Atlantic Multidecadal Variability
As of September, the Atlantic Multidecadal Variability (AMV) is in a Warm phase, 26 years in, with the North Atlantic +0.25°C against the global ocean - and the likeliest path is a weak Warm phase to the late 2030s, then Neutral.
AMV Now and Next
- Now · ERSST v5, to Sep 2026
- Warm phase since 2001 · 26 years in · the North Atlantic +0.25°C on the global ocean in SepCurrent State
- Decades ahead · Zanchettin & Rubino 2024
- A weak Warm phase to the late 2030s, then most likely Neutral by mid-centuryWhat Follows
- Usually · phases since 1854
- Warm and Cool phases of 36 to 41 years · 3 complete since 1854What Follows
What It Is - The North Atlantic’s slow swing between Warm and Cool phases, each lasting decades - a full cycle takes about 60 to 80 years. It is also called the Atlantic Multidecadal Oscillation (AMO).
How It’s Measured - Against the rest of the ocean. The index is the North Atlantic’s sea temperature (the equator to 60°N) minus the whole global ocean’s, so the warming every ocean shares cancels out, and what is left is the Atlantic running ahead of or behind the rest (Trenberth & Shea 2006). The phase is set by its 10-year mean - a single month or year is mostly weather.
The AMOC Link - The Atlantic overturning (AMOC) carries warm water north. When it runs stronger, more heat reaches the North Atlantic and it warms against the rest of the ocean. When it weakens, the Atlantic falls behind. Ocean records and climate models suggest the AMOC paces the phases over decades, but its direct measurements only began in 2004, too short to show a full cycle.
What Flips It - No single switch. The leading idea is a slow loop: a stronger AMOC warms the North Atlantic, and over decades the changes this brings to the water’s saltiness and density weaken the sinking that drives it - so the AMOC slows, less heat arrives and the Atlantic cools, until the loop turns again. The NAO’s winds push the ocean too, and air pollution plays a part: sulphate haze from Europe and North America helped cool the North Atlantic from the 1960s, and its clean-up since the 1980s helped the warm-up (Booth et al. 2012). How much is the ocean and how much the haze is still debated.
What It Brings - A Warm phase brings wetter Sahel summers, busier hurricane seasons and drier US Plains. A Cool phase broadly the reverse - the Sahel drought of the 1970s-80s came in the last Cool phase.
On This Page - The site’s monthly index since 1854, today’s sea temperatures with real example weeks, what it means for the decade ahead, how long phases last and where this one stands, every phase on record, its partners (the AMOC, the PDO and the NAO), its strongest years, the places it drives most and where it bites hardest.
Current State
The North Atlantic’s slow swing against the global ocean - Warm or Cool phases that last decades.
AMV State
since 2001AMV +0.25°C Sep 2026 · 12-month +0.03°C · 10-year +0.20°C
The North Atlantic warmer than the ocean as a whole - wetter Sahel summers, busier hurricane seasons
AMOC · Partner
16.4 Sv12 months to Mar 2024Near Normal · mean 17.0
Near Normal · mean 17.0
Thought to pace the AMV · AMOC ↗
North Atlantic
Reading the Map
- The AMV’s box - the North Atlantic from the equator to 60°N, in its phase’s colour. Its label gives the box against the global ocean on the map’s day, the index’s own sum a day at a time.
- The Gulf Stream and North Atlantic Current - the warm upper limb of the overturning (AMOC), thought to pace the AMV over decades. Thicker in a Warm phase. Live, its label gives RAPID’s latest reading.
- What the phase usually brings over decades - the Sahel’s rains, the hurricane seasons, the Caribbean, the US Plains, north-east Brazil and Europe’s summers.
- The pills swap the latest day for a real week of each phase: a strong Warm AMV (August 2010, a busy hurricane year), a Neutral October (2002) and a strong Cool AMV (August 1992).
