How we measure Atlantic overturning — and why it is not the Gulf Stream

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

The question this tool answers

The Atlantic overturning conveyor — formally the Atlantic Meridional Overturning Circulation, or AMOC — works like this: warm surface water moves north, releases heat, becomes denser, sinks, and returns south at depth. It is related to the Gulf Stream, but it is not the Gulf Stream. The Gulf Stream is largely wind-driven and will not simply switch off; the conveyor is the density-driven heat conveyor whose weakening could change North Atlantic winters, tropical rainfall, and dynamic sea level. This dashboard asks whether that overturning is weaker than in the early observing era, what the “cold blob” adds to the evidence, and how seriously to take collapse headlines.

How we know

The direct measurement comes from the RAPID mooring array at 26°N: an array of moorings stretched across the Atlantic near 26.5°N, from the Bahamas toward the African coast. Since 2004, instruments on those moorings have measured temperature, salinity, pressure, and currents well enough to infer the overturning transport in Sverdrups. One Sverdrup is one million cubic metres of water per second — an oceanographer's unit for continental-scale flow. The RAPID array is the first continuous observing system capable of watching the subtropical conveyor year after year.

Farther north, the subpolar North Atlantic overturning programme watches the subpolar limb, where dense-water formation matters most. The subpolar programme and the RAPID array are not interchangeable; they are different sections through a three-dimensional circulation. The other evidence is indirect: the subpolar North Atlantic has warmed less than most of the planet, and in some periods is anomalously cool. That “cold blob” is a sea-surface temperature fingerprint often associated with reduced northward heat transport. It is not a flow meter. It is a clue.

The choices we made

We lead with Atlantic overturning, not the Gulf Stream. It is tempting to use the familiar name, but doing so repeats the mistake the page is meant to correct. The hero states the conveyor story directly: weaker than the early RAPID-array baseline, not the consensus view that collapse is imminent, but no longer dismissed as science fiction.

The cold-blob swatches are promoted because they are the signature visual. “The rest of the world is warming. This patch is different.” That sentence makes the proxy understandable without pretending it is a satellite map. The RAPID-array chart remains the main direct evidence. The paper annotations are there because this is a live scientific debate: Caesar and colleagues emphasized a warming fingerprint of weakening; Ditlevsen and Ditlevsen argued for early-warning signals; van Westen and colleagues showed collapse pathways in models; Rahmstorf and others keep the observational fingerprint in view. The disagreement is part of the story.

What this tool cannot tell you

It cannot prove a collapse is coming. Twenty years of direct RAPID-array measurements are invaluable, but short for a circulation that varies over decades. Pre-2004 claims require proxies — sediments, temperature fingerprints, salinity, and model-data comparisons — with wider uncertainties. A negative trend since 2004 is evidence of weakening in the observed period, not a countdown clock.

It also cannot produce a city forecast. A weaker conveyor could cool parts of northwest Europe in winter relative to a world without weakening, shift tropical rainfall belts, and raise dynamic sea level along parts of the US East Coast. But the size, timing, and regional expression depend on emissions, ice melt, winds, and model physics. The localized panel is assessment framing, not a postcode forecast.

What's coming next

The strongest upgrade would be automated ingestion from the RAPID array and the subpolar overturning programme when licensing permits, plus gridded sea-surface-temperature and salinity fields so the cold-blob proxy becomes a real map. A second upgrade is an explicit “scientific disagreement” panel: what the United Nations climate panel says, what early-warning papers claim, what model-collapse papers show, and where each line of evidence is strongest or weakest.

Further reading

• Smeed et al. (2018), Geophysical Research Letters — the RAPID array’s observed overturning decline over the first decade-plus of direct measurements. doi.org/10.1029/2018GL078012
• Caesar et al. (2018), Nature — the North Atlantic warming-hole fingerprint of conveyor weakening. doi.org/10.1038/s41586-018-0007-4
• Lozier et al. (2019), Science — the first overturning result from the subpolar North Atlantic overturning programme, showing the subpolar North Atlantic view differs from old assumptions. doi.org/10.1126/science.aau6592
• Ditlevsen & Ditlevsen (2023), Nature Communications — early-warning analysis arguing for a possible conveyor tipping window; debated but influential. doi.org/10.1038/s41467-023-39810-w
• van Westen et al. (2024), Science Advances — modelled conveyor collapse pathway and freshwater transport signal. doi.org/10.1126/sciadv.adk1189

Credits

The direct overturning record exists because oceanographers put instruments across an ocean and kept them there: the RAPID-MOCHA-WBTS teams at 26.5°N, the British Oceanographic Data Centre release process, and the subpolar North Atlantic overturning programme consortium in the subpolar North Atlantic. The cold-blob context depends on the US National Oceanic and Atmospheric Administration Physical Sciences Laboratory sea-surface temperature products and a long literature connecting fingerprints to overturning. This dashboard fetches public indices and embeds published releases. We are downstream of their work.