Trends and Variability of the Hydroxyl Radical in the Tropics Determined from Satellite Observations of its Drivers

Hydroxyl (OH) is the dominant tropospheric oxidant and is responsible for setting the lifetime of many atmospheric constituents, including methane, a potent anthropogenic greenhouse gas. Despite this importance, observational constraints on OH are limited, providing little information on spatial variability at resolutions finer than the hemispheric scale. Here, we present an observationally based tropospheric column OH (TCOH) product determined from a gradient boosted regression tree (GBRT) model constrained with space-based observations of chemical and dynamical drivers of OH. We illustrate the ability of the methodology to reproduce observed OH from the Atmospheric Tomography mission within observational uncertainty. We further demonstrate the suitability of multiple retrievals of OH precursors from the Aura and Terra satellites for use in the product. Using 15 years of this satellite-constrained OH product, we then examine trends and variability in tropical TCOH, finding large interannual variability but little evidence for large trends across the domain. Finally, we highlight the importance of changes in NO₂ abundance as well as the El Nino Southern Oscillation on the observed OH variability. The results presented here demonstrate the viability of this methodology for constraining tropospheric OH and provide a potential new means to constrain OH in global chemistry models.