SPECTROPOLARIMETRY OF THE BIPOLAR PLANETARY-NEBULA M2-9

We present optical spectropolarimetry of the young bipolar planetary n ebula M2-9. The goal of these observations is to determined the origin of the knots or brightness enhancements seen in the lobes of M2-9. Th e line spectra of the lobes of M2-9 are composed of two components, on e that is produced locally in the lobes and one that is scattered from deep in the nebula. The presence of this scattered radiation means th at the total flux line ratios do not accurately describe the local con ditions in the lobes. We have obtained spectropolarimetric data of the N2 and S2 knots and the adjacent nebula, and we use our data to separ ate the scattered and unscattered emission-line components. The spectr um of the core of M2-9 exhibits broad H alpha emission lines. In our h igh-resolution spectra we observe a broad wing on the scattered H alph a line profile at all of the positions in the north lobe. This confirm s that the scattered line emission originates in the core. In addition , we calculate the outflow velocity of the scatterers, approximate to 15 km s(-1), based on the observed wavelength shift between the H alph a peak in scattered and unscattered flux. Using the unscattered spectr a, we derive the local line ratios as a function of position in the no rth and south lobes. The degree of ionization of the spectra decreases in the off-knot regions. We measure the gas temperature as a function of position and find that it is approximately constant across the lob es. This result rules out the simple recombination tail model proposed by Goodrich for the origin of the knots. We suggest instead that the off-knot positions are ionized by a UV spectrum that is attenuated by material between the off-knot positions and the central UV source. We have used the photoionization code CLOUDY to test this idea and find t hat attenuation effects alone cannot accurately reproduce the observed unscattered line ratios. To accurately model the observed line ratios in the knots, we require the presence of both high (greater than or e qual to 10(5) cm(-3)) and low (approximate to 10(3) cm(-3)) density co mponents. For the off-knot positions both attenuation effects and mult iple density components are necessary to reproduce the observed line r atios. We modify the recombination tail model proposed by Goodrich to explain the origin of the knots in M2-9. We propose, as did Goodrich, that at discrete north-south levels, the lobes of M2-9 contain ''rings '' of higher density material. The UV source illuminates a section of the lobe, causing these rings of material to glow, producing the knots and explaining and naturally leading to their ''fixed'' north-south p ositions. Obscuring material near the star keeps one side of the lobe from being illuminated by the full force of UV source, thus explaining the lower excitation off-knot spectra. This model can explain why the knots appear to move only in the east-west plane, while remaining fix ed in the north-south direction. In addition, it explains why the obse rved spectra in the off-knot regions are lower excitation than the kno t spectra.