Continuing Adventures in the Discovery of Multiple Star Systems with AI/ML
Description
Early data-driven analyses of ozone chemistry sensitivity primarily relied on "ratio-based" indicators to partially linearize the non-linear aspects of urban ozone chemistry, which are influenced by pollution levels, light, and water vapor. With the development of more sophisticated algorithms, including machine learning techniques capable of fitting high-dimensional non-linear functions, we have shown that a highly effective parameterization of net ozone production rates (PO3) can be achieved. This approach eliminates the need for empirical linearization of ozone chemistry through various indicators and allows for the primary inputs to be accurately constrained using satellite observations.
Detailed example
All outputs related to the project are well documented in our website: https://www.ozonerates.space/
AI / analytics pattern
Classical/Predictive Machine Learning: Models trained on data to make predictions or classifications based on identified patterns or relationships.
Automation level / stage
a) Pre-deployment – The use case is in a development or acquisition status.
Expected benefit
We need a better understanding of the worldwide spatiotemporal variability of ozone production rates. This is mainly due to the limited information we can gain from supersites or aircraft data by which we can generate observationally constrained PO3. However, we offer a significantly enhanced algorithm to parameterize PO3 using retrospective aircraft observations and a handful of variables that can be primarily informed by satellite observations that provide high spatial coverage. Our work shows the long-term maps of PO3 worldwide. This work has important implications for pollution exposure and regulations and can promote the use of satellite observations as an essential component for informing emission regulations beyond the current capabilities.
Controls / human review
ATO: Not reported; PIA: Not published