What the Thwaites Glacier Expedition Could Reveal About Sea-Level Rise

Scientists have focused on Thwaites Glacier because of two related problems: it sits on ground that slopes inland beneath the ice, and warm ocean water can reach its base. That geometry makes the glacier sensitive to small changes at its seaward edge. Recent field campaigns, remote sensing and robotic dives together form the latest Thwaites Glacier expedition effort to map these processes more precisely.

Think of the glacier as part of a flooded shelf: where the ice detaches from bedrock, small increases in melting can let the floating part widen. That widening can let grounded ice flow faster to the sea. The expedition combines seismic, borehole and under-ice vehicle data with satellite observations to move from general warnings to concrete numbers — distances of retreat, local melt rates, and where fractures are most dangerous. Those numbers matter because they change how cities and insurers estimate long-term coastal risks.

Thwaites Glacier is one of the largest single contributors to potential sea-level rise from Antarctica. Scientists estimate that the ice stored in and behind Thwaites is enough to raise global sea level by around 0.65 m if it were lost over time. That does not mean such a rise would happen quickly; the key question is the rate and pattern of ice loss.

A few concepts help explain why the glacier is monitored closely. The grounding line is the point where the ice sheet detaches from bedrock and begins to float. If this line retreats inland on a bed that slopes downward, retreat can become self-sustaining: more ice floats, flow speeds up, and the grounding line retreats further. Basal melt means melting at the underside of the ice where it meets the ocean; small changes in basal melt near critical pinning points can have outsized effects.

Observations since the 1990s show the grounding line has retreated in places by more than a few kilometres, and parts of the eastern shelf are developing fractures that weaken the ice.

Differences in measurement methods explain some apparent contradictions. Borehole and submersible probes give detailed local melt rates, while satellite interferometry (InSAR) and altimetry show broader trends. Combining both scales is what the Thwaites Glacier expedition aims to do.

If numbers help, a compact table clarifies the scale and recent findings.

The recent Thwaites Glacier expedition combined several approaches to reduce scientific uncertainty. Seismic surveys reveal the shape of the bed under the ice. Boreholes let researchers lower temperature and salinity sensors into pockets under the glacier. Autonomous underwater vehicles map channels and terraces beneath floating ice shelves. Satellites track surface motion over time. Together these methods answer targeted questions: where is melt concentrated, how fast are fractures growing, and which places act as stabilising “pinning points”?

An autonomous underwater vehicle (AUV) is a battery-powered robot that swims beneath the ice and records sonar and water‑property measurements. It is especially useful because it can map narrow sub-ice cavities that satellites cannot see. Borehole measurements, by contrast, give vertical profiles through the ice and immediate conditions at specific locations. The expedition’s novelty lies in synchronising point measurements and wide-area remote sensing so models are constrained by both local physics and large-scale geometry.

Findings reported by the International Thwaites Glacier Collaboration and associated projects in 2024–2025 show heterogeneity: some shallow areas are partly sheltered by freshwater layers, reducing melt there; in contrast, fractures, steep slopes and terraces experience much higher basal melt and mechanical weakening. These local hotspots matter for overall stability because they control where and how fast calving and shelf disintegration can proceed.

Because data types differ in age and resolution, some sources predate the current synthesis. For example, a 2023 NSF-supported study offers detailed borehole and submersible records; this study is more than two years old, but its direct measurements remain important when compared with newer 2024–2025 surveys and satellite trends.

Sea-level rise is often discussed globally, but its local effects determine planning decisions. Small differences in long-term rates change the timing of necessary investments: seawalls, building elevations, and insurance models. Suppose a region expects 0.5 m rise over a century under one scenario and 0.8 m under another; that half-metre difference alters whether low-lying districts remain usable without repeated intervention.

Research from the Thwaites Glacier expedition feeds directly into these risk calculations. Coastal engineers and city planners use probabilistic scenarios produced by ice-sheet models. When those models include process details observed on the ground — such as where basal melt concentrates or where fractures cut an ice shelf into weaker pieces — the resulting sea-level projections tighten, especially for higher-end outcomes that matter for long-lived infrastructure.

Insurers and governments typically plan with median scenarios but also consider low-probability, high-impact cases for critical assets. The expedition’s data reduce ambiguity around some high-impact pathways: for instance, if a previously stabilising pinning point is found to be weakening, the probability of faster retreat increases and planners must update adaptation timelines. That does not mean immediate catastrophe; for many places the effects unfold over decades to centuries. Yet for major ports and large coastal investments, earlier updates improve resilience and allocate funds more efficiently.

Scientific projections for Thwaites range across scenarios because models differ in how they include fractures, terraces and local ocean circulation. Where model physics omit rapid crack propagation or concentrated cavity melt, they tend to under‑represent worst‑case but lower‑probability paths. Adding the expeditions observations into models helps close that gap, but some uncertainties remain intrinsic: exactly how quickly a chain reaction of grounding-line retreat and shelf loss could accelerate is highly sensitive to local geometry.

One clear outcome of recent work is that not all parts of Thwaites behave the same. Some sectors show slow, steady retreat; others reveal hotspots of vigorous melting beneath terraces and near fractures. This spatial variability means projections should avoid relying on a single average rate across the whole glacier. Instead, credible projections incorporate local extremes as well as medians.

For readers interested in what to watch next: improved satellite time series and renewed under-ice missions will refine rates of change within a few years, while updated ice-sheet models that explicitly include fracture mechanics and small-scale ocean processes will change century-scale projections. Policymakers and planners will see the most immediate value as probability estimates for faster change become narrower, enabling better cost–benefit choices for adaptation.

Field data from the Thwaites Glacier expedition bring much-needed detail to a central question in sea-level science: which processes control the rate of ice loss? Combining boreholes, autonomous underwater vehicle surveys and satellite observations reveals where melt and fractures concentrate and which places act as stabilisers. Those findings do not deliver a single countdown to a sea-level threshold, but they do refine the probabilities and timelines that decision-makers use. For coastal communities and infrastructure owners, better probabilities are useful: they turn vague warnings into clearer investment and adaptation choices.