How we measure sea ice — and why the two poles are telling different stories

The dashboard's data, the choices behind it, and the limits of what it can tell you.

The question this tool answers

Sea ice is the planet's most visible thermostat. When it shrinks, the bright white surface that bounces sunlight back to space gives way to dark ocean that soaks it up — so a melting pole warms faster, which melts more ice. This dashboard tracks how much of each polar ocean is ice-covered, every day, and asks two things most sites collapse into one: is the Arctic losing ice (it is, steadily, for decades), and is Antarctica doing the same (it wasn't — until recently). Treating the poles as a single "global sea ice" number hides the most interesting science on the page.

How we know

The numbers come from the National Snow and Ice Data Center (NSIDC) in Boulder, Colorado, which maintains the Sea Ice Index — the canonical satellite record of polar ice extent. The measurement is not a photograph. Sea ice and open water emit microwave radiation differently, and a series of passive-microwave radiometers flown on US Defense and NASA satellites has read that difference from orbit since late 1978: SMMR on Nimbus-7, then the SSM/I and SSMIS instruments on the DMSP weather satellites. Because microwaves pass through cloud and don't need sunlight, these sensors see the poles through the long polar night and through storms — which is why the record is daily and nearly unbroken across more than four decades.

"Extent" has a specific meaning here: the total area of ocean where at least 15% of the surface is covered by ice. That 15% threshold is a convention, chosen because it's the level the microwave sensors can detect reliably. We compare each day's extent to NSIDC's 1981–2010 average for that same calendar date, so "9% below normal" always means below the normal for this day of the year, not below some annual figure.

The choices we made

Two poles, never averaged. The single most important editorial choice on this page is that we keep the Arctic and Antarctic apart. They are different systems: the Arctic is an ocean ringed by land, where ice is increasingly thin and young; the Antarctic is a continent ringed by open ocean, where ice forms freely and is pushed around by winds and currents. Their trends point in different directions and on different timelines, and a "global sea ice extent" number — sometimes quoted to suggest the planet is fine — buries exactly that.

We also distinguish "below normal" from "record low." A day can be well below the seasonal average without setting a new record for that date, because an even lower year already holds it. The record-day calendar exists to make that difference visible: each red square is a day this year that fell below every prior year on that calendar date since 1979. Winter maximum (the Arctic's peak extent between January and April) gets its own number because it sets the starting line for summer melt.

What this tool cannot tell you

It measures area, not thickness or volume. A pole can hold the same extent while the ice beneath thins dramatically — and Arctic ice has been thinning and getting younger for decades, which extent alone cannot show. Ice volume needs a different instrument (radar and laser altimeters such as CryoSat-2 and ICESat-2) and is not on this page. The microwave sensors also struggle with melt ponds in summer, which can look like open water, and with very thin new ice, so day-to-day wiggles are normal noise around a robust multi-decade signal.

The Antarctic story is genuinely unsettled. After holding steady — even growing slightly — through the 2000s and early 2010s, Antarctic sea ice fell to startling lows from 2016 onward. Whether that is a temporary swing or a shift to a "new state" is an active research question, and we present the recent decline as what the data show, not as a settled trend.

What's coming next

NSIDC has signalled changes to its level of service as ageing DMSP satellites are retired, so continuity of the daily microwave record is itself a live concern; newer sensors and reprocessing will carry it forward. On our side, the natural next additions are a daily (Charctic-style) line chart, ice-thickness and volume from the altimetry missions, and regional breakdowns for individual seas — all deferred for now to keep the page readable.

Further reading

• Stroeve & Notz (2018), Environmental Research Letters — "Changing state of Arctic sea ice across all seasons." The clearest summary of the Arctic decline. doi.org/10.1088/1748-9326/aade56
• Parkinson (2019), PNAS — the 40-year Antarctic record showing slow increases followed by abrupt decreases. Essential for the divergence story. doi.org/10.1073/pnas.1906556116
• Purich & Doddridge (2023), Communications Earth & Environment — argues the recent Antarctic lows mark a possible new sea-ice state. doi.org/10.1038/s43247-023-00961-9
• Serreze & Stroeve (2015), Phil. Trans. R. Soc. A — Arctic trends, variability, and the limits of seasonal forecasting, from two NSIDC scientists. doi.org/10.1098/rsta.2014.0159
• NSIDC Sea Ice Index (v4) — the operational dataset behind this dashboard, with documentation and daily files. nsidc.org/data/g02135 · day-to-day commentary at Sea Ice Today

Credits

The polar ice record exists because of the engineers who built and flew passive-microwave radiometers from Nimbus-7 onward, the US Air Force and NASA teams who kept the DMSP series in orbit for four decades, and the scientists at NSIDC — among them Mark Serreze, Walt Meier, Julienne Stroeve, and NASA's Claire Parkinson, whose Antarctic record anchors the divergence story — who calibrate, reprocess, and publish the canonical series. This dashboard fetches from their public files. We are downstream of their work.