van Leeuwen et al. 
participating model). Both pecuniary and technical issues can result 
n gaps in geopolitical coverage of the ensemble result that can 
ainder international acceptance of derived policy products. 
However, the benefits of ensemble modelling are equally clear: 
‚.ncreased confidence in the results (due to the inclusion of different 
models with their own specific strengths), more insight into model 
dynamics and the opportunity for individual model development 
(based on the ensemble results and individual performance) and a 
nigher level of acceptance on the international (policy) stage 
compared to single model results. 
Thus, ensemble modelling is a suitable approach to help tackle a 
variety of ecological issues and their management in the marine 
environment. This could include dispersal of harmful dissolved 
substances, marine litter dispersion (by using particle tracking 
models), circulation pathways of pathogens (by using epidemiological 
bio-physical models) and impacts of these and other stressors on 
ecosystem services (coupled ecosystem models). Models are extremely 
zuited to test different policy options, quantify single and combined 
stressor impacts and predict future marine environmental conditions 
and their impact on anthropogenic derived usage. They can do this on 
5soth small (harbours, estuaries, bays) and large (basins, oceans) scales, 
providing a broad answer to marine ecosystem response that augments 
observational evidence and dedicated experimental work. It is therefore 
anticipated that ensemble modelling will be increasingly used in marine 
management issues. 
5 Conclusions 
This study presented a weighted ensemble modelling approach 
‚oO estimate the pre-eutrophic state of the marine ecosystem on the 
European Shelf. Eight modelling centers from countries around 
Europe participated with their most suited ecosystem model, 
:hough only seven delivered results on time. Inputs and boundary 
conditions were aligned as much as possible to focus on the models’ 
response to pre-industrial riverine and atmospheric nutrient levels. 
As expected, results showed lower nutrient concentrations in the 
pre-eutrophic state in most coastal areas, whereas offshore areas 
showed minimal change compared to the current state. DIN, DIP 
and Chl levels were at most 62%, -40% and „40% lower in the pre- 
eutrophic state than they are now, respectively, with most changes 
occurring in the southern North Sea, the Irish Sea and coastal Bay of 
Biscay areas. Net primary production was also lower in the historic 
scenario, with reductions up to -35% concentrated in the South- 
eastern North Sea and the Irish Sea. N:P ratio showed little change 
ın offshore areas, but strong changes in coastal areas, which moved 
closer to the Redfield ratio in the historic scenario. Pre-eutrophic 
results for near-bed oxygen levels showed improvements in known 
problem areas such as the Oyster Grounds. Overall, coastal areas 
show more sensitivity to DIP reductions than DIN reductions. 
The resulting concentration estimates for key eutrophication 
indicators like surface winter DIN, DIP and growing-season 
chlorophyll-a can be used as a basis for assessments as well as 
nolicy measures to combat marine eutrophication. It also illustrates 
he potential of modelling to support marine management. 
However, the weighted ensemble method relies on observations, 
"rontiers in Marıne © 
lc 4 
10.3389/fmars.2023.1129951 
and more and more spatio-temporally balanced observations are 
needed, particularly in offshore areas, to augment the applied 
weighting method and reduce uncertainty even further. As such, 
this work highlights the need for (more) extensive monitoring 
programmes. Models can help in this respect by optimizing 
existing and new observational efforts. While models are able to 
focus on local ecosystem functioning, they also consider the 
continuity of transboundary transport and processes across large 
areas. This model specificity is particularly useful in systems where 
data collection remains a challenge, such as the ocean. In that sense, 
nodels will continue to be useful for policy initiatives in coastal 
nanagement, and uptake by marine managers is encouraged. 
The ensemble approach presented here has demonstrated its 
use for policy purposes by defining a baseline for nutrient reduction 
measures; it may be useful for other environmental questions as 
well. For eutrophication modelling the next step should be to 
consider climate change impacts on the marine environment, and 
how these changes impact on derived thresholds for eutrophication 
indicators, both in the immediate and intermediate (policy) future. 
Data availlability statement 
The datasets presented in this study can be found in online 
repositories. The riverine input data used for all scenarios can be 
found here: https://doi.org/10.25850/nioz/7b.b.vc. The other open 
sources are mentioned in the manuscript. 
Author contributions 
Model simulations were performed by AB, LV, AvdL, CS, XD, 
GL, OK, IB, TS, RF and MP. All authors contributed to the 
ensemble methodology and conditions, which was led by H-JL. 
TP led the discussions with associated groups within OSPAR, while 
LF performed most of the work related to COMPEAT. SL provided 
the riverine data, collected the individual results and produced the 
final tables and most of figures, with additional analysis figures 
provided by XD and RF. All authors contributed to the article and 
approved the submitted version. 
-unding 
We would like to thank the Swedish Agency for Marine and 
Water Management for their support and financial contribution to 
this work. CS and RF were supported by the Umweltbundesamt 
(UBA, grant no. 3718252110 and 3720252020). Supercomputing 
power was provided to RF by HLRN (North-German 
Supercomputing Alliance) and to CS by Deutsches Klima- 
Rechenzentrum (DKRZ). OK was supported by the Deutsche 
Forschungsgemeinschaft (DFG, KE1970/2-1). MP acknowledges 
the Pöle de Calcul et de Donnees Marines (PCDM) for providing 
supercalculator DATARMOR {storage, data access, computational 
resources}. SL was supported by Rijkswaterstaat and NIOZ. 
Deltares was supported by Rijkswaterstaat and the European 
frontiersin.org