Improved understanding of eutrophication trends, indicators and problem areas using machine learning

Abstract. Nitrate is an essential inorganic nutrient limiting phytoplankton growth in many marine environments. Eutrophication, often caused by nitrogen deposition, is a reoccurring problem in coastal regions, including the North-West European Shelf (NWES). Despite of their importance, nitrate observations on the NWES are difficult to obtain and thus sparse both in time and space. We demonstrate that machine learning (ML) can generate, from sparse observations, a skilled, gap-free, bi-decadal (1998–2020) surface nitrate data-set. We demonstrate that the effective resolution (scales on which the data-set is skilled) is slightly coarser than the 7 km and daily resolution of the product, but still completely sufficient to analyse nitrate dynamics on a monthly scale. With such a data-set we can address questions that would be otherwise hard to answer: (i) We show that nitrate-limited regions on the NWES, potentially vulnerable to eutrophication, extend beyond the eutrophication-problem areas already identified by the monitoring bodies (i.e. OSPAR). The newly identified regions include southern Irish coastline and parts of Irish Sea, indicating that these areas could become problematic under sub-optimal policy, or management changes. (ii) We demonstrate that bi-decadal 1998–2020 trends in coastal nitrate, responding to long-term policy-driven reduction in riverine discharge, are mostly modest with a notable exception of the Bay of Biscay. (iii) We show that winter nitrate plays relatively minor direct role in the phytoplankton bloom intensity the following spring, which can have some implications for using winter inorganic nitrogen as eutrophication indicator (as often included by OSPAR).