Skip to content

Antarctica briefing

Research and Response

Last reviewed October 13, 2026

Antarctic science is an exercise in assembling partial views. Satellites can measure an entire ice sheet but not every process beneath it. Field instruments can observe the ice–ocean boundary directly but only in a few difficult places. Models connect those observations through physical laws and test possible futures. None is sufficient alone.

The observing system

Satellite gravity missions measure changes in ice mass. Radar and laser altimeters track elevation. Synthetic-aperture radar measures ice velocity even through cloud and polar darkness. Autonomous floats and moorings record the ocean; aircraft map bedrock hidden beneath kilometres of ice; ice cores preserve past atmospheres.

Field programs provide the fine-scale measurements needed to interpret the continental view. The International Thwaites Glacier Collaboration, for example, has combined oceanography, glaciology, geophysics, remote sensing, and modelling to study one of West Antarctica's most consequential glacier systems. Its seven-year programme also shows why observation must continue: seawater intrusion at grounding zones and highly uneven melting beneath ice shelves are now known to matter, but are not yet fully represented in forecasts.

When the measurements continue but the explanation pauses

A working satellite does not guarantee a working public information service. NSIDC's account of its analysis programmes records that monthly Sea Ice Today analyses were paused from October 2025 through August 2026 after NASA funding ended. They resumed in September 2026 with support from the Climate Science and Communication Fund, administered by Surge Climate Talent.

The August 19 restart announcement makes the limits clear. Sea-ice maps and charts had continued during the commentary pause. Monthly analysis was returning, but the additional mid-month summer posts were not. As of this October 13 review, NSIDC still described Ice Sheets Today as suspended; occasional ice-sheet discussion within Sea Ice Today was not a full restoration of that separate service.

This is not evidence that the underlying satellite observations stopped. It is evidence that different parts of the information system can lose support at different times. Collecting a measurement, preserving its provenance, interpreting it, and explaining its significance to people making decisions are distinct jobs. Keeping one does not automatically keep the others.

The case gives a practical example of Krigsvold's position that Antarctic observation is global public infrastructure. The relevant question is not simply whether an instrument still works. It is whether the people and services needed to make its record useful can continue their work. This assessment describes the services documented on October 13, 2026; it is not a claim about every NSIDC programme or a prediction of future funding.

Models are not crystal balls

A climate or ice-sheet model is an explicit set of physical relationships, assumptions, and initial conditions. Researchers test models against historical observations, compare independent models, and run ensembles to see how outcomes change under different emissions and uncertain parameters.

The resulting range is information. A narrow range means different plausible assumptions produce similar outcomes. A wide range identifies a system or timescale in which planning must account for poorly bounded risk. A mechanism can also become less likely as models improve: recent high-resolution work has reduced concern that marine ice-cliff instability will collapse Thwaites this century, even while continued rapid ice loss remains the central projection. Uncertainty does not turn a measured trend into a guess, and revision does not mean the earlier research was pointless.

What the evidence supports

The major findings are consistent across observing systems: the climate is warming because of human greenhouse-gas emissions; Antarctica is losing ice; the Southern Ocean has absorbed a disproportionate share of human-added heat and carbon; and the severity of future change depends substantially on future emissions.

Those findings support four broad responses at different scales:

  1. reduce greenhouse-gas emissions to limit additional warming;
  2. use local sea-level scenarios—not a single global average—to plan infrastructure and communities;
  3. manage fisheries and other direct pressures so ecosystems retain more room to adapt;
  4. sustain the observations, field programs, data systems, and researchers needed to detect change early.

Science does not choose the distribution of costs or write a policy by itself. It does constrain honest choices. A response that ignores the measured direction of change is not made prudent by calling the future uncertain. Nor does acknowledging a lower estimate for one mechanism justify ignoring the larger body of evidence.

Primary sources