Quality-controlled bottom pressure dataset to analyse for continental shelf waves: Contribution to WP1 OI (Deliverable D1.6)

Ocean Cryosphere Exchanges in Antarctica: Impacts on Climate and the Earth system Quality-controlled bottom pressure dataset to analyse for continental shelf waves: Contribution to WP1 OI Deliverable D1.6 UK Partners are funded by UK Research and Innovation (UKRI) under the UK government’s Horizon Europe funding Guarantee. https://ocean-ice.eu/ OCEAN ICE – GA 101060452 Deliverable D1.6 About this document Deliverable: D1.6 Quality-controlled bottom pressure dataset to analyse for continental shelf waves: Contribution to WP1 OI Work Package: WP1 Subpolar circulation, heat delivery and water mass export Delivery date: 31 October 2025 Type of document: Report Dissemination level: Public Lead beneficiary and author: PP13: United Kingdom Research and Innovation - British Antarctic Survey (UKRI-BAS): Pierre Dutrieux, Shenjie Zhou Contributors: PP3: Alfred-Wegener-Institut Helmholtz-Zentrum für Polarund Meeresforschung (AWI): Markus Janout PP4: Centre National de la Recherche Scientifique (CNRS): Cosme Mosneron-Dupin PP11: Norwegian Polar Institute (NPI): Torre Hattermann Review: PP1: Danish Meteorological Institute (DMI), Chiara Bearzotti ([email protected]) Cover sheet: Schematic depiction of a mooring recovery in polar environments. 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 European Research Executive Agency (REA). Neither the European Union nor the granting authority can be held responsible for them. 2 OCEAN ICE – GA 101060452 Deliverable D1.6 Table of contents 1. Publishable summary.....................................................................................................................................4 2. Work performed and main achievements..................................................................................................... 5 2.1 Description of the work performed.......................................................................................................5 2.1.1 A compilation of historical time series of near-bottom pressure around Antarctica...................5 2.1.2 Comparison between state-of-the-art pressure records and our new dataset............................7 2.2 Open Science.......................................................................................................................................10 3. Results..........................................................................................................................................................10 4. Impact.......................................................................................................................................................... 11 3 OCEAN ICE – GA 101060452 Deliverable D1.6 1. Publishable summary In this deliverable, we present the first version of a compilation of historical moored timeseries of pressure within 50 m of the seabed that have been deployed over the past 50 years over Antarctic continental shelves and slopes. We further compare the type of records contained within our compilation with state-of-the-art pressure time series typically observed by higher-sensitivity sensors directly seated on the seabed and demonstrate how this new dataset can be used to evaluate related satellite products in polar settings. The standardized data collection of near-seabed pressure records we present is restricted to the region south of 60°S (Figure 1) and is freely accessible from SEANOE (https://www.seanoe.org/data/00983/109453/ ), and we hope to complement it with additional records and maintain it in the future. We refer to this data compilation as the OCEAN ICE near-seabed pressure compilation herein. Though beyond the scope of this data release report, this dataset provides an opportunity for a systematic study on pan-Antarctic ocean wave dynamics. Fig. 1: Position of moored records of ocean pressure within 50m of the seabed since 1970 are marked by white circles. Colour indicates the mean time (central year) covered in individual time series, and circle diameter indicates the length of each record. 