The hydroxyl radical (OH) dictates the lifetime of many tropospheric species, including carbon monoxide (CO), methane (CH4), and numerous volatile organic compounds (VOCs). Knowledge of OH is therefore necessary to understand the abundance, distribution, and variability of these species.
Combining machine learning, chemical transport model (CTM) output, and satellite data has the potential to constrain tropospheric column OH (TCOH). A variety of machine learning techniques, such as neural networks, self-organizing maps, random forest regression, and gradient boosted regression trees (GBRTs), show promise in helping to solve problems in atmospheric chemistry. GBRT models use an ensemble of decision trees to predict the value of a target based on multiple inputs, even for targets with highly non-linear dependencies on the inputs.
We present a methodology to infer tropospheric column OH (TCOH) in the tropics over the open oceans using a combination of a machine learning model, output from a simulation of the GEOS model, and satellite observations, including from the NASA EOS flagship missions, Terra, Aqua, and Aura. Our overall goals are to assess the feasibility of our methodology, to identify potential limitations, and to suggest areas of improvement in the current observational network.
The methodology reproduces the variability of TCOH from independent 3D model output and of observations from the Atmospheric Tomography mission (ATom). With slight modifications, a similar methodology could likely be expanded to the extratropics and over land, with the benefits of increasing our understanding of the atmospheric oxidation capacity and, for instance, informing understanding of recent CH4 trends.
Tropospheric Column Hydroxyl Radical (TCOH) calculated with the machine learning model using satellite inputs from EOS Terra (carbon monoxide), Aqua (water vapor, aerosol optical depth), and Aura (total ozone, nitrogen dioxide, formaldehyde) for the months of each ATom deployment: February 2017 (a), May 2018 (b), August 2016 (c), and October 2017 (d). The mean, domain-wide TCOH value in molec. cm−2 for each month is also indicated.
References:
Daniel C. Anderson, Bryan N. Duncan, Julie M. Nicely, Junhua Liu, Sarah A. Strode, and Melanie B. Follette-Cook
https://doi.org/10.5194/acp-23-6319-2023
07.2023
