The 4byo Climate Symphony
A compound teleconnection stack on a single 15-year timeline - ENSO, NAO, PNA, AMV, and the Polar Vortex - past and present with the official forecast. Also shows how the post-El Niño temperature floor - informally, the “ratchet” - has stepped higher after each major event, layered on the long-term warming trend. Toggle any driver in What-If mode to see how the combined picture shifts.
The 4byo Climate Symphony - Current State & Forecast
Click any panel to see how the drivers interact.
After each major El Niño, global temperatures decline during the La Niña that follows, but they rarely fully return to the pre-event baseline. Each event has left a higher temperature “floor” than the one before, producing the staircase shape shown by the gold line - though as the step-size card below shows, how big that step is varies a lot by event. The treads slope upward with the background anthropogenic warming rate (~0.027 °C/yr, derived from an OLS fit to the most recent 25 years of land+ocean data); most of each near-vertical step is that same background trend accumulating between events, with the El Niño peak as the moment it becomes visible rather than a separate cause.
† 2023–24 floor lift is provisional — post-event temperatures are still settling 12–24 months after the peak. These card values are approximate empirically-observed raw floor lifts (including trend); the forecast step model uses detrended coefficients derived from 24 El Niño events back to 1950 (see below).
What updates automatically: The observed temperature line, trend rate, staircase floors, step and peak coefficients, and forecast profiles are all dynamically derived each time new snapshot data arrives — no manual calibration. The ENSO forecast lines update when the NOAA and SNU CNN model outputs are refreshed.
The long-dashed staircases extend the floor into the forecast period: gold uses NOAA's official RONI outlook median peak; purple uses the SNU CNN peak (typically larger), producing a higher riser and post-event floor. Step height is estimated using a category-tiered coefficient derived from 24 historical events. The fine dotted lines (and amber shading) show the predicted month-by-month actual temperature: the ratchet floor plus ~0.19 × Niño 3.4 for the ENSO contribution, spiking above the floor during the event then settling to the new level.
Why doesn’t ENSO average out to zero on its own?
Even the small piece of the staircase not explained by the background greenhouse-gas trend doesn’t obviously come from a single, settled cause. A few possible contributing mechanisms have been proposed in the literature - each plausible, none of them the dominant story here (that’s still the AGW trend above), and the size of any one is genuinely debated:
1. Power asymmetry (positive skewness)
Super El Niños (1997–98, 2015–16, 2023–24) have spiked global temperatures more sharply than La Niñas have cooled them. La Niña cooling has a physical floor (set by how cold deep Pacific water actually is) while El Niño has no equivalent ceiling, which could leave a small residual warming signal averaged over many cycles. [1, 2]
2. Nonlinear atmospheric rectification
The atmosphere is thought to respond nonlinearly to ocean heat: El Niño’s warm water triggers large-scale cloud and circulation changes that may not fully reverse when La Niña flips the Pacific back, potentially leaving a small fraction of the shifted heat behind - sometimes called “ENSO rectification”. [2, 3]
3. Decadal dominance cycles
ENSO appears to oscillate through multi-decadal phases where one state dominates. During the early-to-mid 20th century (~1900–1941) El Niño was more frequent and powerful than La Niña, which may have added an independent multi-decadal warming contribution on top of the greenhouse signal. [1, 2]
Is ENSO getting more extreme?
The baseline frequency of ENSO events (~every 2–7 years) has not conclusively increased, but the intensity distribution has shifted. IPCC AR6 and WMO assessments note that episodes since 1950 have been demonstrably stronger than those in 1850–1950. Climate models project the frequency of extreme El Niño and La Niña events could roughly double (from once every ~20 years to once every ~10 years) by the end of the 21st century under high-emissions scenarios. [4, 5]
A separate trend is the rise of multi-year clusters: the 2020–23 “triple-dip” La Niña is an example of consecutive cool phases that research (published in Nature) links directly to intensified El Niños; global warming may be increasing the chance of such consecutive La Niña years as a rebound response. [5]
[1] IPCC AR6 WGI Ch. 2 & 3 - ENSO observed changes; WMO State of the Global Climate 2024.
[2] Foster & Rahmstorf (2011), Environ. Res. Lett. 6 044022 - annual-mean regression 0.10 °C/°C; empirical monthly peak ~0.18 °C per RONI unit used for the dotted profiles (derived via OLS from 23 events, 1950–2024). Floor step-lift coefficient by category: moderate ~0.034 · strong ~0.058 · very strong ~0.049 per RONI unit, derived from detrended pre/post-event floor analysis across 23 El Niño events (1950–2024); Pinatubo 1991–92 cooling window excluded. RONI (Relative Oceanic Niño Index) is NOAA’s official ENSO index since Feb 2026; coefficients and the forecast peak used to scale them are both on the RONI basis, so the projected floor step is not inflated by the raw-ONI warming background. Model forecasts (raw Niño 3.4) are converted to RONI by subtracting the observed ONI−RONI gap season by season (~0.5°C cooler; Stockdale 2026, ECMWF).
