New paper published! – OMI aerosol height over land from the O2-O2 visible absorption band, based on neural networks

Retrieving aerosol height from passive satellite sensors providing then daily global observations of our atmosphere is a very important issue and a burning challenge. This is very important for air quality and climate purposes, but also for better correcting the aerosol scattering effects present in the spectra measurements when quantifying the atmospheric trace gas amounts (e.g. NO2, CH4, CO2, etc…).

For the first time, we have developed an algorithm devoted to the OMI 477 nm O2-O2 measurements and applied it over land surfaces. The employed approach relies on machine learning (cf. neural network) approach. The obtained product shows encouraging results with uncertainty below 800 m.

These results are now published in the Atmospheric Measurement Techniques (AMT) peer-review journal.

Great thanks to all the co-authors KNMI and the GRS research department for their support on this very nice study!

More details can be read here, and paper available here.

For sure, more results to come in a near future…