El Niño Does Not Dry the Planet: Global Humidity Rises with Warming Tropics

New analysis of 60+ years of data reveals a predictable physical sequence and shows that global tropospheric humidity rises, not falls, during El Niño events.

Special acknowledgement to Dr. Henri Masson, University of Antwerp, Belgium, for his assistance with the extensive data analytics presented in this article.

This study is published at a critical moment to counter the widespread pseudoscientific alarmism surrounding the anticipated 2026/27 El Niño event. While media outlets and many social commentators warn of an “unprecedented” super El Niño triggering global mega-droughts, this analysis of six decades of NCEP Reanalysis data demonstrates a far more orderly and predictable physical sequence.

Using first-derivative (rate-of-change) analysis, the paper reveals a consistent ENSO “heartbeat” of the Pacific Trade Wind deceleration that leads tropical sea surface warming by ~3 months, which in turn drives increases in global specific humidity.

This single, robust observation directly contradicts claims of impending global drought.

Rather than drying the atmosphere, El Niño events make the global troposphere measurably moister.

By quantifying the timing and causal chain of ENSO teleconnections, this paper provides an evidence-based counter to sensationalist narratives and reminds readers that El Niño is a well-understood, repeating physical process — not an existential climate catastrophe.

ABSTRACT

The global tropics experience a recurring, quasi-periodic disruption to their normal seasonal energy state. These disruptions — best known as El Niño and La Niña (ENSO) — are set in motion by oscillations in the speed of the Pacific Trade Winds (PTWS) and the complex feedback responses they trigger. When Trade Winds slow, a domino effect unfolds: global tropical sea surfaces warm, the atmosphere spins faster, moisture rises, and temperatures across the planet respond in sequence. This study measures and quantifies the average time delays — the lags — between each step in that chain. Using cross-correlation statistics applied to first-derivative (rate-of-change) time series, we show that: global tropical sea surface temperature (GTSST) peaks about 3 months after Trade Wind deceleration; global atmospheric angular momentum (GAAM) peaks 3 months after GTSST; global specific humidity (GSH) at 700 mb peaks 3 months after GAAM; and global lower tropospheric air temperature (GAT) peaks 9 months after the initial Trade Wind slowdown. The methods employed — 12-month rolling averages combined with month-over-month differencing — strip away the seasonal cycle and reveal the underlying ENSO heartbeat.

1. Introduction

After the seasonal cycle of the sun, the El Niño–Southern Oscillation (ENSO) is the single most powerful driver of year-to-year climate variability on Earth (Pavlakis et al., 2008). While ENSO is born in the tropical Pacific, its effects ripple across every ocean basin and every continent. When headlines in 2026 describe a coming El Niño as “unprecedented,” they are invoking a phenomenon with deep scientific roots — one that can be tracked, timed, and understood through the physics of ocean-atmosphere interaction.

ENSO’s signature is felt through a chain of interconnected changes. The Pacific easterlies slow, warm water piled up again Indonesia spreads across the equatorial Pacific towards Latin Ameriaca, and sea surface temperatures (SST) rise across the Indian and Atlantic tropics as well through a process called teleconnections and eventually the whole global atmosphere responds. The reverse is true during La Niña: Trade Winds strengthen, cold water upwells, the tropics cool, and the atmosphere contracts.

Scientists track this continuum of states using a six-variable tool called the Multivariate ENSO Index (MEI), developed by Wolter and Timlin (2011). The MEI captures sea level pressure, SST, surface air temperature, cloudiness, and wind components to provide a single number summarising the state of the tropical Pacific at any given moment — making it one of the most comprehensive single-number summaries of ENSO available.

The physics behind ENSO’s global reach comes down to energy transport.

The tropics and sub-tropics receive more solar energy than they can radiate back to space — they run an energy surplus.

Higher latitudes do the opposite, radiating more than they receive. The atmosphere and ocean are the transport systems that move this excess heat poleward. When ENSO disrupts the tropical energy state, it alters the efficiency of that transport and the effects cascade outward with measurable time lags (Trenberth et al., 2002).

While previous studies have used anomalies or index values to track ENSO, this study takes a different approach: treating the first derivative (month-over-month rate of change) of the Pacific Trade Wind Speed (PTWS) as the key signal.

This highlights the transitions — the accelerations and decelerations — rather than the absolute state, and applies cross-correlation statistics to pin down the timing of each downstream response like a branching domino chain. The result is a sharper picture of the ENSO domino sequence from first cause to final atmospheric response.

