ALBERT

Description: Sediment fluxes to the seafloor govern the fate of elements and compounds in the ocean and serve as a prerequisite for research on elemental cycling, benthic processes and sediment management strategies. To quantify these fluxes over seafloor areas, it is necessary to scale up sediment mass accumulation rates (MAR) obtained from multiple sample stations. Conventional methods for spatial upscaling involve averaging of data or spatial interpolation. However, these approaches may not be sufficiently precise to account for spatial variations of MAR, leading to poorly constrained regional sediment budgets. Here, we utilize a machine learning approach to scale up porosity and 210 Pb data from 145 and 65 stations, respectively, in the Skagerrak. The models predict the spatial distributions by considering several predictor variables that are assumed to control porosity and 210 Pb rain rates. The spatial distribution of MAR is based on the predicted porosity and existing sedimentation rate data. Our findings reveal highest MAR and 210 Pb rain rates to occur in two parallel belt structures that align with the general circulation pattern in the Skagerrak. While high 210 Pb rain rates occur in intermediate water depths, the belt of high MAR is situated closer to the coastlines due to lower porosities at shallow water depths. Based on the spatial distributions, we calculate a total MAR of 34.7 Mt yr -1 and a 210 Pb rain rate of 4.7 · 10 14 dpm yr -1 . By comparing atmospheric to total 210 Pb rain rates, we further estimate that 24% of the 210 Pb originates from the local atmospheric input, with the remaining 76% being transported laterally into the Skagerrak. The updated MAR in the Skagerrak is combined with literature data on other major sediment sources and sinks to present a tentative sediment budget for the North Sea, which reveals an imbalance with sediment outputs exceeding the inputs. Substantial uncertainties in the revised Skagerrak MAR and the literature data might close this imbalance. However, we further hypothesize that previous estimates of suspended sediment inputs into the North Sea might have been underestimated, considering recently revised and elevated estimates on coastal erosion rates in the surrounding region of the North Sea.