Our system employs a high-precision Blue Wave spectrometer by StellarNet, which is positioned to observe the sky. The spectrometer collects incoming light across various wavelengths, capturing the spectral signature of the atmosphere. The presence of pollutants in the air alters the way light is absorbed and scattered, resulting in distinct spectral patterns.

The raw spectral data is then processed using advanced algorithms. We analyse specific wavelength absorption features that correspond to pollutants such as PM2.5, NO₂, and SO₂. Each of these pollutants has a unique spectral signature that allows us to identify and quantify their concentration in the air. The collected spectral data is decomposed into its constituent wavelengths using Fourier Transform techniques. This helps isolate the specific absorption bands associated with pollutants. Environmental noise and natural atmospheric variations can affect readings. Our system applies baseline correction algorithms to ensure accuracy. The processed data is compared against reference spectra stored in our database. These reference spectra are obtained from controlled laboratory experiments and previous field studies to improve identification accuracy. Our system can analyse multiple pollutants simultaneously, detecting overlapping absorption features and differentiating them using statistical models. By continuously collecting spectral data, our system can track pollution fluctuations over time, identifying peak pollution hours and seasonal variations.