[3] Rodgers et al. (2021), Nature Climate Change - nonlinear ENSO atmospheric response.
[4] Cai et al. (2014 & 2015), Nature - doubling of extreme El Niño & La Niña frequency under warming.
[5] Geng et al. (2023), Nature - increased consecutive La Niña years linked to global warming.
NOAA NCEI land+ocean monthly anomaly. SNU CNN Niño 3.4 forecast (Ham et al. 2019, Nature).
Sources: NOAA NCEI (temp) · Niño 3.4 + ONI + RONI (ENSO obs) · NOAA RONI Outlook + SNU ACE Lab CNN (ENSO forecast) · NOAA CPC (NAO, PNA, AO) · NOAA PSL (AMV) · EM-DAT (extremes)
Combined Global Footprint
Climate is ENSO, NAO, AMV and the Polar Vortex acting together.
El Nino with -NAO Surprise low confidenceSPV has no clear live signal and is inferred from climatology. This downgrades confidence one level from the base pattern.
Stratosphere Strong vortex but surface NAO bucks the trend - watch for weakening.
1 pillar is currently neutral or unknown - we've matched to the closest signed combination and reduced the confidence rating accordingly.
Outcomes are precomputed from peer-reviewed teleconnection literature - deterministic lookup.
Sources: NOAA CPC ONI (ENSO) · NOAA CPC NAO index · NOAA PSL AMO · NOAA CPC Stratosphere Monitor (SPV).
Explore
Explore Climate Data
Data Sources
How the Climate Symphony is Built
Chart vs map: PNA appears in the charts but not the map - its footprint is a complex wave pattern across the North Pacific, not a simple region. SPV and AO represent the same polar system at different altitudes: AO is the surface measurement tracked in the charts; SPV is the stratospheric driver also shown on the map.
The Symphony is built from the same per-pillar snapshots that drive the individual tracker pages (ENSO, NAO, Polar Vortex), fused into a single multi-pillar forecast-stack JSON. Every number on the chart, the world map, and the pillar pills is sourced from the same artefact - they cannot disagree.
Observed (history) data
- ENSO / Niño 3.4 / ONI - NOAA Climate Prediction Center Oceanic Niño Index (3-month running mean ERSSTv5 SST anomaly, Niño 3.4 region) plus weekly OISST.v2.1. Monthly updates.
- NAO - NOAA CPC daily & monthly NAO index (500 hPa EOF-based, 1950-present).
- PNA - NOAA CPC daily & monthly Pacific-North American teleconnection index.
- AO (proxy for stratospheric polar vortex state) - NOAA CPC daily & monthly Arctic Oscillation index.
- SSW catalogue - Major Sudden Stratospheric Warming events from the Free University Berlin / NOAA SSW chronology.
- AMV - NOAA Physical Sciences Laboratory Atlantic Multidecadal Variability (unsmoothed monthly SST anomaly, North Atlantic basin).
- Global surface temperature - NOAA Climate at a Glance Global Land + Ocean monthly anomaly.
Forecast data & models
- ENSO - official NOAA CPC ENSO consensus probability forecast, 9 overlapping 3-month seasons (DJF, JFM, FMA, ...). Updated monthly. Confidence: high.
- NAO / PNA / AO - persistence-decay model:
value₀ · exp(-Δmonths / τ)with τ = 2 months. This honestly reflects how little operational seasonal skill these atmospheric indices have at lead. Confidence: moderate at 1 month, downgraded to low by 3 months and climatological beyond. - ENSO (CNN) - the SNU deep-learning model (Ham et al. 2019, Nature) trained on CMIP5 climate model output and verified against observations. Predicts Niño 3.4 SST anomaly up to 18 months ahead. Shown as the purple dashed line and shading on the ENSO and post-El Niño floor panels; the purple shading fills only the gap where the CNN projects a stronger El Niño than the NOAA consensus. Confidence: high at leads up to ~6 months, moderate beyond.
- AMV - persistence (carry-forward of the latest observed value). Justified by AMV being a decadal background phase with autocorrelation r ≈ 0.8 over a year. Confidence: moderate.
Operational seasonal NAO products (Met Office GloSea, ECMWF SEAS5, DePreSys) do exist with modest skill in some winters - when their data feeds are wired in they will replace the persistence-decay defaults for NAO and AO. The method field in the forecast-stack JSON already supports the swap.
