Preliminary assessment of the marine extreme indicators (Milestone MS18)

Milestone MS18 Preliminary assessment of the marine extreme indicators MS18 Preliminary assessment of the marine extreme indicators MILESTONE Milestone MS18 Preliminary assessment of the marine extreme indicators Work Package WP5 EOVs and ECVs in support of climate modelling Due Date Delivery to EC: 30 January 2026 Submission Date 11 December 2025 Milestone Lead Giulia Bonino <[email protected]> (CMCC) Author(s) Giulia Bonino (CMCC) Francesco De Rovere (CMCC) Simona Masina (CMCC) Sourav Chatterjee (NCPOR) Roshin Raj (NERSC) Emmanuel Eresanya (NUIM) André Düsterhus (DMI) Veera Haapaniemi (FMI) Reviewer Chiara Bearzotti (DMI) DISCLAIMER Funded by the European Union. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the European Research Executive Agency (REA). Neither the European Union nor the granting authority can be held responsible for them . 2 MS18 Preliminary assessment of the marine extreme indicators TABLE OF CONTENTS 1. What is this milestone about 4 2. Milestone report 4 2.1 Means of verification 4 2.2 Work performed 4 2.2.1 Marine heatwaves 5 2.2.2 Sea level extremes 7 2.2.3 Sea ice extremes 8 3. Contribution to the ObsSea4Clim objectives 9 4. References 9 3 MS18 Preliminary assessment of the marine extreme indicators 1. What is this milestone about This milestone provides an update on the work carried out so far in the assessment of the marine extreme indicators within Task 5.3 “Skill assessment of climate-relevant marine extreme indicators from ESMs”. The indicators considered were those proposed in MS5. This milestone, together with the deliverables D4.1 “Sea ice Essential Ocean Variables uncertainties”, D4.4 “Sea Level Assessment” and D4.5 “Marine Heat Waves Assessment”, contributes to refining and updating the definition of the marine extreme indicators, which will be consolidated in their final form in Deliverable D3.2 “Extreme indicators”. All the results from these reports will also be summarised in the final deliverable of task 5.3, the D.5.3 “Marine extreme indicators in ESMs”. 2. Milestone report 2.1 Means of verification According to the Description of the Action, this milestone is achieved when the following is in place: “A preliminary assessment to feed into the requirements (feeding into Milestone MS5 and deliverables D4.1, D4.4 and D4.5) is available”. The preliminary assessment of marine extreme indicators, intended to feed into the requirements, such as Milestone MS5 “Identification of Extreme Indicators for Task 3.2”, Deliverable D4.1 “Sea ice Essential Ocean Variables uncertainties, D4.4 Sea Level Assessment, D4.5 “Marine Heat Waves Assessment”, and D5.3 “Marine extreme indicators”, is now ready and available to the partners. 2.2 Work performed This milestone presents the progress achieved in the regional assessment of the marine extreme indicators identified in Milestone MS5. The assessment covers the following three main types of extremes: ● Marine Heat Waves (MHWs) have been analysed in the Nordic and Barents Seas as well as in the Mediterranean Sea. 4 MS18 Preliminary assessment of the marine extreme indicators ● Sea level extremes have been assessed in the Nordic and Barents Seas and over the European Shelf. ● Sea ice extremes have been evaluated in the Arctic region. The following paragraphs give an update on the work performed on the specific extreme indicators. 2.2.1 Marine heatwaves The work has focused on evaluating the capability of different climate model configurations to reproduce observed MHW characteristics. Based on the definitions provided by Task 3.2 in Milestone MS5 (1), daily temperature fields have been analysed to detect MHWs, following Hobday et al. (2016), using both moving and fixed baselines to explore methodological sensitivities. In the Mediterranean Sea, the performance of Med-CORDEX (Ruti et al., 2016; Somot et al., 2018) Regional Climate System Models (RCSMs) has been evaluated against their driving Global Climate Models (GCMs) in reproducing observed MHW properties over the period 1982–2020, using historical simulations (1982–2005 and 1982–2014) and the early years from scenario experiments. Overall, GCMs tend to overestimate MHWs' duration and underestimate their intensity, while RCSMs show a clear improvement in the simulation of duration but exhibit less consistent behaviour in intensity. Most RCSMs improve the representation of MHW duration, mainly by better simulating short-lived events. However, for intensity, improvements are less systematic, and regional models often underestimate observed MHW intensities even more than their corresponding GCMs. The next step will be to assess the ability of Med-CORDEX models to represent MHW properties in the subsurface, first by validating the simulated MHW characteristics in historical and early scenario years using reanalysis products, and then by investigating the projected subsurface MHW properties under future scenarios. 