Ozone absorption enables the Green Flash
Note added on 13 July 2020: The upper range of ozone column densities that appear in these simulations extend beyond values that are expected on Earth but, without good observed spectra of the phenomenon, we cannot be sure of the relative contributions of refraction, scattering and absorption (by ozone, water and tetraoxygen) to the appearance of the green flash.
In April 2018, I posted three spectral plots along with an explanation of the essential role of ozone absorption in creating a spectral gap between the red and the green components of the image of the setting (or rising) Sun. See the comments below to find these posts.
I argued that an angular separation between the red and green images of the Solar limb resulting from the wavelength dependence of the atmospheric refraction was insufficient to generate the phenomenon of the green 'flash' where the green image becomes spatially distanced from that of the solar limb.
I also explained that under rare conditions of very low atmospheric aerosol content and a mirage-like choreography of atmospheric temperature gradients, the sunlight grazing the limb of the Earth at sunset/rise might result in a high effective pathlength through the ozone layer between altitudes of about 12 and 40 km.
Such a combination of high transparency and high ozone column density can contrive to shift the colour of the green flash towards the blue to create the very rarely-seen 'blue flash' See: www.eso.org/public/images/eso0812b/ for the wonderful photo by Guillaume Blanchard at the European Southern Observatory at Paranal in Chile. Note that this image was taken from Paranal at an altitude of close to 2,600m.
The animation shown here uses a model of atmospheric extinction (scattering + absorption) to illustrate the spectral behaviour under the conditions when a green flash might be expected. This is typically a view from a mountain over a clear ocean horizon when the prevailing wind over the ocean has a very low aerosol content.
The two relevant variables in the extinction model are the aerosol content and the assumed ozone column density. For the Rayleigh and aerosol scattering, I use the prescription in Astrophysical Quantities (Third Edition) by C. W. Allen, The Athlone Press, 1973. For the molecular absorption cross-sections, I use the relevant data from the Molecular Spectroscopy and Chemical Kinetics Group studies at the IUP, University of Bremen: www.iup.uni-bremen.de/gruppen/molspec/index.html
The plot shows the entire UV–near-infrared spectrum of the Sun from above the Earth's atmosphere (grey line) taken from the Hubble Space Telescope calibration database. The appearance of the Sun with its refracted centre on the horizon is shown as the blue line spectrum. For the purpose of this demo, the aerosol content has been taken to be just 10% of what is considered to be a normal clear atmosphere, ie. very clear. [Note that my extinction model does not include telluric absorption lines other than ozone, notably due to water, molecular oxygen and tetraoxygen. The collisionally induced absorption (CIA) from tetraoxygen does actually contribute to an increased absorption close to the centre of the Chappuis ozone absorption and so will enhance the effect discussed here.]
As shown in my previous descriptions, the dominant effect of the aerosol scattering is on the brightness of the green flash which even a normal aerosol content will render invisible.
The normal ozone content of the atmosphere would result in a layer of 3mm depth (0.3 atmo-cm) at sea level (standard temperature and pressure STP). The ozone is, however, not distributed uniformly through the atmosphere but restricted to a range of altitudes as mentioned above.
This means that the usual method of calculating the atmospheric path traversed by sunlight at different altitudes is not a good representation of the ozone column density when the Sun is on the horizon since the atmospheric refraction, driven by the presence of layers of air at different temperatures, can guide the sunlight along different paths which can result in longer than normal traverses within the ozone layer.
The animation shown here illustrates the effect of increasing the effective ozone content of the sunlight on the horizon from zero by steps corresponding to the standard 0.3 atmo-cm to ten times that amount. In practice, this corresponds to paths that increasingly favour passages within the ozone layer, the higher values being somewhat beyond realistic values.
As the animation proceeds, notice first the appearance of a dip at the centre of the Chappuis absorption band of ozone in the orange part of the spectrum. As the ozone increases, this band deepens to result in a clear separation of the red and green light from the Sun. The differential atmospheric refraction can then result a clear spatial separation of the green from the red/orange of the rest of the Sun.
But notice also that the increasing absorption results in the green band moving significantly towards the blue which can explain the rare blue flash phenomenon.
The logarithmic energy scale on the vertical axis indicates how faint the green flash will be compared with full direct sunlight from high in the sky.
The refractive separation of the green/blue flash from the red part of the solar image has been nicely simulated as the result mirage effects by Andy Young: aty.sdsu.edu/explain/simulations/simintro.html#inf-mir
who also proposes that the transition from a green to a blue flash can be the result of a very low aerosol content, See aty.sdsu.edu/explain/simulations/inf-mir/colors/GFcolors....
Young points out, quite correctly, that it is difficult to understand how to achieve a high enough ozone column density in the sunlight path to achieve a sufficient level of ozone absorption to produce the strength of the effect that I illustrate in these simulations. While this may be possible for observations made at altitude, it appears unlikely for observations from the beach! Until it is possible to examine good spectroscopic observations of this phenomenon, it will be difficult to really assess the relative contributions of refraction, scattering and absorption to this complex and fascinating phenomenon.
I thank Guillaume Blanchard for his superb blue flash photograph and John Law for constructing this animation from the plots coming from my atmospheric extinction model. I also thank Andy Young for very helpful comments and information.
