Rayleigh inclusions than to be of homogeneous composition, the results presented here may be substantially improved if DDA is used. On the other hand, due to changes in solar illumination geometry during the apparition of the comet, there is also reason to suspect that the requirement of change in composition may at least be partially real due to possible dichotomy in the surface composition in the northern and southern regions of comet Hale-Bopp. The latter regions being sunlit during and prior to 1996 and the former regions in 1997. The porcupine like appearance of the jets (O'Meara et al. 1996) have been explained due to dust emission from discrete southern sources at latitudes of $-70 ^\circ$, $-35 ^\circ$ and $0 ^\circ$ by Sekanina & Boehnhardt (1998) and of $-65^\circ$, $-35 ^\circ$ and $\pm 5^\circ$ by Vasundhara & Chakraborty (1998). The prominent shell structures near perihelion have been attributed by these authors to a source near $60 ^\circ$ latitude. As shown in the top panel of Fig. 9, the solar phase angle while oscillating between oppositions steadily increased with heliocentric distance. The dates of the available polarimetric observations in red are marked with the same symbols as in Figs. 4 and 6. Variation of the angle $B^{\prime }$ (Sect. 2.2) with heliocentric distance is shown in the middle panel.

Figure 9: Top panel: variation of the solar phase angle with heliocentric distance. The dates of the available polarimetric observations in red are marked with the same symbols as in Fig. 6. Middle panel: variation of the angle $B^{\prime }$ (Sect. 2.2) with heliocentric distance. Lower panel: the activity of the sources near $-65^\circ$ (solid line) and $+65^\circ$ (dashed line) as determined by the average value of the cosine of the local zenith distance of the Sun over a rotation period of the comet, ${<}\cos{z_{\odot}}{>}$.

The activity of the sources near $-65^\circ$ and $+65^\circ$as determined by the average value of the cosine of the local zenith distance of the Sun over a rotation period of the comet, ${<}\cos{z_{\odot}}{>}$, at different heliocentric distances are shown as solid and dashed lines respectively in the lower most panel. The activity of the latter source increased steeply during February 1997. The best fit to polarization vs. phase curves corresponds to a 15 fold increase in the relative number of Mg poor grains at the phase angle of $48^\circ$ compared to their composition at <$30 ^\circ$.

The proposed dichotomy may also explain the presence of the most volatile gaseous species to be more abundant at large heliocentric distances before perihelion reported by Biver et al. (1997). The stalling or reduction of production rates between 3 and 1.6 AU before perihelion reported by them coincides with the epoch when the solar illumination in the southern regions started decreasing rapidly. Wooden et al. (1999) note that the $9.3~\mu {\rm m}$ shoulder in the 7.5-13.4 IR spectra of the comet is present only in the images when the heliocentric distances were less than 1.2 AU. From Fig. 9 it is seen that the activity of the source at $+65^\circ$increased steeply at heliocentric distances closer than $\approx$1.2 AU, hence the extra feature at $9.3~\mu {\rm m}$ may have come from the northern region of the comet.