Digital image processing for atmospheric monitoring at Colombian Andes

Abstract

ABSTRACT As an alternative to the current technologies, we explored the feasibility of using low cost and massive use of digital cameras as photometric sensors to retrieve the atmospheric total optical depth (τ) in the urban area of a city in the Colombian Andes. This study proposes a simple way to estimate τ from digital processing of images of the Sun based on the Beer-Bouguer-Lambert law Langley’s linear fitting for the colour levels in channels red, green, and blue registered by the pixels of cameras’ sensors. From February to March 2022, the τ values retrieved from the images were correlated to the retrieved values from a solar spectral radiometer (SSR). We found that τ is sensible to the featured changes in the local atmosphere and to the cameras’ exposure parameters setup. Under conditions of partly clear sky, around 80% (r > 0.8) of the τ values from cameras showed a linear correspondence to those retrieved from SSR system. Its spectral dependency (τ _red < τ _green < τ _blue) is in accordance with the physical phenomena in light-atmosphere interaction. The results suggest that the methodology applied can be used for monitoring the atmosphere at any geographical location in the world.