Ryder Glacier in northwest Greenland is shielded from warm Atlantic water by a bathymetric sill

The study investigates how fjord bathymetry and ocean circulation control the exposure of Greenland’s marine-terminating outlet glaciers to warm Atlantic-origin subsurface water, which drives basal melting, thinning, calving, and dynamic retreat. Outlet glaciers terminating in floating ice tongues are sensitive to warm water inflow that enhances basal melt, reduces buttressing, and can lead to grounding-line instability and increased discharge, contributing significantly to sea-level rise. The research focuses on Sherard Osborn Fjord, where Ryder Glacier drains into the Lincoln Sea, and compares it with nearby Petermann Fjord. The central research question is whether the geometry of bathymetric sills in Sherard Osborn Fjord shields Ryder Glacier from warm Atlantic Water (AW), thereby reducing basal melting relative to Petermann Glacier, and whether such bathymetric controls help explain the differing recent behaviors of these two glaciers. The study is motivated by the need to reduce uncertainties in sea-level projections by resolving the role of fjord bathymetry and warm-water inflow in ice–ocean interactions.

Prior work shows that marine outlet glaciers respond strongly to oceanic forcing by warm subsurface waters of Atlantic origin, which increase basal melt, thin ice tongues, and destabilize flow. Historical observations document substantial changes at Petermann Glacier, including major calving events in 2010 and 2012 and net retreat since 1948, whereas Ryder Glacier exhibited smaller net advance with multi-year advance–retreat cycles possibly influenced by basal topography at the grounding line. Petermann Fjord bathymetry, previously mapped, shows a single sill at the fjord mouth; Jakobshavn’s rapid thinning was tied to Irminger Sea warming despite a sill that did not effectively block AW, emphasizing sill depth and geometry as key controls. The 79°N Glacier case similarly highlights how narrow channels across sills can restrict AW inflow. Atlantic Water enters the Arctic via Fram Strait and Barents Sea and has been implicated broadly in Greenland’s outlet glacier changes. These studies situate the present investigation into how Sherard Osborn Fjord’s double-sill configuration may modulate AW access to Ryder Glacier compared to Petermann’s single sill.

Field data were collected during the 37-day Ryder 2019 Expedition (Aug 5–Sep 10, 2019) aboard icebreaker Oden, with 15 days dedicated to Sherard Osborn Fjord and adjacent Lincoln Sea, and additional work in Nares Strait and Petermann Fjord. Multibeam bathymetry: Kongsberg EM122 (12 kHz, 1°×1°) echosounder with Kongsberg Seapath 320 navigation and MRU5 motion sensor. Sound speed profiles from the oceanography program were used for refraction correction; a patch test calibration (2019-08-21) was applied. Processing used QPS Qimera v1.7.6 with refraction re-corrections, total propagated uncertainty, and CUBE gridding to 15 m, with QA/QC and GIS products generated (Fledermaus DTMs and GeoTIFFs). Projection was North Polar Stereographic (true scale 75°N, Greenwich meridian). Oceanographic stations: 25 CTD profiles inside Sherard Osborn Fjord and 3 just outside in the Lincoln Sea; 21 CTD profiles in Nares Strait and Petermann Fjord. Instrumentation: Seabird 911 CTD with 24×12 L Niskin rosette; dual SBE3 temperature, dual SBE04C conductivity, SBE43 oxygen, WET Labs ECO-AFL/FL turbidity/fluorescence, Benthos altimeter. Data converted to potential temperature and absolute salinity using TEOS-10. Lowered ADCP: Two RDI Workhorse 300 kHz (upward and downward looking) mounted on CTD rosette; processed with LDEO IX13 LADCP package. Comparative sill geometry analysis: Cross-sectional areas of the warm-water gateways were computed across the shallowest passages of the inner sill in Sherard Osborn Fjord and the mouth sill in Petermann Fjord. Areas were constrained to waters warmer than 0 °C (depths where potential temperature ≥ 0 °C: ~275 m in Sherard Osborn; ~221 m in Petermann) to represent AW with high basal melt potential. An alternative comparison used a common reference depth (248 m) to evaluate sensitivity. Volume estimates: Volumes of ≥0 °C waters between the grounding lines and the constraining sills were estimated by extrapolating observed temperature structure beneath the ice tongues to the grounding lines (subtracting ice-occupied volume) using BedMachine v3 bathymetry/topography. Plume modeling: A one-dimensional buoyant-plume model (Jenkins 1991, 2011) based on conservation of mass, momentum, heat, and salt was used to estimate subglacial melt rates under Ryder’s tongue. Two ambient boundary conditions were tested: (1) colder profiles from inside the inner sill near the terminus (CTD 17) and (2) warmer profiles from the main basin seaward of the inner sill (average over selected CTDs). Model setup and parameters are detailed in Supplementary Information. Profiles and sections: Temperature–depth transects along- and across-fjord were constructed using selected CTD stations to evaluate thermal structure relative to sill positions and to compare with Petermann Fjord.

