4.1. Uncertainty in the data and special conditions
The only parameters which enter the determination of the Zanstra temperature are the observed stellar magnitude, the observed H flux and the extinction. The latter is unimportant because the other two quantities are measured at almost the same wavelength. The H flux has been measured four times using diaphragms of about 40": the same flux was measured each time. Since a substantially brighter central star would have certainly been seen in the HST measurement, we conclude that no important uncertainties in the data exist.
Are there conditions which could trick us into thinking that the Zanstra temperature is anomanously high? One possibility is that the extinction of the starlight is much greater than the extinction of the nebular light. This could be done by having a small cloud of highly absorbing material around the star. This material should have the property that it lets ionizing radiation pass freely, so that the ionization of the nebula is not impeded. Alternatively this material could be in the form of a disk placed so that it absorbs starlight in our direction but allows it to pass freely in other directions, so that the nebula will be ionized. Such a condition is not impossible: witness the dark dust lane passing through the central regions of NGC 6302. But there is nothing to see of such anomalous extinction on the HST images of NGC 6537. Other conditions could exist as well. The star could emit radiation much different than a blackbody. But none of these conditions are likely enough to ignore the possibility that the star is indeed very hot.
4.2. The nature of the star
An upper limit to the radius of the star can be obtained, assuming it radiates as a blackbody of 500 000 K and has an mv=22.4. The distance to the nebula must also be known but is poorly determined. The best value is that of Gathier et al. (1986): 2.4 kpc. It is based on 21 cm absorption line measurements which show clear absorption feature from the local gas and from the Sagittarius arm as well. No trace of the Scutum arm is seen. The Sagittarius feature has the same optical depth as is seen in the spectrum of a very close background source. Gathier et al conclude that NGC 6537 is at the far side of the Sagittarius arm but in front of the Scutum arm. With this distance the radius of the star becomes 1. This is what would be expected from a 0.7 white dwarf carbon-oxygen core (Hamada and Salpeter,1961and Suh and Mathews, 2000). However the star probably has a substantial hydrogen atmosphere, making the carbon-oxygen core at least 10% and perhaps 40% smaller than the stellar radius. This would correspond to a star of 0.9 to 1.0. For example, the models of Blocker (1995) show a substantially bigger atmosphere for hot stars in this stage of stellar evolution. His hottest model, which has a core mass of 0.94 and reaches a temperature of 400 000 K, has a radius of on the cooling curve about a factor of 20 below maximum luminosity. This is almost 50% greater than the carbon-oxygen core of this mass. It is clear that we are dealing with an abnormally high core mass central star. This result is distance dependent: if the star is closer(as several recent statistical determinations indicate) its radius will be even smaller and the core mass even higher.
There is good agreement between the observations and the models in one respect. The models predict that the higher core mass central stars are able to reach a much higher surface temperature than the lower mass stars, and this is what we believe we see in NGC 6537. In another respect the models do not agree with the measurements in this nebula. All models predict that the evolution occurs more quickly as the core mass increases. In the 0.94 model of Blocker (1995) the evolution from the AGB to the white dwarf stage happens in less than 100 years. If the mass were higher the time would be shorter still. In NGC 6537 the evolutionary time can be estimated by noting the bright shell of material surrounding the central star at a distance of about 3 or 4" from it. If this has always moved at about 18km/s, which is the accepted expansion velocity of the nebula, then its age, measured from when the shell was expelled, is 2500 years. This value is similar to other young nebulae, and gives no evidence for exceptionally fast evolution. There is a photograph of the nebula taken about 85 years ago by Curtis(1918) for which a diameter of 5" is given. This is comparable to its present dimension considering the great difficulty of measuringdiameters on uncalibrated photographs. In addition, Curtis remarks that no central star was seen. If the central star had a temperature of 30 000 K 100 years ago, it would have been 10.1 magnitudes brighter in the visible if it followed the 0.94 model of Blocker (1995). It would thus have had a visual magnitude of 12.3. Such a bright star would have easily been seen by Curtis, who probably could have seen a star as faint as the 17th magnitude. It must be concluded that the evolution of the central star did not take place nearly as quickly as predicted by the 0.94 model of Blocker.
© European Southern Observatory (ESO) 2000
Online publication: October 24, 2000