2008A&A...486..545S

Strong mass loss off stars at the tip of the asymptotic giant branch (AGB) profoundly affects properties of these stars and their surroundings, including the subsequent planetary nebula (PN) stage. With this study we wanted to determine physical properties of mass loss by studying weakly emitting halos, focusing on objects in the galactic disk. Halos surround the, up to several thousand times, brighter central regions of PNe. Young halos, specifically, still contain information of the preceeding final mass loss stage on the AGB. In the observations we used the method of integral field spectroscopy with the PMAS instrument. This is the first committed study of halos of PNe that uses this technique. We improved our data analysis by a number of steps. In a study of the influence of scattered light we found that a moderate fraction of intensities in the inner halo originate in adjacent regions. As we combine line intensities of distant wavelengths, and because radial intensity gradients are steep, we corrected for effects of differential atmospheric refraction. In order to increase the signal-to-noise of weak emission lines we introduced a dedicated method to bin spectra of individual spatial elements. We also developed a general technique to part the temperature-sensitive oxygen line [OIII]λ4363 from the adjacent telluric mercury line Hgλ4358 - without using separate sky exposures. By these steps we avoided introducing errors of several thousand Kelvin to our temperature measurements in the halo. For IC3568 we detected a halo. For M2-2 we found a halo radius that is 2.5 times larger than reported earlier. We derived radially densely sampled temperature and density structures for four nebulae, which all extend from the central regions and out into the halo. NGC7662, IC3568, and NGC6826 show steep radially increasing temperatures and a hot halo, indicating that the gas in the halo is not in thermal equilibrium. M2-2 shows a larger temperature in the central region and an otherwise constant value. From the density structures we made estimates of core and halo masses and - for the first time reliable - mass loss rates at the tip of the AGB. All four objects show inwards radially increasing mass loss rate structures, which represent a rise by a factor of about 4-7, during the final mass loss phase, that covers a time period of approximately 10⁴years. Within a factor of two, the average of the maximum mass loss rates, which are distance dependent, is {dot}(M)_max≃10^–4M_☉/yr.