- Sherard Osborn Fjord contains two prominent sills: an outer sill at the fjord mouth and an inner sill located ~5 km seaward of the present Ryder ice-tongue terminus. The inner sill has a broad crest shallower than ~300 m (minimum ~193 m), with the deepest passage restricted to a ~1 km-wide channel at 390 m depth on the eastern side. The outer sill has deepest channels of ~475 m (east) and ~375 m (west). The main fjord basin reaches ~890 m depth. - Petermann Fjord has a single sill at the fjord mouth with a maximum depth of 443 m and a fjord maximum depth of 1158 m. - Atlantic Water (AW; θ > 0.3 °C, S > 34.7) is present in both fjords below ~350 m and in adjacent Lincoln Sea/Nares Strait. However, temperatures landward of Sherard Osborn’s inner sill are significantly colder: θ remains ≲0.2 °C from ~250 m to the bottom inland of the inner sill, compared to >0.3 °C in the main basin and outside the fjord. - Temperature–salinity relationships inland of the inner sill indicate cooling and freshening consistent with glacial melt at depth and influence of subglacial discharge at shallower depths, evidencing glacier–ocean modification of inflowing water. - Gateway size for warm-water inflow differs strongly: cross-sectional area for θ ≥ 0 °C water is ~264×10³ m² across Sherard Osborn’s inner sill (from ~275 m) versus ~1991×10³ m² across Petermann’s mouth sill (from ~221 m), i.e., Petermann’s passage is ~7.5 times larger. Using a common reference depth of 248 m, Petermann remains ~3.4 times larger. - Estimated volumes of ≥0 °C waters between the sills and grounding lines differ markedly: ~180 km³ for Ryder (inner sill to grounding line, ≥275 m) versus ~980 km³ for Petermann (mouth sill to grounding line, ≥225 m), i.e., Ryder’s is < one-fifth of Petermann’s. - Plume-model experiments suggest that if the warmer main-basin water seaward of the inner sill were to reach Ryder’s terminus, average subglacial melt rates would be ~15% higher than under the observed colder inland conditions. - Independent remote-sensing estimates indicate submarine melt flux (km³/yr) is ~5× larger for Petermann than Ryder; normalized by area, Petermann’s basal melt rate per unit area is still ~20% higher. - LADCP at the inner sill’s narrow channel shows a ~50 m-thick near-bottom inflow layer of ~0.3 °C water toward the glacier with speeds >0.2 m s⁻¹; a Richardson number estimate (Ri ≈ 0.6) implies strong shear-driven vertical mixing with overlying colder, glacially modified waters as flow traverses the inner sill and descends into the cavity. - The geometry and narrowness of Sherard Osborn’s inner sill enhance mixing and restrict AW penetration to Ryder’s cavity, effectively cooling waters reaching the vulnerable grounding line, thereby shielding Ryder Glacier relative to Petermann.

The findings directly address the research question by demonstrating that Sherard Osborn Fjord’s inner sill substantially restricts and cools Atlantic Water inflow, reducing basal melt potential at Ryder Glacier’s ice tongue and grounding line compared with Petermann. The smaller, shallower, and narrower gateway confines warm inflow to a thin bottom layer that undergoes significant mixing with colder, glacially modified waters, leading to sub-zero-to-near-zero temperatures landward of the sill. In contrast, Petermann’s broader and deeper mouth sill allows more voluminous, warmer AW to access the cavity with weaker vertical mixing, exposing the glacier to greater ocean heat. These oceanographic controls are consistent with the observed differences in recent glacier behavior: Petermann’s large calving events and net retreat versus Ryder’s modest net advance with cyclic fluctuations. Analogues (e.g., Jakobshavn’s response to Irminger Sea warming and the restricted AW inflow at 79°N Glacier) reinforce the importance of sill geometry and channelization. The study implies that sill configurations can modulate outlet glacier sensitivity to external ocean warming, thereby influencing ice dynamics and sea-level contributions. However, differences in geographic setting (Ryder’s more northerly location) and fjord geometry may also play roles and warrant further investigation. Overall, the results underscore the need to incorporate realistic fjord bathymetry and exchange processes into predictive models of ice–ocean interactions and sea-level rise.

This study provides new multibeam bathymetry and oceanographic observations from previously uncharted Sherard Osborn Fjord and demonstrates that Ryder Glacier is partly shielded from warm Atlantic Water by an inner bathymetric sill. The sill’s shallow crest and narrow deep passage constrain and mix inflowing warm water, yielding significantly colder conditions inside the inner fjord and near the ice tongue compared to Petermann Fjord. Comparative analyses of sill cross sections, warm-water volumes, observations (CTD/LADCP), and plume modeling collectively indicate reduced basal melt potential at Ryder relative to Petermann, helping explain their different recent behaviors. The key contribution is highlighting sill geometry as a critical control on grounding-line thermal forcing in fjord systems, with implications for projections of ice-sheet mass loss and sea-level rise. Future work should include: (1) year-round moored observations and repeated CTD/LADCP sections to capture temporal variability and confirm long-term conditions; (2) direct measurements beneath the ice tongue and at the grounding line; (3) targeted studies of how the distance between the ice-tongue terminus and sills influences cavity exchange and residence times; (4) incorporation of detailed fjord geometry and side-wall stress effects on glacier dynamics into coupled ice–ocean models; and (5) regional assessments of sill controls across Greenland’s fjords to refine vulnerability frameworks.

- Temporal coverage is limited to a ~2-week late-summer 2019 field campaign; results may not represent multi-seasonal or interannual variability. - No direct oceanographic measurements beneath the ice tongue or at the grounding line (~26 km from the terminus) were obtained; subsurface temperature structure there is inferred. - The estimated warm-water volumes rely on extrapolations beneath ice tongues and bathymetry from BedMachine v3, introducing uncertainty. - The plume-model estimates are first-order and idealized; they do not capture full 3D circulation, tides, or transient forcing. - Potential influences of Ryder’s more northerly location relative to Petermann on offshore AW properties are not fully resolved by the limited 2019 snapshots. - The study does not explicitly address side-wall stresses, fjord narrowing effects, or detailed ice dynamics that may influence advance–retreat behavior.