- Sea surface temperature against 1991-2020 (NOAA OISST, a recent day). The map keeps the world’s warming in, so most of the ocean looks warm - the AMV asks whether the Atlantic is warmer still.
- Alerts since 12 Nov 2025 - 1 AMV-linked of 20 in its reach. Context, not cause.
Key Terms
- Atlantic Multidecadal Variability
- - The North Atlantic’s slow swing against the global ocean, in phases that last decades. Also called the AMO.
- Warm AMV
- - The Atlantic warmer than the ocean as a whole - wetter Sahel summers, more hurricanes.
- Cool AMV
- - The reverse - the Sahel drought of the 1970s-80s came in the last Cool phase.
- AMOC
- - The Atlantic overturning, thought to pace the AMV over decades.
- The phase
- - Set by the 10-year smoothed series: ±0.1°C or beyond counts. A single month or year is mostly weather.
The AMV is the site’s index from NOAA ERSST v5, monthly since 1854: the North Atlantic (0‑60°N) anomaly less the global ocean’s, against 1991-2020 (Trenberth & Shea 2006). The phase is the 10-year smoothed series: Warm at ≥ +0.1°C, Cool at ≤ −0.1°C, starting where it last crossed zero.
The map is NOAA OISST v2.1’s daily sea surface temperature against 1991-2020; the box figure is its 0-60°N, 80°W-0° mean less the 60°S-60°N ocean’s. The examples are OISST on those days, with the site’s AMV for the month. The AMOC is RAPID’s 26.5°N array, released a year or more late.
Alerts from GDACS (Orange and Red, the last 12 months), flagged AMV-linked when the hazard fits how a Warm or Cool AMV usually moves the place in that season, on the site’s own measured links - or, for Atlantic hurricanes, the busier seasons a Warm AMV brings (Goldenberg et al. 2001).
What It MeansWhat It Means - AMV Briefing
For the decade ahead
The AMV is in a Warm phase, and it moves over decades, measured against the rest of the ocean. It shapes the odds over many years, not any one season. A Warm AMV usually brings wetter conditions to the Sahel and the Caribbean. Atlantic hurricane seasons tend to be busier. The US South and Plains lean drier and warmer. North-east Brazil leans drier. European summers lean warmer.
The partners
RAPID last measured the AMOC at 16.4 Sv, near normal. It is thought to pace the AMV. The PDO is in a Cool phase, and the two tend to run opposite. Over the strong AMV years since 1950 they were opposite in 8 of 12.
How this compares
This Warm phase is 26 years old, shorter than the 3 complete phases, which ran 36 to 41 years. Its 10-year mean peak of +0.20°C is close to the Warm phase of 1927-1964, which peaked at +0.22°C. The published outlook is a weak Warm phase to the late 2030s, then most likely Neutral.
What to watch
The site rebuilds the monthly AMV from NOAA's ERSST each month, with October's value due in early November. The next hurricane season runs from 1 June to 30 November. RAPID publishes the AMOC a year or more late.
Written by Claude from this page's own data, every figure checked against it · 9 Nov 2026
What Usually Follows
How Long Phases Last
Every phase since 1854 by its length, this one still growing - then each kind side by side from the year it began.
Phase Lengths · Years
- Warm 1854-8532+
- Cool 1886-2641
- Warm 1927-6438
- Cool 1965-0036
- Warm 2001Now26
32+ - began before the record in 1854 (dashed below). Faint - still running. Hatched - likeliest end, 2036-2039 (Zanchettin & Rubino 2024).
This Phase
A Warm phase, 26 years in, its 10-year mean at +0.20°C. The last Warm phase ran 38 years (1927-1964), and at this point it stood at +0.12°C. If this one runs as long, it would turn about 2039.
What That Means - The North Atlantic’s warm tilt most likely holds another decade, weakening - wetter Sahel summers and busier hurricane seasons stay more likely than not into the 2030s. A turn to Cool would bring the reverse, but none is expected before then.