Section 2 describes the datasets and methods. Section 3 presents results across five pairs of variables. Section 4 discusses the findings in the context of prior literature and draws conclusions relevant to the ongoing debate about ENSO predictability and the 2026/27 outlook.

2. Materials and Methods

2.1 Data Sources and Selection Criteria

Much of the primary data come from NCEP Reanalysis, which is one of the most widely used, long-running atmospheric datasets in climate science, covering 1948 to the present. The following time series were extracted:

● Regional and global tropical sea surface temperatures (SST): latitudes 5°N to 5°S for tropical Pacific; 90°N to 90°S for global parameters

● Pacific Trade Wind Speed (PTWS) at 850 mb: longitudes 240°W to 90°W, latitudes 5°N to 5°S

● Global Specific Humidity (GSH) at 700 mb: full globe (360°W to 0°W, 90°N to 90°S)

● Global Atmospheric Angular Momentum (GAAM): sourced from NOAA Earth System Research Laboratory (ESRL)

● Global lower tropospheric air temperature: Remote Sensing Systems (RSS) TLT (Temperature of the Lower Troposphere) and University of Alabama Huntsville (UAH) TLT satellite records

2.2 Data Processing

The processing chain is straightforward and applied identically to every variable:

1. Step 1 — Seasonal detrending: Apply a 12-month rolling average to remove the seasonal cycle. This acts as a low-pass filter, leaving only the interannual and longer-period signals.

2. Step 2 — First derivative: Take the month-over-month rate of change of the smoothed series. This converts the anomaly time series into a rate-of-change time series, highlighting transitions and peaks rather than absolute levels.

3. Step 3 — Cross-correlation: Apply cross-correlation statistics between paired variables to identify the time lag (Tau) at which correlation is maximised. Cross-correlation calculates the correlation coefficient between two time series at progressively shifted time offsets and identifies the shift that produces the strongest match.

3. Results

The results are presented in five subsections, each examining a specific pair of variables. The goal in each case is the same: determine how many months one variable leads or lags the other, and how tightly they are coupled.

3.1 Global Tropical Sea Surface Temperatures versus the Pacific Trade Wind Speed

The tropical oceans are connected. What happens in the Pacific doesn’t stay in the Pacific. ENSO-like warming and cooling patterns develop simultaneously in the Indian Ocean and the tropical Atlantic — a phenomenon called teleconnection — and these three ocean basins warm and cool in rough synchrony during ENSO events (Alexander et al., 2002; Wang, 2004; 2006). Figures 1a and 1b compare the tropical Indian Ocean sea surface temperature (SST) anomalies to those in the Easter Pacific and Atlantic regions.

Note the clear temporal coherence in changes in SST - see a recent article I wrote that highlights the atmospheric circulation dynamics behind this tropical teleconnection.

Figures 1a and 1b Comparisons of the seasonally detrended (12 month average) anomalies (1950 – 1980 baseline) of the Indian versus the eastern Pacific (1a) and the Atlantic (1b) tropical (5N – 5S) sea surface temperatures (SST) from NCEP Reanalysis from 1950 to 2013.

When we look at what drives these changes, the zonal Pacific Trade Wind Speed (PTWS) emerges as the initiating signal. Figures 2a and 2b show that decelerations of the Trade Winds — the winds slowing down (i.e., positively trending anomaly) — precede growth in global tropical sea surface temperature (GTSST) by approximately 3 months. Here, GTSST is the average of SSTs in full circumference along 5N to 5S.

The decelerating Trade Winds is the trigger; warming tropical seas are the first response.

Figures 2a and 2b Comparisons of the seasonally detrended (12 month average) anomalies of the tropical Pacific Trade Wind Speed (240W – 90W x 5N – 5S, 850 mb) versus the global average tropical sea surface temperature (5N – 5S) from NCEP Reanalysis (2a), and their associated monthly rates of change (first derivatives) time series (2b) from 1950 to 2013.

Figure 2c Cross correlation statistics of the average lag between the peak monthly rates of change of the global tropical sea surface temperature versus the Pacific Trade Wind Speed (Figure 2b) over the period 1950 to 2013. The cross-correlation shows that the peak in dGTSST/dt lags the peak in dPTWS/dt by approximately 3 months.

