Thoughts on this (albeit old) webpage / article - http://alperen.cepmuvakkit.com/alperen/makale/index.htm#Namaz - mainly points below: It has been recently debated, especially in England and U.S.A., whether the true dawn is valid at -18°. It has been asserted that the value of -18° is the boundary for astronomical twilight and it is not conforming to the definition of the true dawn which requires the horizontal spread. The observation of the zodiacal light by the Belgian astronomer Marcel Minnaert was around a depression angle of -17°, and he could distinguish the onset of twilight at -16/17° using photometric instruments. Dr. Mohammad Ilyas found similar results and he could only detect small changes in sky illumination after -16° with optical tools. A study in 2005 at ESO-Paranal Observatory (Chile) has akin outcomes (Figure 9). The graph shows the zenith brightness in mag arcsec-2. X-axis is the Sun's zenith angle in degrees, so it must be subtracted from 90° to convert it into elevation angle. The measured data shows that the twilight brightness diminishes down to the night-sky illumination and then stays flat after around -16°. The solid line represents the simple-scattering model, which clearly knees at -10°, implying the end of red dusk. Following the results of such observations, some scholars accepted a vertical angle of -15° both for Fajr and Isha. To overcome this difficulty about the Fajr/Isha timings, some researchers attempt to perform long-term observations. One of the most comprehensive observations was started in 1987 September in Blackburn/Lancashire (England) and lasted near one year with the attendance of the religious scholars. Some interesting findings were obtained at the end of this observation: Although dawn and dusk are two symmetrical phenomena astronomically, it has been perceived that the relevant depression angles are not always identical, which is commented as the atmospheric conditions are not the same for dawn and dusk. In fact, the detection of the first illumination on the sky by an eye adapted to darkness is easier than the detection of the last illumination by an eye adapted to brightness. Besides, the humidity, fog and especially temperature levels are rather different at dawn compared to dusk. Yet another factor is that the dawn and dusk occur at opposite locations on the horizon with unlike air and surface layers. Observed angle values for Fajr/Isha are not constant throughout the year but are seasonally changing. Values increase a bit in winter for example, which can be explained as the higher refraction of the light by the colder (and denser) atmosphere. So it was deducted that the angle values will vary in different locations with distinct climate. The angles were altering within one year between -12°/-16° for Fajr (tabayyun) and -9°/-11° for Isha (red dusk). Regarding all these researches, we conclude that it is not possible to assign a constant and reliable vertical angle for Fajr/Isha, they change in time and location. Although observation is essential, it is not generally possible especially under city illumination. So we suggest to allow a gray zone of angle for Fajr as well as Isha, say -18°/-15° for the onset of true dawn, -16°/-12° for tabayyun, -9°/-12° for the red dusk and -15°/-17° for white dusk. Ideally we should refrain from these zones of uncertainty as much as possible (for example to start fasting at -18° but perform Fajr prayer at -12° or perform Maghrib prayer before -9° but perform Isha prayer after -17°, whereby conforming to sunnah). Nevertheless, this ambiguity should be considered as a blessing not a burden; in any case of difficulty, we believe that we can use these zones, similar to the case of Early/Late Asr.