1 McAdam, R., & Bonino, G. (2025). Identification of Extreme Indicators for Task 3.2 "Marine extremes in Europe and the Arctic area" (Milestone MS5). Zenodo. https://doi.org/10.5281/zenodo.14623263 5 MS18 Preliminary assessment of the marine extreme indicators In the Barents and Nordic Seas, an initial assessment based on OSTIA (Donlon et al., 2012; Good et al., 2020) observations (1992–2024) has documented a clear seasonal intensification and spatial expansion of MHWs, especially in the Barents Sea. A more concise basin-wide evaluation of surface and subsurface MHWs will follow, primarily to provide a reference for assessing reanalyses and model performance. A primary focus of the work is the evaluation of MHW representation in the CMIP6 HighResMIP (Haarsma et al., 2016) ECMWF-IFS simulations at three ocean resolutions (LR, MR, HR) over 1980–2014. Using ESA CCI (Embury et al., 2024) or OSTIA SST as benchmarks, the first objective is to determine whether increasing ocean resolution improves the simulation of MHW frequency, intensity, duration, and spatial structure. The second is to assess how enhanced ocean resolution, while keeping the atmospheric grid fixed, modifies key MHW drivers, including surface heat fluxes, stratification, mixed layer depth, and Arctic inflow. A similar comparison will then be extended to the highest-resolution configurations. In the Baltic Sea, the core activity focuses on assessing SST and MHW characteristics using the high-resolution SmartSea NEMO model for the Gulf of Bothnia (Hordoir et al., 2019). The regional model provides six long-term projections (1976–2100) forced by downscaled CMIP5 (Taylor et al., 2012) atmospheric scenarios. Modelled MHW events are being compared with available in-situ and L4 satellite products, with a streamlined extension planned using L4 reprocessed SST (1982–2020) and OSTIA. The next phase will expand the assessment to subsurface temperature MHWs. The DestinE Digital Twin for Climate Change Adaptation (Climate DT, Doblas-Reyes et al. 2025) output will be analysed as soon as it becomes available: MHWs from the 30-year historical run and the extended projection will be assessed and compared to previous results for the Baltic Sea. Publications and dissemination: Results from Med-CORDEX models will be summarised in a paper entitled "The added value of Med-CORDEX regional models in the representation of sea surface temperature and marine heatwaves," planned for submission in early 2026 and presentation at the Ocean Sciences Meeting, to be held in Glasgow in February 2026. The Barents Sea study manuscript titled “Seasonal 6 MS18 Preliminary assessment of the marine extreme indicators Intensification and Spatial Evolution of Marine Heatwaves in the Barents Sea” will be submitted in the early part of 2026. 2.2.2 Sea level extremes Based on the definitions provided in Task 3.2 of MS5, sea level extremes have been detected using the 99th percentile as the threshold. For the Northwest European shelf, both historical and projected future changes in dynamic sea level (DSL) have been analysed using simulations from the Coupled Model Intercomparison Project Phase 6 (CMIP6, Eyring et al., 2016). The objective was to assess how the ocean's dynamic state is expected to evolve under various global warming thresholds and emissions scenarios. Historical CMIP6 DSL simulations were compared with observed detrended dynamic sea level data from the Copernicus Marine Service to evaluate the performance of CMIP6 models over the historical period. This comparison helped establish confidence in the models' ability to reproduce past oceanic conditions over the study region. Future changes in DSL were then explored across SSP1-2.6, SSP2-4.5, and SSP5-8.5 scenarios. The analysis shows a clear sensitivity of regional DSL to global warming, with spatially varying increases across the shelf. The Irish–Norwegian region emerges as particularly responsive, with projected rises of about 20–40 cm by the end of the century under high-emission pathways. Sea-level extremes analysis indicates a tendency toward more frequent and intense high sea levels as warming progresses, highlighting growing coastal vulnerability. The next planned step is to incorporate the thermohaline component into the ocean-dynamic