4 OCEAN ICE – GA 101060452 Deliverable D1.6 2. Work performed and main achievements 2.1 Description of the work performed 2.1.1 A compilation of historical time series of near-bottom pressure around Antarctica Despite its crucial role in global climate dynamics, the Southern Ocean remains one of the most mysterious oceans on Earth. A major obstacle to understanding this region is the difficulty of observing areas covered by seasonal sea ice. Oceanic bottom pressure is a crucial tool to measure ocean dynamics. Typically, it is measured with sensors that are directly seated static on the seabed, thereby allowing high accuracy observations of the weight of the water column thickness above it. In turn, this weight can be affected by changes in the density of the water mass above the sensor and/or its height, both of which are determined by the kinematic displacement of water masses or by dynamical rearrangements of ocean properties through mixing and interactions with the atmosphere and the cryosphere. At high frequency (hours to weeks), it is typically assumed that the main driver of change is the former, whilst changes in water mass properties become increasingly important over time. In both cases, bottom pressure variability is closely connected to that of ocean surface height (hereafter referred to as sea surface height, or SSH), which is similarly affected by the same processes. Over temperate, open ocean, SSH can also be measured by satellite altimetry, and for this reason, bottom pressure records have been used extensively to evaluate and calibrate satellite-derived SSH products. But in polar settings, sea ice also affects the satellite-derived SSH and our ability to easily deploy and recover bottom pressure records. There is therefore a clear need to obtain joint SSH observations from satellites and in situ bottom pressure proxies. As part of OCEAN ICE, a novel SSH dataset leveraging altimeters' ability to detect leads in sea ice was developed and published (https://doi.org/10.5281/zenodo.17467408 ). The goal of the following work is to provide a proxy for in situ bottom pressure time series that can be used alone for analysis of ocean dynamics, to evaluate satellite-derived SSH products, or both. Accessing observational time series of ocean near-bottom pressure can be an extremely cumbersome exercise, as datasets tend to reside in disparate national data centers, under varying names and formats. The same is true for time series of ocean temperature, salinity and current. As part of the OCEAN ICE project, our team previously aggregated moored time series for all the above variables, at all depths. The larger dataset is available here: https://www.seanoe.org/data/00887/99922/, and a publication presenting the observation and providing a first overview of the potential it offers for science is published here: https://doi.org/10.5194/essd-2025-54. The database provided above was created to be as complete and easy to access as possible, offering a unified nomenclature for variable names and links to the original sources of the data. However, extracting singular variables for singular scientific purposes still requires data mining skills. Here, we go one step further and select from the above dataset those pressure time series that are within 50 m of the seabed. The new compilation dataset is available here (https://www.seanoe.org/data/00983/109453/). We aim to maintain it in the future as additional observations are provided to us and may, in the future, add a compilation of bottom pressure records if the community requests it, though this is beyond the scope of the present publication. 5 OCEAN ICE – GA 101060452 Deliverable D1.6 Fig. 2: Comparison between BPR and moored CTD. Time series of observed BPR (a) and moored CTD (b) pressure anomaly records from the Norwegian Polar Institute at site DML1. Blue lines show the full resolution records, the orange line shows a 15-60-day band-pass filtered version of the same. c)/d)/e) Direct comparison between bottom pressure anomalies from BPR (blue) and moored CTD (orange) in the 15-60 days band pass at sites DML1/DML3/DML4. 6 OCEAN ICE – GA 101060452 Deliverable D1.6 2.1.2 Comparison between state-of-the-art pressure records and our new dataset As a prerequisite to ensure the viability of our bottom pressure proxy dataset, we compare state-of-the-art seabed-mounted bottom pressure (BPR) observations with pressure records from sensors (typically measuring conductivity, temperature and depth, aka CTD) moored at the same location and same time but 20m higher in the water column. In addition to measuring pressure with a lower frequency range and (sometimes) lower sensor accuracy, one could expect that movements of the mooring line associated with varying ocean currents could affect the moored sensor depth and therefore its measured pressure variability. CTD observations are used as proxies in our new database, whilst BPR are much rarer in the Circum-Antarctic context (and in general). Between 2023 and 2025, the Norwegian Polar Institute deployed and recovered such joint BPR/moored observations at three locations north of the Fimbulisen ice shelf, East Antarctica, aligned across the continental shelf break. Sites DML1 and DML3 were seated slightly down the continental shelf break, whilst site DML4 was closer to the top of the break in slope. Full resolution observed time series are shown in Figure 2 (a and b) for DML1, both displaying signals at tidal, fortnightly and longer frequencies. Because current variability is affected by