An interesting (fairly new) paper I came across:
http://www.mnras.org/content/442/3/2600.full
Some interesting quotes:
Quote:
|
It has always been appreciated that the traditional naked-eye magnitude limit of 6 is merely approximate. Weaver (1947) commented on the magnitude limits of nineteenth-century naked-eye star catalogues, which ranged from 5.7 (Argelander) to 6.7 (Heis). The latter observer was renowned for his visual acuity, and his Atlas Coelestis is unusual in including the galaxy M33 as a naked-eye object (Heis 1872). Gould's Uranometria Argentina had a stated magnitude limit of 7 but modern photometry has shown the actual limit to be 6.5 (Gould 2010). As an example of exceptional eyesight, Weaver cited Meesters’ ability to see stars to 6.9 mag. Weaver's study upheld a value of just over 6 for the typical dark-sky naked-eye limit, yet more recently there has been a substantial raising of achievement and expectation. The Bortle Scale (Bortle 2001) suggests that for a Class 1 (‘excellent’) site, the limiting magnitude should be ‘7.6–8.0 (with effort)’ and for Class 2 (‘typical truly dark site’) 7.1–7.5. Schaefer (1990) reported O'Meara's extraordinary ability to see stars as faint as 8.4 mag against the sky. Apart from unusual acuity or special observing techniques, such high limits may in many cases be explained by scintillation, with momentary glimpses being taken as typical threshold. Subjective estimates may not always be reliable or accurate; the survey by Schaefer (1990) yielded many responses (about half of the total) in which naked-eye limit was given only to the nearest whole number.
|
and...
Quote:
The International Dark-Sky Association (IDA 2013) currently recognizes three classes of dark sky: bronze (SQ 20.00–20.99), silver (SQ 21.00–21.74) and gold (SQ ≥ 21.75). A reading greater than 22 is ‘unlikely to be recorded’ (Unihedron 2014b). The suggested limiting magnitudes (based on the Bortle Scale) are bronze: 5.0–5.9, silver: 6.0–6.7, gold: ≥ 6.8. It is questionable whether 20 mag arcsec−2 can be considered dark, given the results shown in Fig. 18. Also, it has been argued here that a definition of naked-eye limiting magnitude based on momentary glimpses is overly susceptible to scintillation, which is a local and variable effect. Consequently, it has been suggested that currently recommended magnitude limits may be excessive, compared with limits that would be obtained for targets visible for an extended period. It has also been shown that the recommendations of the Bortle Scale with regard to the visibility of M33 are contradicted by the present model. Since the scale appears to be based on subjective judgment rather than rigorous data, its reliability appears questionable.
A practical definition of a dark sky would be one in which the Milky Way is capable of being seen. The non-uniformity of the Milky Way makes this problematic to model, and even if a particular region were chosen as standard, there remains the problem that existing luminance measurements are based on a surface-brightness limit fainter than that of the eye, and are usually filtered to remove bright stars, whereas unresolved stars just beyond the visual limit may contribute a significant proportion of the light detectable by eye. An equivalent limiting magnitude could be found empirically: observers would view the sky through a variable filter, adjusting it until a chosen portion of the Milky Way was considered just visible, and they would also note the faintest stars visible at this setting.
|
The whole article is worth reading. I always thought Bortle's NELMs were unrealistic and suggested visual 'trickery' in some form (scintillation, momentary glimpses, etc. i.e. not what a visual observer is really interested in).
From my personal experience, I could see mag 5.5 (with difficulty) from suburban Adelaide, and at my new dark-sky site in southern Tasmania, I have glimpsed mag 6.8 and can easily do 6.5 (the biggest difference is in the Milky Way and nebulosity, and sky background in photographs - chalk and cheese between the two sites!)