Compound stacking matrix
The 16-cell matrix is an evidence-based lookup keyed by the sign (+ or -) of each pillar. Cell narratives are synthesised from the peer-reviewed teleconnection literature and the documented impacts maintained by NOAA CPC, the Met Office and ECMWF. Each cell carries region-specific impact text for the UK / NW Europe, central + eastern Europe, North America and the ENSO-impacted Southern Hemisphere regions.
The post-El Niño temperature floor
The amber dashed lines on the Global Temperature panel mark the post-El Niño temperature floor: the 12-month mean global anomaly for the year starting 6 months after each major event’s ONI peak. After the 1997–98, 2009–10, 2015–16, and 2023–24 El Niño events, temperatures never returned to the pre-event baseline during the following La Niña - each event left a higher floor than the one before in absolute terms (+0.13, +0.08, +0.21, and +0.32 °C respectively), though the size of that step varies widely by event. This is our own descriptive label for the pattern, not a separate physical mechanism: it reflects primarily the underlying anthropogenic warming trend - each El Niño spike rides a rising baseline, so even the La Niña cooling that follows still lands above the pre-event level. Once that background trend is removed, some events (2009–10 in particular) show next to no additional ENSO-specific step - see the FAQ below for the detrended figures.
A secondary contribution may come from El Niño-induced feedbacks - Arctic sea-ice albedo loss, ocean heat redistribution - that are not fully reversible on multi-year timescales. With a new El Niño now developing (NOAA CPC consensus probability >90 % by JJA 2026), the global temperature baseline can be expected to step higher again; the exact magnitude depends on event strength but historical analogues suggest a floor lift of roughly+0.1–+0.18 °C above the current post-2023-24 level (NOAA plume scenario ~+0.11 °C; CNN ~+0.16 °C).
Foster & Rahmstorf (2011, Environ. Res. Lett. 6 044022) provide the canonical framework for separating the ENSO signal from the underlying warming trend. Their analysis shows ENSO has a total range of ~0.39 °C on the global surface temperature record, with global warming lagging the MEI index by 2–5 months.
FAQs
Climate Symphony - Common Questions
What is the Climate Symphony?
The Climate Symphony is an interactive view of how Earth's major climate "drivers" stack together to shape weather and climate at any moment. It combines four pillars - ENSO (El Niño / La Niña), the NAO (North Atlantic Oscillation), the AMV (Atlantic Multidecadal Variability) ocean background, and the stratospheric polar vortex (SPV) - into a single compound state. The same page also shows the PNA, AO and global surface temperature on one timeline, so you can see past, present and the official forecast in one place.
Why look at all the climate systems together rather than one at a time?
Because no single driver acts in isolation. A +NAO winter feels very different depending on whether ENSO is in El Niño or La Niña, whether the AMV ocean background is warm or cool, and whether the stratospheric polar vortex is strong or disrupted. The four pillars stack into 16 possible sign combinations and each combination has a documented signature in the historical record. Looking at one pillar at a time hides the most important question: what is the combined state right now, and what is it expected to be next season?
What does the compound stacking matrix show?
The matrix is a 16-cell lookup keyed by the sign (+ or -) of each pillar - ENSO, NAO, AMV and SPV. Each cell carries an evidence-based narrative for what that combined state means for the UK and northwest Europe, central and eastern Europe, North America, and the ENSO-impacted Southern Hemisphere regions (Australia, India / SE Asia, South America, Southern Africa, East Africa). The system selects the cell that matches the live observed state, the forecast state for any upcoming CPC season, or any combination you build in What-If mode.
Where does the forecast come from?
The fused multi-pillar forecast uses different methods per pillar because operational seasonal skill varies enormously. ENSO uses the official NOAA CPC consensus probabilistic forecast (high skill at 1-9 months). NAO, PNA and AO use a persistence-decay model (last observed value with a 2-month e-folding decay toward climatology) - this honestly reflects the fact that seasonal atmospheric skill is genuinely low. AMV uses persistence (carry-forward of the latest value), valid because the AMV is a decadal background phase with autocorrelation r ≈ 0.8 over a year. Each pillar carries its own confidence flag (high / moderate / low / climatological) so the front-end is honest about what is well-known and what is a baseline guess.
Why does the NAO forecast often look near zero?
Because that is the honest answer at seasonal lead. The atmospheric drivers (NAO, PNA, AO) have very little operational predictive skill beyond a few weeks. Persistence-decay correctly collapses the forecast to climatology (≈ zero) within a couple of months. Showing a confident +NAO or -NAO signal 6-12 months out would be misleading. Operational seasonal NAO products (Met Office GloSea, ECMWF SEAS5, DePreSys) do exist with modest skill in some winters - we plan to swap them in for the persistence-decay default once their data feeds are wired up.