Usually - Phases have run 36 to 41 years, alternating Warm and Cool, and their 10-year mean has peaked 9 to 25 years in. Each turn took the North Atlantic’s weather with it - the Sahel drought came with the last Cool phase, the busy hurricane seasons of the 2000s with this Warm one.
| Phase | Lasted | Peak |
|---|---|---|
| Warm 1854-1885 | 32+ | +0.14°C yr 23 |
| Cool 1886-1926 | 41 | −0.18°C yr 24 |
| Warm 1927-1964 | 38 | +0.22°C yr 9 |
| Cool 1965-2000 | 36 | −0.20°C yr 25 |
| Warm 2001-now | 26So Far | +0.20°C yr 26 |
The site’s AMV from NOAA ERSST v5: the North Atlantic (0‑60°N) less the global ocean, against 1991-2020 (Trenberth & Shea 2006). A phase is a run of the 10-year mean’s sign, runs under five years folded in; its peak is the 10-year mean’s furthest point from zero. The likeliest end is from Zanchettin & Rubino 2024 (a weak Warm phase to the late 2030s, then Neutral).
Phases
Warm and Cool Phases Since 1854
The yearly AMV and its 10-year mean, which sets the phase - each Warm or Cool run shaded.
AMV by Year · 1854-2026
The Phases - 3 complete since 1854, each of 36 to 41 years, alternating Warm and Cool. The present Warm phase began in 2001. The likeliest path is a weak Warm phase to the late 2030s, then Neutral by mid-century (Zanchettin & Rubino 2024).
Built by 4 Billion Years On from NOAA’s ERSST v5: the North Atlantic (0‑60°N) anomaly less the global ocean’s, against 1991-2020 (Trenberth & Shea 2006). The 10-year mean is centred, so its last five years use a shorter window and can still move. A phase is a run of the 10-year mean’s sign, with runs under five years folded into the phase around them.
History
The AMV Since 1854
Every year since 1854 against the global ocean, with the decades that leaned each way, the years remembered and the monthly record.
Years Since 1854
22 / 19 / 131Warm, Cool, Neutral
Warm, Cool, Neutral
Beyond ±0.2°C, of 172 years
By Decade
+0.21 / −0.18The 2020s and the 1970s
The 2020s and the 1970s
2 of 6 years warm in the 2020s, 0 of 10 in the 1970s
Year to Year
r +0.64One year leans like the last
One year leans like the last
Year to year
AMV Years · Jan-Dec Mean, 1854-2025
Notable Years
- 1955+0.31°CThe 1950s US drought, with a Cool PDO ↗
- 1970−0.08°CThe Sahel drought begins as the Cool phase sets in ↗
- 1995+0.08°CThe yearly AMV turns warm, six years before its 10-year mean ↗
- 2005+0.22°CThe record hurricane season of Katrina ↗
- 2024+0.38°CThe warmest year since 1933, with Hurricane Helene ↗
Any Long-Term Change? - No clear trend: flat since 1854 (p 0.75). Warm years - 13 in 1854-1939, 9 in 1940-2025. Cool - 9, then 10.
Built by 4 Billion Years On from NOAA’s ERSST v5: the North Atlantic (0‑60°N) sea surface temperature anomaly less the global ocean’s, against 1991-2020 (Trenberth & Shea 2006). Years beyond ±0.2°C are shaded.
The AMV's Partners
The AMV and Its Partners - Every Year Since 1950
The AMV each year beside the PDO, the NAO’s winters 15 years before and the AMOC - where the bars line up with the shaded years, the partner moves with it.