3.2 Global Tropical Sea Surface Temperatures versus Global Specific Humidity

▶ Key Finding — Challenging the Drought Narrative

Here is where media narratives about El Niño-driven drought deserve serious scrutiny. When El Niño events develop — driven by warming tropical sea surfaces — global average specific humidity at 700 mb (the base of the free troposphere, roughly 3 kilometres altitude) rises measurably.

This means the atmosphere as a whole becomes more moist during El Niño events, not drier.

The widespread alarm about tropical drought during El Niños reflects regional redistributions of rainfall — some areas do dry out — but the global atmospheric moisture budget moves in the opposite direction to what the drought narrative implies. Global moisture at 700 mb tracks GTSST with an approximate 3-month lag, and the relationship is robust across the full 1948–2013 record.

Figures 3a and 3b The seasonally detrended (12 month average) global average specific humidity (360W – 0W x 90N – 90S, 700 mb) and the comparison of the monthly rates of change of the of the global average tropical sea surface temperature (NCEP Reanalysis) versus the global average specific humidity at 700 mb from 1948 to 2013.

Figure 3c Cross correlation statistics of the average lag between the peak monthly rates of change of the global tropical sea surface temperature versus the 700 mb global average specific humidity (Figure 3b) over the period 1948 to 2013. The cross-correlation shows that the peak in dGSH/dt lags the peak in dGTSST/dt by approximately 3 months.

3.3 Global Specific Humidity versus Global Lower Tropospheric Air Temperature

The moisture signal leads the temperature signal by approximately 1 month. The largest pathway by which heat moves from the tropics to higher latitudes is meridional (south to north) atmospheric circulation via convective overturning of the Hadley and Ferrel cells (Trenberth et al., 2002).

As humid, warm air rises over the tropics and is transported poleward, it releases latent heat in the middle and upper troposphere, warming the lower troposphere at higher latitudes.

▶ Key Finding — Bjerknes Feedback and the Rise of Westerlies

Bjerknes Feedback — the positive reinforcing loop between weakened Trade Winds, reduced cold upwelling, warmer SSTs, and further Trade Wind weakening — amplifies this process during El Niño events.

The critical consequence is a strengthening of the westerlies at middle-to-higher latitudes, driven by accelerated Hadley Cell circulation and the Coriolis Effect. The result is measurably warmer lower tropospheric air temperatures globally, lagging the global average humidity peak by just one month.

I have seen no precedent for such a lag between specific humidity and the global tropospheric air temperature anomaly in the literature. Either this is an error or it may suggest the pulse of moisture emerging from the tropics precipitates out along lower latitudes than does the subsequent pulse in global tropospheric air temperature.

Figures 4a and 4b The seasonally detrended (12 month average) anomaly of the global average lower tropospheric air temperature (RSS TLT) and the comparison of its associated monthly rate of change versus the monthly rate of change of the global average 700 mb specific humidity (NCEP Reanalysis) from 1987 to 2013.

Figure 4c Cross correlation statistics of the average lag between the peak monthly rates of change of the global average 700 mb specific humidity (NCEP Reanalysis) versus the global average lower tropospheric air temperature (RSS TLT) from 1987 to 2013. The cross-correlation shows that the peak in dGAT/dt lags the peak in dGSH/dt by approximately 1 month.

3.4 Pacific Trade Wind Speed versus Global Tropical SST and Lower Tropospheric Air Temperature

Bringing the chain together: Trade Wind deceleration is the first domino. It leads global tropical SST (GTSST) by 3 months, and GTSST in turn leads the tropical (20N to 20S) lower tropospheric air temperature (TAT) by 3 months and the global lower tropospheric air temperature (GAT) by 6 months. The difference in lag factors is indicative of the time for warm - moist air masses to propogate poleward through Bjerknes Feedback, following the emergence of an El Nino State.

The changes in Walker Circulation and Bjerknes Feedback explain why the tropical atmosphere (TAT) warms before the global (GAT) atmosphere — tropics first, then poleward via a strengthened Hadley Cell and rising westerlies at mid-to-higher latitudes.

Figure 5a compares the TAT versus GAT anomaly, as measured using microwave sounding from orbiting satellites and 5b shows their first derivatives versus the first derivative of the Pacific Trade Wind Speed.

The peaking Trade Wind deceleration (blue) is clearly seen to lead the max rates of change of the TAT (tan) and GAT (brown) anomalies.

Figure 5c contrasts the cross correlations of the rates of change of the Trade Winds versus the GTSST (3-month), the TAT (6-month) and GAT (9-month) anomalies.