Ozone absorption enables the Green Flash
Note added on 13 July 2020: The upper range of ozone column densities that appear in these simulations extend beyond values that are expected on Earth but, without good observed spectra of the phenomenon, we cannot be sure of the relative contributions of refraction, scattering and absorption (by ozone, water and tetraoxygen) to the appearance of the green flash.
In April 2018, I posted three spectral plots along with an explanation of the essential role of ozone absorption in creating a spectral gap between the red and the green components of the image of the setting (or rising) Sun. See the comments below to find these posts.
I argued that an angular separation between the red and green images of the Solar limb resulting from the wavelength dependence of the atmospheric refraction was insufficient to generate the phenomenon of the green 'flash' where the green image becomes spatially distanced from that of the solar limb.
I also explained that under rare conditions of very low atmospheric aerosol content and a mirage-like choreography of atmospheric temperature gradients, the sunlight grazing the limb of the Earth at sunset/rise might result in a high effective pathlength through the ozone layer between altitudes of about 12 and 40 km.
Such a combination of high transparency and high ozone column density can contrive to shift the colour of the green flash towards the blue to create the very rarely-seen 'blue flash' See: www.eso.org/public/images/eso0812b/ for the wonderful photo by Guillaume Blanchard at the European Southern Observatory at Paranal in Chile. Note that this image was taken from Paranal at an altitude of close to 2,600m.
The animation shown here uses a model of atmospheric extinction (scattering + absorption) to illustrate the spectral behaviour under the conditions when a green flash might be expected. This is typically a view from a mountain over a clear ocean horizon when the prevailing wind over the ocean has a very low aerosol content.
The two relevant variables in the extinction model are the aerosol content and the assumed ozone column density. For the Rayleigh and aerosol scattering, I use the prescription in Astrophysical Quantities (Third Edition) by C. W. Allen, The Athlone Press, 1973. For the molecular absorption cross-sections, I use the relevant data from the Molecular Spectroscopy and Chemical Kinetics Group studies at the IUP, University of Bremen: www.iup.uni-bremen.de/gruppen/molspec/index.html
The plot shows the entire UV–near-infrared spectrum of the Sun from above the Earth's atmosphere (grey line) taken from the Hubble Space Telescope calibration database. The appearance of the Sun with its refracted centre on the horizon is shown as the blue line spectrum. For the purpose of this demo, the aerosol content has been taken to be just 10% of what is considered to be a normal clear atmosphere, ie. very clear. [Note that my extinction model does not include telluric absorption lines other than ozone, notably due to water, molecular oxygen and tetraoxygen. The collisionally induced absorption (CIA) from tetraoxygen does actually contribute to an increased absorption close to the centre of the Chappuis ozone absorption and so will enhance the effect discussed here.]
As shown in my previous descriptions, the dominant effect of the aerosol scattering is on the brightness of the green flash which even a normal aerosol content will render invisible.
The normal ozone content of the atmosphere would result in a layer of 3mm depth (0.3 atmo-cm) at sea level (standard temperature and pressure STP). The ozone is, however, not distributed uniformly through the atmosphere but restricted to a range of altitudes as mentioned above.
This means that the usual method of calculating the atmospheric path traversed by sunlight at different altitudes is not a good representation of the ozone column density when the Sun is on the horizon since the atmospheric refraction, driven by the presence of layers of air at different temperatures, can guide the sunlight along different paths which can result in longer than normal traverses within the ozone layer.
The animation shown here illustrates the effect of increasing the effective ozone content of the sunlight on the horizon from zero by steps corresponding to the standard 0.3 atmo-cm to ten times that amount. In practice, this corresponds to paths that increasingly favour passages within the ozone layer, the higher values being somewhat beyond realistic values.
As the animation proceeds, notice first the appearance of a dip at the centre of the Chappuis absorption band of ozone in the orange part of the spectrum. As the ozone increases, this band deepens to result in a clear separation of the red and green light from the Sun. The differential atmospheric refraction can then result a clear spatial separation of the green from the red/orange of the rest of the Sun.
But notice also that the increasing absorption results in the green band moving significantly towards the blue which can explain the rare blue flash phenomenon.
The logarithmic energy scale on the vertical axis indicates how faint the green flash will be compared with full direct sunlight from high in the sky.
The refractive separation of the green/blue flash from the red part of the solar image has been nicely simulated as the result mirage effects by Andy Young: aty.sdsu.edu/explain/simulations/simintro.html#inf-mir
who also proposes that the transition from a green to a blue flash can be the result of a very low aerosol content, See aty.sdsu.edu/explain/simulations/inf-mir/colors/GFcolors....
Young points out, quite correctly, that it is difficult to understand how to achieve a high enough ozone column density in the sunlight path to achieve a sufficient level of ozone absorption to produce the strength of the effect that I illustrate in these simulations. While this may be possible for observations made at altitude, it appears unlikely for observations from the beach! Until it is possible to examine good spectroscopic observations of this phenomenon, it will be difficult to really assess the relative contributions of refraction, scattering and absorption to this complex and fascinating phenomenon.
I thank Guillaume Blanchard for his superb blue flash photograph and John Law for constructing this animation from the plots coming from my atmospheric extinction model. I also thank Andy Young for very helpful comments and information.