What is the difference between Now, Forecast and What-If modes?
Now: the live observed state of all four pillars from the latest snapshot of each index. Forecast: step through the next ~12 months of CPC seasons; the matrix and the world map update to show the dominant fused stack for each season. What-If: toggle any pillar to any sign you want and see which historical analogue (and which combined regional impacts) you have just constructed - useful for asking questions like "what would a +NAO winter look like if AMV were cool?"
Why do the chart and the map sometimes disagree on the NAO sign?
They no longer do. The map dot colour is driven by the numeric forecast value with the same dead-band the chart uses (NAO / PNA / AO ±0.5 standardised anomaly, AMV ±0.10 °C). A near-zero climatological forecast now reads as a neutral grey dot on the map AND a near-zero dashed line on the chart - matching the value shown on the pillar pills.
What is the AMV and why does it matter?
AMV stands for Atlantic Multidecadal Variability - the slow basin-wide warm/cool background phase of the North Atlantic sea surface temperatures, with a typical period of 60-80 years. A warm AMV tilts the NAO negative on average, boosts Atlantic hurricane activity, and adds ~0.05 °C to global mean temperature; a cool AMV does the opposite. AMV phase is one of the main reasons certain decades feel persistently stormier or drier than others, even when ENSO is neutral.
What is the stratospheric polar vortex (SPV) and how is it linked to weather?
The SPV is the ring of fast westerly winds that circle the North Pole each winter in the stratosphere (~30 km up). When it is strong, midlatitude weather tends to stay zonal and mild (+NAO bias). When it weakens or is disrupted (a sudden stratospheric warming, or SSW), the signal couples downward over 2-4 weeks and pushes the surface into a -NAO regime - cold spells across northern Europe and the eastern US. The Symphony page uses the AO (Arctic Oscillation) as an observational proxy for the SPV state.
How often is the data updated?
All observed pillars are refreshed monthly (and most also have daily updates). The CPC ENSO probability forecast updates monthly. The fused forecast-stack JSON is rebuilt whenever any underlying pillar snapshot changes, so the chart, the map and the pillar pills always carry the same numbers.
Why call it the Climate Symphony?
Because no single driver tells the whole story. Each pillar is one instrument; the weather we experience is the chord they play together. The Symphony page is the score - the live arrangement and the forecast for the next few months - so you can see the whole orchestra at once instead of listening to one violin in isolation.
Why do global temperatures stay high after an El Niño ends?
After each major El Niño event, global temperatures decline during the La Niña that follows - but they usually don't return to the pre-El Niño baseline. Each event has left a higher temperature "floor" than the one before, producing what we call the "staircase" (or informally, "ratchet") pattern in the global temperature record - our own descriptive label for the pattern, not a separate physical mechanism. The main reason is simply the underlying anthropogenic warming trend: each El Niño spike rides a rising baseline, so even the La Niña cooling that follows still lands above the pre-event level - the same reason any hot year sits above the equivalent year a decade earlier. Measured against RONI (the Relative Oceanic Niño Index - NOAA CPC's primary ENSO index since Feb 2026, which isolates the ENSO signal from background tropical warming) to see what, if anything, is left over once that background trend is removed, the 1997-98, 2009-10, 2015-16, and 2023-24 events lifted the post-event floor by +0.17, essentially flat (-0.04), +0.12, and +0.25 °C respectively. In other words: one of these four "steps" (2009-10) isn't really a step once the trend is stripped out, which is a reminder that this is a loose, event-by-event pattern rather than a fixed-size mechanism that fires every time. A secondary, less certain contribution may come from feedbacks such as Arctic sea-ice albedo loss or ocean heat redistribution, which some research suggests could make part of a step only partially reversible on multi-year timescales - see the Symphony page's "Why doesn't ENSO average out to zero on its own?" panel for the (debated) candidate mechanisms. The gold staircase line on the Global Temperature chart shows these post-event floors: the treads slope gently upward with the background warming trend (~0.027 °C/yr, derived from a 25-year OLS fit to land+ocean data), and the near-vertical steps mark where each El Niño event's floor became visible. All staircase parameters - trend rate, step coefficients, and forecast profiles - are dynamically derived from 23 historical El Niño events back to 1950, classified by RONI (Van Oldenborgh et al. 2021, ERL 16 044003; L'Heureux et al. 2024, J. Climate 37(4)). As of mid-2026, a new El Niño is developing with close to 100% probability; the forecast step is +0.15 to +0.16 °C above the post-2023-24 level, depending on peak intensity (NOAA plume vs CNN model).