Figures 5a and 5b The average lower tropospheric air temperature anomalies for the tropics (20N – 20S) and the globe (UAH TLT) and the comparison of their associated monthly rates of change relative to the monthly rate of change of the Pacific Trade Wind Speed from 1979 to 2013.

Figure 5c Cross correlation statistics of the average lag between the peak monthly rates of change of the Pacific Trade Wind Speed (NCEP Reanalysis) versus the average global tropical sea surface temperature and the lower tropospheric air temperature anomalies for the tropics (TAT) and globe (GAT) from UAH TLT from 1979 to 2013.

3.5 Pacific Trade Wind Speed versus Global Atmospheric Angular Momentum

As Trade Winds decelerate, the atmosphere spins up. Global Atmospheric Angular Momentum (GAAM) — a measure of the rotation rate of the entire atmosphere — rises during El Niño events because the weakening of the equatorial easterlies reduces the drag that normally slows the atmosphere’s rotation and the strengthening of the middle latitude westerlies.

Bjerknes (1969) and subsequent studies identified this as a signature of accelerated tropical lattude Hadley Cell circulation, which strengthens the westerlies at mid-latitudes. The cross-correlation shows that the rate of change of the Pacific Trade Winds leads the rate of change of the global atmospheric angular momentum by approximately 3 months — the same lag as for GTSST — consistent with a single initiating mechanism driving both ocean and atmospheric responses simultaneously.

Figures 6a and 6b The seasonally detrended anomalies of the Pacific Trade Wind Speed (NCEP Reanalysis) and the global average atmospheric angular momentum (NOAA ESRL) and their associated monthly rates of change from 1959 to 2014.

Figure 6c Cross correlation statistics of the average lag between the peak monthly rates of change of the Pacific Trade Wind Speed (NCEP Reanalysis) versus the global average atmospheric angular momentum (NOAA ESRL) from 1959 to 2014. The cross-correlation shows that the peak in dGAAM/dt lags the peak in dPTWS/dt by approximately 3 months.

3.5 The Pacific Trade Wind Speed as a Leading Climate Indicator

Figure 7 summarises the full domino sequence. Reading from left to right, it shows the average lag (Tau, in months) of each climatic variable relative to the peak deceleration of the Pacific Trade Winds. The 1-month gap between the global specific humidity (GSH) peak and the global air temperature (GAT) anomaly peak reflects the latent heat release pathway: moisture rises, condenses, releases heat, and warms the lower troposphere. Tropics warm first; then the signal travels poleward via changing atmospheric circulation and rising westerlies.

Figure 7. Comparisons of the average lag (Tau in months) between the peak monthly rates of change of the global tropical sea surface temperature, the global atmospheric angular momentum, the global specific humidity at 700 mb, the tropical air temperature (20N – 20S, UAH TLT) and the global air temperature (UAH TLT), relative to the peak monthly rate of change of the Pacific Trade Wind Speed from NCEP Reanalysis.

4. Discussion and Summary

The results of this study tell a coherent story. Every major interannual climate variable examined — tropical SST, atmospheric angular momentum, specific humidity, and lower tropospheric air temperature — responds to changes in the Pacific Trade Winds in sequence, with predictable time lags that have remained consistent across more than six decades of data. This is not a new discovery — Bjerknes (1969) described the feedback loop bearing his name over fifty years ago — but this study provides a quantified, multi-variable, derivative-based empirical framework that places the Trade Wind deceleration as the measurable starting gun of the entire ENSO sequence.

When commentators describe the potential 2026/27 El Niño as “unprecedented,” the physically meaningful question is whether the initiating Trade Wind deceleration and its downstream lags are behaving differently from prior events. The evidence from this study suggests that the ENSO sequence itself — from Trade Wind slowdown to global temperature response — is a stable, repeatable physical process governed by Walker Circulation and Bjerknes Feedback dynamics. What varies between events is the amplitude of the Trade Wind anomaly, not the underlying mechanism. An event can be large without being structurally novel.

The specific humidity finding (Section 3.2) is particularly important in the current media environment. Global atmospheric moisture at 700 mb rises during El Niño — the atmosphere gets wetter globally, even as regional drought conditions develop in specific locations (e.g., tropical terrestrial environments). The conflation of regional drought signals with a global drying narrative is not supported by the 700 mb humidity data spanning 1948 to 2013.