sea-level analysis. For the Norwegian and Arctic coasts, ongoing work focuses on estimating sea-level extremes from tide gauges, satellite altimetry, and HighResMIP datasets. The analysis will start with monthly data and explore the possibilities of extending it to higher temporal resolution. The indicator data will be provided at different points, coinciding with the locations of tide gauges along the Norwegian coast and in the Arctic. The indicator data will contain two time series: a monthly sea level time series and an index 7 MS18 Preliminary assessment of the marine extreme indicators (1, 0) indicating whether a given month is categorised as an extreme, based on the 99th percentile threshold. Publications and dissemination: The results for the Northwest European shelf were presented at EGU 2025 (2) and will be submitted to Geophysical Research Letters and made available in open access (3). 2.2.3 Sea ice extremes Sea ice extremes in the Arctic Ocean have been analysed through an investigation of the CMIP6 archive of Global Circulation Models (GCMs) to determine the necessary temporal resolution to adequately represent these extreme events. Based on the definitions provided by WP3.2 in MS5, the analysis focused on the yearly minimum sea-ice extent. The investigation identified 19 models whose published output includes daily sea-ice concentrations. A new statistical approach was developed to determine whether the models adequately resolve extremes. The methodology involved sorting the extremes over the available historical period (1850-2014) and applying a running window of various lengths to investigate how this sorting changes. Through this analysis, it was identified that the window length necessary to identify the extremes in a useful manner was below a monthly time resolution (around 20 days). Therefore, the analysis concludes that daily sea-ice data are required in future model outputs to make useful statements about sea-ice extremes, such as minimum sea-ice extent. The following steps will include examining high-resolution simulations in the CMIP6 archive to investigate how results vary with spatial resolution and across time periods. Furthermore, the models will be compared with observational satellite data to gain further insights into model biases and their consequences for statements about sea-ice extremes. 3 Future projection of the Ocean Dynamic Sea Level over the Irish-Nordic-Arctic Seas under different global warming thresholds. (Authors: Emmanuel Eresanya, Gerard D. McCarthy, Veeranjaneyulu Chinta, and Hyacinth C. Nnamchi). 2 Eresanya, E., McCarthy, G. D., Chinta, V., and Nnamchi, H. C.: Future projection of the Ocean Dynamic Sea Level over the Irish-Nordic-Arctic Seas under different global warming thresholds, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-2008, https://doi.org/10.5194/egusphere-egu25-2008, 2025. 8 MS18 Preliminary assessment of the marine extreme indicators 3. Contribution to the ObsSea4Clim objectives This milestone contributes to the following specific objective of the project. SO2 To advance the use of EOVs and ECVs for improved ESM and reduced uncertainty in projections Implementation: ObsSea4Clim will advance the use of EOV and ECVs to develop the next generation of Earth System Models (ESMs) in preparation for CMIP7, constrain uncertainty in climate predictions (WP5), and improve regional-scale projections in the Application Areas (WP4). Improved process understanding in combination with advanced statistical methods for ensemble screening and AI optimisation of model parameters will be applied to reduce model biases with a focus on processes of both physical and biogeochemical importance, including ocean stratification, AMOC, sea ice loss, ice sheet interaction, and extremes such as marine heat waves (WP4, WP6). This milestone summarises the preliminary skill assessment of a set of indicators proposed in MS5 in ESMs. These results are crucial for further enhancing the milestone MS5 and achieving the deliverables D5.3, D3.2, D4.1, D4.4 and D4.5. 4. References Doblas-Reyes, F. J., Kontkanen, J., Sandu, I., Acosta, M., Al Turjmam, M. H., Alsina-Ferrer, I., Andrés-Martínez, M., Arriola, L., Axness, M., Batlle Martín, M., Bauer, P., Becker, T., Beltrán, D., Beyer, S., Bockelmann, H., Bretonnière, P.-A., Cabaniols, S., Caprioli, S., Castrillo, M., Chandrasekar, A., Cheedela, S., Correal, V., Danovaro, E., Davini, P., Enkovaara, J., Frauen, C., Früh, B., Gaya Àvila, A., Ghinassi, P., Ghosh, R., Ghosh, S., González, I., Grayson, K., Griffith, M., Hadade, I., Haine, C., Hartick, C., Haus, U.-U., Hearne, S., Järvinen, H., Jiménez, 9