Amazon river runoff contribution to regional ocean circulation changes over 1993-2015 in the Caribbean Sea.
Supervisor: William Llovel (IFREMER)
In addition of contributing the most to the world’s global riverine discharge to the oceans, the Amazon River discharge experiences a large interannual variability in the same order of the time mean discharge. Yet, in Ocean General Circulation Models (OGCMs) riverine freshwater input is forced with climatological signals only. Recent studies showed that the Amazon River plume influences the circulation in the Tropical North Atlantic and the Caribbean Sea. Here we perform a sensitivity analysis using a modelling approach based on a regional configuration of NEMO to investigate the contribution of Amazon River discharge variability to regional ocean circulation changes in the Caribbean Sea over the period 1993-2015. The impact of river discharge interannual variability on sea level and circulation remains largely unknown. This study is unique in that it uses interannual daily runoffs from the most recent ISBA-CTRIP land surface model which was validated with in situ data from Obidos station. We studied the Amazon discharge variability impact on the Atlantic Inflow to the Caribbean Sea before focusing on its impacts on the Caribbean circulation using 3 sections following satellite tracks from Topex/Poseidon and Jason constellations spanning the Caribbean Sea from east to west.

The presence of the Amazon discharge itself, regardless of its variability, reduces the Atlantic inflow while increasing its variability into the Caribbean Sea. The reduction mainly occurs at Windward Islands passages south of 15°N corresponding to water from South Atlantic origin. It is most probably due to the increased transport to the east by the North Equatorial Counter Current in the presence of Amazon discharge. The eastern and western Caribbean basin circulation follow two different regimes. While the eastern basin is mainly sensitive to the presence of the Amazon discharge, the western basin reacts strongly to the Amazon discharge variability. It suggests that interannual current variability in the eastern basin is controlled by atmospheric forcing, most likely wind stress. The presence of Amazon discharge is also the main sensitive factor of mean dynamic topography and time-mean vertical velocity structure over 1993-2015. Regarding the contribution of Amazon discharge interannual variability, the missing non-seasonal Amazon discharge variability in OGCMs yields significantly different punctual transports of the order of several Sverdrups. It also modifies the volume transport temporal variability especially at annual and interannual frequencies, and even the phase of these transports in the western basin. Interannual river discharge variability plays a substantial role punctually, even though it is not significant when considering time-mean circulation. This suggests the incorporation of this temporal variability in models to help understanding and predicting climate change and extreme events occurrence.