Readers who follow ENSO news closely should treat with scepticism any claim that El Niño is making the world drier in aggregate: the tropospheric moisture budget says otherwise.

Equally important is what drives the tropospheric temperature response.

The warming of the lower troposphere during El Niño events is primarily a result of changed atmospheric circulation — altered Walker Circulation in the tropics feeds an accelerated Hadley Cell, strengthening the westerlies at middle-to-higher latitudes through Bjerknes Feedback. This poleward transport of heat and moisture is the dominant mechanism.

Understanding this helps explain both the timing lags and the spatial pattern of the global temperature response.

Finally, this study acknowledges that Trade Wind decelerations are themselves triggered by higher-frequency intra-annual atmospheric and oceanic westerly equatorial waves. At a quasi-decadal level, numerous publications have shown that up to 50% of El Niño events between 1900 and 2005 arise through non-linear phase locking with the 11-year solar cycle (White, 2003; 2008, Wang, 2004).

ENSO is not random — it is a structured sub-harmonic of the Earth’s seasonal cycle, amplitude-modulated by solar forcing and ocean memory (Jin et al, 1996; Stueck et al, 2013; Douglas, 2011). Understanding its mechanics is the most reliable antidote to the recurring media temptation to call every new event “unprecedented.”

References

● Alexander, M. A. et al, J. Climate, 15, 2205-2231, 2002.

● Angell, J. K. et al J. Geophy. Research 105, 11, 841–11,849, 2000.

● Brady, E. C. et al, J. Physical Oceanography, 24, 2658-2673, 1994.

● Bjerknes, J., Monthly Weather Review, 97 (3), 163-172, 1969.

● Bjerknes, J., Tellus, 18 (4), 8, 820-829, 1966.

● Bjerknes, J., Analysis of the rhythmic variations of the Hadley Circulation over the Pacific during 1963 – 67, Technical Report, ONR contract No. N0014-69-A-0200-4044 NR 083-287, 1969.

● Christy, J. R. et al, J. Climatology., 12, 2451–2473, 1999.

● Douglas, D. H., Physics Letters A, 376, 128–135, 2011.

● Eastman, R. et al, J. Climate, 24, 5914-5934, 2011.

● Fasullo, J. T., et al, J. Climate, 21, 2313-2325, 2008.

● Glantz, M. H., Currents of Change, Cambridge University Press, Cambridge, UK, 2001.

● Huang, J. et al, J. Climate, 27, 2656-2666, 2014.

● Jin, F. F. et al, Physica D, 98, 442, 1996.

● Jin, F. F. et al, J. Atmos. Sci., 54, 811-829, 1997.

● Lee, S. et al, J. Climate, 27, 6519-6525, 2014.

● Loeb, N. G., Climate MDPI, 6, 62, 2018.

● Mansbach, D. K., J. Climate, 20, 1555-1570, 2007.

● McLean, J. D. et al, J. Geophy. Research, 114, D09110-D09129, 2009.

● McPhaden, M. J, J. Climate, 21, 6070-6079, 2008.

● Navarro-Perez, E. et al, S. Afr. J. Mar. Sci., 19, 61-73, 1998.

● Pavlakis, K. G. et al, Atmos. Chem. Phys., 8, 5565-5577, 2008.

● Philander, S. G., El Nino, La Nina and the Southern Oscillation, Academic Press, 1990.

● Pinker, R. T. et al., J. Geophy. Research: Oceans, 122, 10, 7880–7896, 2017.

● Rosen, R. D., et al, Science, 225 (27), 411, 1984.

● Stefanick, M., J. Geophy. Research, 87, 428, 1982.

● Stuecker, M. F. et al., Nature Geoscience, 6, 540–544, 2013.

● Sun, D., J. Climate, 13, 3533-3550, 2000.

● Trenberth, K. E. et al, J. Geophy. Research, 107, 2002.

● Wallace, J. M., Monthly Weather Review, 109, 784-812, 1981.

● Wang, C. et al., Progress in Oceanography, 69, 239–266, 2006.

● White, W. B., J. Geophy. Research, 108 (C3), 30,70, 2003.

● White, W. B., Geophysical Research Letters, 35, L19607, 2008.

● Wang C., Geophysical Monograph Series, 147, 21–48, 2004.

● Wolter, K. et al, Inter. J. Climatology, 31 (7), 1047-1087, 2011.

● Zelle, H. et al, J. Physical Oceanography, 34, 643-655, 2004.