2. Observations and data reduction
Observations of comet Wirtanen were obtained on a total of six nights using either the 42-inch (1.1-m) Hall telescope or the 31-inch (0.8-m) telescope at Lowell Observatory. The single observational set from the 1991 apparition was measured 21 days prior to perihelion, when the comet was at a heliocentric distance r =1.12 AU. Our earliest observations during the 1997 apparition were obtained in mid-February, one month before perihelion, and our final observation was made in early July when Wirtanen was at a heliocentric distance of 1.72 AU. A conventional photoelectric photometer equipped with pulse-counting electronics was employed for all observations. Generally, a total of seven narrowband filters were used to isolate the emission from five gaseous species - OH, CN, C2, C3, and NH - and the continuum in the blue-green (4845 Å) and the near-UV (3650 Å). In June and July 1997, however, the comet was too faint to be observed with all of the filters. Most of the observations were obtained with the International Halley Watch (IHW) filter set (A'Hearn 1991); however, on 5 March 1997, one observational set was also obtained with the new Hale-Bopp (HB) filter set (Farnham et al. 1998), in which the continuum is measured in the near-UV (3448 Å), blue (4450 Å), and green (5260 Å). A typical observational set (see Table 1) consisted of several integrations totaling 60 seconds or more with each filter, while automatically tracking at the comet's rate of motion, with associated sky measurements of between 10 and 60 seconds taken at least 15 arcminutes away from the comet. Sufficient sky measurements were obtained to compensate for the effects of changing twilight. Wirtanen appeared very diffuse due to its relatively high gas-to-dust ratio, so large photometer entrance apertures were generally utilized to ensure that the comet was centered. Because we could not detect Wirtanen visually in June or July 1997, we pointed by means of a relative offset from the nearest PPM star, tracked at the comet's rate of motion, and used the largest available aperture. Tests of this offset pointing technique indicate that the resulting pointing accuracy should have been better than 10 arcseconds. This pointing uncertainty, combined with previous tests of the effects of pointing accuracy on resulting fluxes in several other comets, indicate that, except for the June data, uncertainties in our measurements due to centering inaccuracy are much lower than other uncertainties discussed in the next paragraph. Possible errors in centering in the July data are overwhelmed by the low contrast of the comet signal to the sky signal.
Table 1. Observing circumstances and photometric fluxes for Comet 46P/Wirtanen
Because of the unfavorable observing circumstances (high airmass, changing twilight sky, and/or bright moon on some nights), we performed an extensive analysis to determine the sources and the size of the uncertainty for each data point. Possible sources included the photon statistics of both the comet measurement and its corresponding sky measurement, the effects of changing sky brightness during twilight, and uncertainties associated with the determination of the underlying continuum for each emission band. The results of this analysis provided a qualitative understanding of the dominant uncertainties for each measurement, and allowed us to quantitatively incorporate each of the components in quadrature to produce a formal value of sigma. As it turns out, changes in sky due to twilight were determined sufficiently accurately in all cases that it was a minor component compared to other factors. More importantly, the sky often comprised a large fraction of the signal in the comet measurement, so we performed tests in which we changed the sky value by one sigma to investigate how uncertainties in the underlying sky value propagated through to affect the final result. Results of the overall analysis show that the errors are dominated by photon statistics in the measurement of the continuum in the comet and its associated sky value; an additional, sometimes dominant, uncertainty was introduced in the emission measurements by the effects of continuum subtraction.
Photometric reductions and subsequent standard modeling were performed following our normal procedures and using our current model parameters (cf. A'Hearn et al. 1995), and so will be only briefly summarized here. Numerous standard star measurements were obtained to determine nightly extinction and absolute flux calibration coefficients. After continuum subtraction, emission band fluxes were converted to molecular abundances and, with application of the Haser model, to production rates (Q ). Continuum fluxes were converted to a proxi for dust production, , the product of the albedo, A , at the observed phase angle, ; the filling factor, f , of the grains within the field-of-view; and the projected radius of the aperture, . No correction for phase angle was applied because the phase angles ranged from only 24 to 45 degrees, a region over which the phase function is expected to vary by less than about 30% (cf. Schleicher et al. 1998).
© European Southern Observatory (ESO) 1998
Online publication: June 18, 1998