Global reanalysis products cannot reproduce seasonal and diurnal cycles of tropospheric ozone in the Congo Basin
Inês Vieira, Hans Verbeeck, Félicien Meunier, Marc Peaucelle, Thomas Sibret, Lodewijk Lefevre, Alexander W. Cheesman, Flossie Brown, Stephen Sitch, José Mbifo, Pascal Boeckx, Marijn Bauters
Abstract
Tropospheric ozone (O 3 ) is a secondary pollutant and a greenhouse gas with a positive radiative forcing . Many studies have documented its negative impacts on plant growth and human health. Historically, studies have focused on determining levels of exposure in mid- and high-latitude regions. In the tropics, high O 3 concentrations are expected due to large concurrent and future precursor emissions. In Africa, seasonal biomass burning (from both natural and anthropogenic fires) during the dry season plays a crucial role in O 3 precursor production. However, O 3 observational studies in tropical Africa are currently missing. To fill this major knowledge gap, we established in November 2019 a continuous monitoring of near-surface O 3 in the Congo Basin at the Yangambi research centre in the Democratic Republic of the Congo. Using this unique dataset in the heart of the second-largest tropical forest in the world, we assessed the ability of current remote sensing products to capture the magnitude and temporal dynamics of in situ tropospheric O 3 concentrations, especially O 3 concentration variation between dry and wet seasons until March of 2022. We compared near-surface atmospheric O 3 measurements collected in Yangambi and four different reanalysis products: European Centre for Medium-Range Weather Forecasts Reanalysis (ECMWF) v5 (ERA5), Copernicus Atmospheric Monitoring Service reanalysis (CAMSRA), Modern-Era Retrospective Analysis for Research and Applications version 2 (MERRA-2) and Japanese Reanalysis (JRA-55). The results show that reanalysis products overestimated the magnitude of near-surface O 3 across the region with a mean bias of 27.3 ppbv, 19.9 ppbv, 10.8 ppbv and 1.0 ppbv for ERA5, CAMSRA, MERRA-2 and JRA-55, respectively. ERA5 and CAMSRA reanalysis were the only products able to capture, to some extent, the observed annual variation, showing higher O 3 concentrations during dry season months, despite the inability to reproduce the daily cycle of near-surface O 3 .