processes with varying amplitudes and frequencies, one can expect co-variability to be frequency-dependent. A direct comparison between time series in the 15-60 day time scales (Figure 2c-e) indicates that the moored CTD tracks BPR variability with remarkable accuracy at these time scales, especially at sites DML1 and DML3. Site DML4 shows periods of strong cross-correlation (between October 2023 and July 2024) and periods of mismatch (July 2023 and August 2024). The record is too short to draw extensive conclusions about the reasons for the decorrelation, but it is interesting to note that both periods of poor correlation coincide with the austral winter, perhaps indicative of a seasonal bias at that location. Fig. 3: Correlation between BPR and moored CTD. Band-limited correlation between BPR and CTD as a function of frequency. Time series are band-pass filtered around the circled frequencies and correlated with each other for sites DML1 (blue), DML3 (orange), and DML4 (green). Regardless, the cross-correlation between the proxy and the state-of-the-art record is actually very promising (Figure 3). Starting from tidal and longer time periods (>6h), band-limited correlations show high values (>0.9) at sites DML1 and DML3. For site DML4, correlation reaches 1 at the tidal and inertial 7 OCEAN ICE – GA 101060452 Deliverable D1.6 frequencies (close to 1 day period), is greater than 0.7 for periods greater than 60 days, but drops between the two frequency ranges, mostly due to the austral winter decorrelation periods (Figure 2e). Though limited in space (Fimbulisen shelf slope) and time (2023-2025), this preliminary analysis indicates strong potential for moored proxies to capture the variability of full-water-column ocean pressure. Because moored observations are much more numerous than BPR, they may be extremely useful for spatio-temporal analyses of pressure variability in the circum-Antarctic context. Fig. 4: Comparison between BPR anomaly proxies (moored CTD) and SSH anomaly. Direct comparison between observed SSH (orange) and bottom pressure proxy (blue) at site 6 (top) and site 1 (bottom). Middle panel: Band-pass filtered (30-60 days) time series of SSH and BPR proxy at site 6. Correlations are noted in panel subtitles. 8 OCEAN ICE – GA 101060452 Deliverable D1.6 2.1.3 Demonstration of the utility of the new dataset: comparison with satellite-derived SSH One example of use for our new dataset is a direct comparison with SSH as measured from satellite altimetry (Figure 4, top). Indeed, observing SSH through sea ice leads is a fairly arduous exercise and is necessarily error-prone. Similarly, our proxy dataset for BPRs (the abovementioned CTD) contains some noise. But their respective and joint spatio-temporal coverage is sufficiently large that we can hope to obtain a statistically meaningful comparison over the observational ensemble. For each joint observational time period at each location, we extracted SSH observations. Limited by the frequency of satellite pass repeat cycles, SSH is necessarily smoother than BPR proxies (hereafter referred to as BPRs in the figures, for simplicity). However, band-passed comparisons between SSH and BPR proxies (Figure 4, middle panel) can indeed show strong correlation (>0.8) in a few locations, and lower or no correlations in others. The reason for the lack of correlation between the two types of observations can be reconducted to multiple factors. In the case represented in Figure 4 (bottom panel), the moored CTD time series of pressure varies far beyond the range of typical deep water pressure variability, reaching daily variation of order 4 m, and a sudden jump of about 10 m in the early part of the record. We surmise that such large-amplitude variability is associated with a coastal setting, the jump being associated with an iceberg displacing the mooring, or an anomalous CTD record. A coastal setting would be extremely hard to capture with satellite observations, but in any case, the associated satellite record is completely uncorrelated to the in situ record in that instance (correlation of 0.03). Fig.5: Correlation between BPR anomaly proxies (moored CTD) and SSH anomaly. Left) Band-passed correlation between observed SSH anomaly and bottom pressure anomaly proxy across all records (coloured circles). Right: probability density function of band-passed correlation between SSH anomaly and BPR anomaly proxies. The dashed black line indicates a perfect match. The red line indicates the linear regression across all records. A broader look at the entirety of the database suggests that some regions, especially those offshore and in the Weddell Sea sector, show high correlation between satellite and in situ observations, while others (tendentially more coastal) are not well correlated (Figure 5, left panel). Taken as a whole, a linear 9