Astron. Astrophys. 338, 340-352 (1998)

4. Observational photometric data

Our data sets contain thermal emission observations from 7 to , taken from 1983 to 1997. We have individual photometric points as well as multi-filter photometry and thermal lightcurve measurements. The observation geometries include phase angles from to , with negative angles related to after opposition and positive angles related to before opposition. We converted the available filter photometry to monochromatic fluxes at the effective or the isophotal wavelengths (Golay 1974). This allowed us to compile data from a number of references in a uniform format. Each entry i in the database consists of:

where .

The database contained 678 individual observations in December 1997, and because of its size it is only available in electronic format ⁷. A quantitative summary of all measurements is given in Table 2. In the following we describe the data sets per instrument or telescope.

Table 2. Data coverage per instrument and asteroid . Summary of the number of data points selected (Sect. 5.3) in the analysis, from the data set available in December 1997.

4.1. IRAS data

In 1983 IRAS ⁸ surveyed the sky in 4 wavelength bands centred at 12, 25, 60 and (Neugebauer et al. 1984). Photometry for more than 1 800 asteroids has been obtained during the mission time. A description of the data products can be found in The IRAS Minor Planet Survey (IMPS) (Tedesco 1992). Starting from the IMPS table No. 108, we applied colour corrections to obtain monochromatic fluxes at 12, 25, 60 and . In cases where the was lower than 10 a flux overestimation correction (Tedesco et al. 1992) has been applied. We ignored the Band-to-Band corrections. They have been introduced by Tedesco et al. (1992) in order to bring the individual diameters and albedos, derived with the STM for each band flux, into agreement. According to the IMPS the measurement errors are between 10 and . Some Ceres observations are already in the non-linearity range of the detectors and therefore questionable, we excluded them. The IRAS Explanatory Supplement (Beichmann et al. 1988) gives an additional absolute band uncertainty of (), (), () and (). Cohen et al. (1996) conclude from a statistical comparison of IRAS photometry with models for 12 bright standard stars that, at and , flux densities measured by IRAS should be revised downwards by about and , respectively. We root-sum-squared the stated errors.

4.2. JCMT data

All JCMT observations have been obtained during several observing runs with the UKT14 ⁹-Bolometer (Duncan et al. 1990) between September 1989 to April 1996. The last campaign in April 1996 took place in support for the ISO mission (D. Hughes, J. Stevens, priv. communication). All data are available from the JCMT archive (http://cadcwww.dao.nrc.ca/jcmt). The standard data reduction is described in Redman et al. (1992) or in Emerson (1994). In principle the measured signal is multiplied by the flux conversion factor, which is based on calibration measurements. The extinction can be obtained from the -meter at CSO ¹⁰. Colour corrections are not relevant. The overall uncertainty is derived from the -ratio, the data reduction process and the absolute errors of the submillimetre calibrators. Typical total uncertainties are in the order of 10 to .

4.3. UKIRT mid-IR N and Q data

During the ISO mission several observing campaigns have been performed at the UKIRT with the Si:As-BIB ¹¹ MAX ¹²-Camera from the MPIA ¹³ (http://www.mpia-hd.mpg.de/MPIA/Projects/IRCAM/MAX/index.html). We obtained N and Q band photometry in chopping-nodding mode for all 10 asteroids and parts of thermal lightcurves for selected targets. The definition of the photometric system, including instrumental and celestial calibration, has been established in collaboration with M. Cohen, Berkeley (private communication, 1997). A full description of the data reduction process can be found in Müller (1997), including passbands, calibration stars and error calculation. The highest quality N and Q band results have a relative uncertainty of less than , the absolute total errors are at 8 to .

4.4. IRTF mid-IR N and Q data

The STM was calibrated by Lebofsky et al. (1986), based on K, M, N and Q band measurements on Ceres and Pallas. Here we consider only the N and Q band data, where the flux contribution from reflected light is negligible. We applied the given colour correction factors to obtain monochromatic fluxes at the central filter wavelengths. The stated accuracy of in Lebofsky et al. (1986) seems to be too optimistic considering the fact that the absolute photometric system at N and Q is only known to (Rieke et al. 1985). Adding a photometric system offset of (Hammersley et al. 1998) and typical atmospheric variations during the nights, we ended up at around uncertainty for the highest quality data. This is roughly equivalent with increasing all published uncertainties by a factor of 3.

4.5. ISOPHOT

ISOPHOT performed observations of photometric standards on a weekly basis to calibrate the internal flux reference (FCS ¹⁴). At wavelengths beyond it is necessary to observe stars, asteroids and planets to cover the full dynamic range of the detectors. The selected 10 asteroids provide the intermediate flux interval (see Fig. 1). It is possible to deselect all asteroid observations from this calibration program and base the determination of the FCS at high flux levels only on the planets and at low flux levels only on stars. The asteroid measurements can then be treated as independent scientific observations (Müller 1997). The bright ones can be compared against the planets Uranus and Neptune, whereas the faint ones are close to flux densities of the brightest standard stars. The internal reference source FCS allows a comparison of observations, taken at different epochs. The observing mode of centring the source on each detector pixel provides in case of C100 (9 pixels) and C200 (4 pixels) several independent measurements of the same object within a few minutes. The observing sequence is accompanied by background and dark current measurements, leading to reliable asteroid fluxes with an uncertainty of . All measurements are exclusively calibrated against well known standard stars and planets. Examples for this method and a discussion of the possibilities, as well as the limitations can be found in Müller (1997). In total we derived 35 individual photometric data points between 50 and .

4.6. Other data

Altenhoff et al. (1996) measured the intensity ratio of the planet Mars and the asteroid Ceres at a frequency of with a wideband bolometer at the HHT ¹⁵ (Kreysa 1990; Baars & Martin 1990; Martin & Baars 1990). From 58 pairs of observations they derived at a flux density ratio of Mars and Ceres of 268.5, corrected for the partial resolution of Mars. The brightness variation with the rotation phase was smaller than . They claim that their obtained flux density has an accuracy of better than , but adding the Mars uncertainties of we get a total error of for those Ceres observations.

© European Southern Observatory (ESO) 1998

Online publication: September 8, 1998
helpdesk.link@springer.de