- PDOClear Link
Both Ways, Debated
Jan-Dec · r −0.53 · opposite way in 8 of 12
- NAO, 15 years beforeDecadal Link
Drives It · Years Ahead
10-year winter mean · r +0.85 · decade to decade
- AMOC (RAPID)Too Short to Measure
Thought to Drive It · Years Ahead
Against its mean, from 2005 · record too short to measure
| Partner | SONDJFMAMJJA | vs AMV | In Strong Years | Link |
|---|---|---|---|---|
| AMV | The reference - the AMV’s yearly mean | |||
| PDOBoth Ways, Debated | r −0.53 | Opposite way in 8 of 12 | Clear Link | |
| NAO, 15 years beforeDrives It · Years Ahead | 10-year winter mean | r +0.85 | Decade to decade | Decadal Link |
| AMOC (RAPID)Thought to Drive It · Years Ahead | - | Record too short to measure | Too Short to Measure | |
AMVJan-Dec, °C
PDOJan-Dec
NAO, 15 years before10-year winter mean
AMOC (RAPID)Against its mean, from 2005, Sv
PDO. The two oceans’ slow swings have run opposite since 1950 - together, a Warm AMV and a cool PDO brought the great US droughts (Zhang & Delworth 2007).
Opposite way in 8 of 12 strong AMV years · r −0.53PDO Tracker ↗
NAO. Strong NAO winters are thought to spin up the AMOC, warming the Atlantic a decade or more later (Delworth et al. 2016).
Same way in 0 of 0 strong AMV years · r +0.85NAO Tracker ↗
AMOC. The overturning is thought to pace the AMV over decades, but RAPID has measured it only since 2004 - about 20 years against a 60-80 year swing, too short to measure the link. Its sharp dip in 2009-10 came with a strongly Negative NAO winter (McCarthy et al. 2012) (Zhang et al. 2019).
Same way in 0 of 0 strong AMV yearsAMOC Tracker ↗
The AMV is the site’s index from NOAA ERSST v5, each calendar year’s mean. The PDO is NOAA NCEI’s index over the same year; the NAO is NOAA CPC’s Dec-Feb index, averaged over the ten winters ending 15 years earlier, and compared with the AMV’s own 10-year mean. The AMOC is the RAPID array’s yearly mean at 26.5°N (full years from 2005), against its own mean, in Sverdrups - shown, not correlated.
“vs AMV” is the correlation over every complete year since 1950. “In strong years” counts the years with the AMV beyond ±0.25°C in which the partner was on the expected side. A Clear Link is r 0.4 or more at p < 0.01. The NAO’s link is between 10-year means, with only a handful of independent decades, so it has no p-value - a Decadal Link is r 0.4 or more.
The arrow under each partner is the direction the science gives, with its source in How Each Partner Works; the Link badge is what this record measures. The two can differ - a one-way influence can still be faint by the time it reaches the AMV, so its verb follows the badge (Drives It, Partly Drives It, Tilts It).
This season (2026) is provisional and not in the statistics - AMV Jan-Sep so far, PDO Months so far.
Strongest Years
Strongest AMV Years - and What They Brought
Every year with the AMV beyond ±0.25°C since 1950, strongest first, with the Sahel’s and the Caribbean’s rainy seasons and Texas’s year.
Rain as a share of its 1961-1990 normal.
The AMV is the site’s index from NOAA ERSST v5, each calendar year’s mean. Rain is GPCC as a share of each place’s 1961-1990 normal - Senegal, Chad and Haiti over July-September, Texas over the year. Typical is the mean over every year of that kind in the list. Warm and cool years cluster in the phases (cool through the 1970s-90s, warm since), so a year’s rain also carries the decade’s.
AMV-Driven Places
AMV’s Pull by PlaceHow Much the AMV Drives Each Place
Each place’s strongest 3-month link to the Atlantic Multidecadal Variability since 1950, beyond ENSO’s share and the world’s warming - around the Atlantic.
Most AMV-Driven Places
Strongest link, either measure
One entry per country, its strongest part. Figures - the share of each season’s swing the AMV explains beyond ENSO’s share and the world’s warming. Tap a place for its figures, the arrow for the rest of the country.
