Interactions between the biosphere and the atmosphere are an important controlling factor for regional to global atmospheric chemistry and composition. This ultimately has wide impacts on the modern environmental challenges of air quality and climate change. However, there are still substantial uncertainties in the biosphere-atmosphere interaction processes that drive the global abundance and variability of many critically important atmospheric constituents, including ozone, aerosol, and Volatile Organic Compounds (VOCs). New methods from the data science and artificial intelligence (A.I.) literature, when informed and guided by scientific understanding, present a valuable tool in addressing these knowledge gaps. In this seminar, I will present results from recent work using a variety of A.I. methods to better constrain biosphere-atmosphere interactions as they are relevant to atmospheric chemistry and composition. Specifically, we applied a set of tools known as deep neural networks to develop an improved data-driven model for the interactions between ozone and the plant biosphere and used a statistical learning variable selection approach to enable a detailed representation of canopy physics within global models of atmospheric chemistry.
Content coming soon!