The blue color of the sky is hard to remain unnoticed. What is less noticeable is the opportunity available if one knows how blue it numerically is. Since the color is determined by the air properties, these numbers are the keys to any abnormalities in the atmospheric composition. Our project facilitates quantitative analysis of the sky color in any location, based on both processing of the images of the sky, theoretical predictions and statistical analysis of the data shared between the users.

This project is solving the My Sky Color challenge.


Since our planet is surrounded by a layer of gases, the light we observe from the heavens differs from the light that is being radiated towards us from the Sun. The transformations underwent by light wavefronts as they penetrate the atmosphere strictly depend on the molecular geometry of the substances in the air and, hence, the structure of the atmosphere. Therefore, from the color of the sky, given that the original solar spectrum is known, it must be possible to deduce the composition of the atmosphere or, at least, detect an abnormality in it. Our project – ChromoSky – aims at development of a tool for a simple quantitative analysis of the sky color available to users all around the world. The project consists of two major parts: the processing desktop application and the website for data sharing, both of which are to be described below.

The processing application provides users with the ability to, firstly, theoretically predict the color of the sky they should observe if the atmospheric composition was perfect (i.e. followed the primary atmospheric standard). The calculation relies on the Rayleigh scattering model, NRLMSISE-00 atmospheric density model and basic geometry of sun rays, and it is done based on the time, date and location selected by the user. The details of the mathematics behind the project can be found in the attached PDF file, which includes all the necessary tabulated values used and equations.

Secondly, the processing application helps a user to remove any obstacles on the stellar pictures they have taken and analyze these pictures to compute the actual color values of the sky color observed. Finally, when both the theoretical and the experimental sky colors are computed, they are compared and outputted as a single number, quantifying their difference. This number can be later statistically correlated with the data of aerosol abundance from the NASA observatories and satellites.

The second part of the project is crucial to collect enough statistical data for this correlation. The processing application allows users to submit their data in one click to the processing website, which will accumulate all the data received based on the time and the location of measurements taken. The users are able to use the tools provided by the website to monitor how the aerosol abundance changes chronologically and dimensionally in both their observation site and worldwide.

Project Information

License: Academic Free License 3.0 (AFL-3.0)

Source Code/Project URL:


The Mathematical Model -


  • Antonios Papadopoulos
  • Roman Gerasimov