Typical country’s share explained · strongest place
Figures are the share of a place’s year-to-year swing in that season that the AMV explains (r², 1950-present), the best of its twelve 3-month windows, after three things are taken out of both: the long-term trend, ENSO’s share, and the world’s warming (the global land-and-ocean temperature over the same months, as the index itself takes the global ocean out of the North Atlantic). That last step matters here: since 1950 the warming has run flat then steep, the same shape as the AMV’s one cool-then-warm swing, and without it most places looked AMV-driven. A Strong or Moderate tier also needs a second window to agree.
The AMV moves over decades, so its years are not independent: a warm phase is a block of consecutive warm years. Every tier and every phase count here is therefore judged on the effective sample size (Bretherton et al. 1999), so the same badge means the same confidence as on the other driver pages. In practice, once the warming is taken out, what is left of the AMV keeps little memory from one year to the next and a place’s rainfall almost none, so the effective count stays close to the full one and the test rarely moves a tier. Even so, one cool-then-warm swing is a thin record: Europe’s summers warm with a Warm AMV in the right season and direction, but too weakly over one swing to rank, and the split between the AMV and the clean-up of air pollution since the 1980s (Booth et al. 2012) is uncertain for Europe and the Sahel alike.
Ranked inside the AMV’s documented reach - Europe and the UK, North and Central America with the Caribbean, northern South America and West Africa with the Sahel (Zhang & Delworth 2006, Knight et al. 2006, Sutton & Dong 2012); places elsewhere can still be found in the search. With one cool-then-warm swing in the record, these are tendencies over decades, not a forecast for any season. A Warm or Cool AMV season is a 3-month mean of ±0.15°C or beyond. Sources are the site’s AMV from NOAA ERSST v5, Copernicus ERA5 temperature and GPCC rainfall by country and US state.
Contains modified Copernicus Climate Change Service information 2026.
Regional Impacts
Where Does the AMV Bite Hardest?
What each AMV phase usually brings around the Atlantic - over decades, most clearly in the Sahel’s rains and the hurricane seasons.
Warm AMVToday
Cool AMV
West African Sahel
Wetter
The monsoon reaches further north - stronger July-September rains.
Drier
The 1970s-80s Sahel drought came in the last Cool phase (Zhang & Delworth 2006).
Atlantic hurricanes
More, stronger
Warm seas and weaker wind shear - busier seasons since 1995 (Goldenberg et al. 2001).
Fewer
The quieter seasons of about 1970-1994.
Caribbean
Wetter
A wetter July-September rainy season - Haiti’s rain moves closely with the AMV.
Drier
Weaker rainy seasons.
US South and Plains
Drier · warmer
Drought more likely - with a Cool PDO, the 1930s and 1950s droughts (McCabe et al. 2004).
Wetter
Fewer long droughts.
Northern Brazil
Drier
The tropical rain belt shifts north, away from north-east Brazil (Knight et al. 2006).
Wetter
The rain belt sits further south.
Europe in summer
Warmer
Warmer summers, wetter in the north-west (Sutton & Dong 2012) - a weak link in the site’s data.
Cooler
Cooler summers.
Tendencies over decades, not forecasts.
From Zhang & Delworth (2006, GRL), Goldenberg et al. (2001, Science), McCabe et al. (2004, PNAS), Knight et al. (2006, GRL) and Sutton & Dong (2012, Nature Geoscience), checked against the site’s own measured links (Strongest AMV Years).
Explore
Where Next
- NAO Tracker
North Atlantic Oscillation - a key driver of Europe's winds, storms and winter weather
- ENSO - El Niño / La Niña
The Pacific now and forecast - RONI, NOAA strength odds and likely impacts
- Global Climate Update
Whole-planet temperature, CO₂, sea ice and ENSO in one monthly update
- Coral Bleaching Tracker
Daily heat stress for 214 reef regions, bleaching alerts and the NOAA four-month outlook