Oxygen is critical for sustaining all life on planet Earth and for regulating the biogeochemical cycles of carbon and major nutrients in the ocean. However, global oxygen concentrations in the ocean have been decreasing for the past six decades, with trends varying in time and space. Despite oxygen being the third most-measured physicochemical variable in the ocean, sparse in situ observations alone cannot be used to quantify temporal changes concentrations on regional and global scales. Conversely, ocean biogeochemical models, while available with global coverage, are unable to reproduce the observed spatial and temporal variability of O2 concentrations and air-sea O2 fluxes. 

Machine learning algorithms represent a powerful tool to quantify global ocean deoxygenation over time, as they allow to reconstruct global oxygen concentrations directly from observations. In my work, as part of the InMOS project, I am developing a new high-resolution gridded product of oxygen concentrations for the past 30 years. Here below is a little preview:

Lead (Pb) is a heavy metal, naturally occuring in the ocean at very low concentrations. Since the late 1800s, and especially in the last century, the natural biogeochemical cycle of Pb in the ocean has been severely perturbed by anthropogenic emissions generated by the use of leaded gasoline, waste incineration, coal combustion and smelting activities. Lead isotopes are a powerful tool that can be used to reveal the sources and sinks of Pb, as well as trace the movements of water masses in the ocean. 

As part of the international mearine geochemistry programme GEOTRACES, I have combined laboratory and computational techniques to investigate the distribution of Pb and its isotopes in the South Atlantic and Arctic Ocean.  My work revealed that Pb pollution in the surface South Atlantic decreased by more than 30% between the 1990s and 2011, and that, despite the stringet environmental policies in place, Pb of anthropogenic origin has reached the deep ocean aided by sinking particles. 

Marine debris, most of which consists of plastic items, poses significant threats to the environment. Plastic production rates have been increasing steadily in the past decades, and so have the negative consequences of pollution in the (marine) environment. 

In my research, I built numerical and statistical models to study the pathways of microplastics in the South Atlantic Ocean, and the distribution of litter items in the coastal areas of Australia. With the latter, I discovered that size and density of marine debris increase from the water line to the backshore vegetation, where larger items tend to accumulate. I also worked on the development of machine learning models to detect and identify plastic items floating on rivers, to prevent them from reaching the ocean.