El Niño 2023: a year-long warning ✎

Original: "Very early warning of a moderate-to-strong El Niño in 2023"
arXiv:2301.10763 · 2023-01-25 · arXiv non-exclusive · independent analysis · 2 min
Two independent methods predicted the 2023 El Niño with 89% probability and estimated its strength at 1.49±0.37°C.

In May 2022, sea surface temperatures off the coast of Peru were a degree below normal. Nothing hinted at trouble, but two independent "seismographs" had already picked up a rumble from below: El Niño was coming in a year. That's the name for the warm phase of the Pacific oscillation, when a vast tongue of hot water spreads along the equator and redraws weather from Australia to Africa.

The first instrument monitors the ocean's "nervous system." Fourteen points in the central and eastern Pacific are linked to 193 points worldwide, and the strength of these connections shifts before an event. A year before El Niño, the links tighten, as if the sensors start whispering to each other more often. The second instrument measures the "temperature of chaos": it looks at how disorderly daily temperature fluctuations become in the key Niño3.4 region. It turns out that the higher this disorder in a given year, the stronger the El Niño the following year.

Both instruments raised the alarm almost simultaneously: between May 24 and June 7, 2022. In the entire observational record, there have been seven cases where they agreed, and all seven times El Niño did arrive.

The forecast came true. The 2023 El Niño began, and its strength, predicted as 1.49±0.37°C, proved close to reality. This event temporarily raised global temperature by about 0.2°C, making 2024 one of the hottest years in the instrumental record. The probability that the planet would for the first time breach the +1.5°C threshold relative to pre-industrial times was estimated at 21.1%.

Previously, a reliable El Niño forecast was only possible a few months ahead, and the event's strength was a matter of guesswork. Now, two independent methods, validated on historical data, allow us to warn of climate risks a year in advance. It's like learning to hear a storm approaching long before the first waves hit the shore.

The seismographs caught the tremor, and we had time to prepare. Next time the ocean starts whispering again, we'll already have time to reinforce levees, reallocate water, and save crops.

From 1981 to 2022, the climate network model issued 11 alarms, of which 8 were correct, and in the stricter version, all 8 alarms were correct.

SysSampEn(m,p,l_eff,γ) = -log(A/B)
SysSampEn is approximately equal to the negative natural logarithm of the conditional probability that two subsequences that are similar over m consecutive points remain similar over the next p points.
Concepts
ENSO El Niño El Niño-Southern Oscillation spring predictability barrier La Niña recharge oscillator model Multivariate ENSO Index ERSST Jin-Timmermann model Southern Oscillation relative ONI Warm water volume NOAA CPC Westerly wind burst TAO/TRITON moorings interannual variability North Equatorial Countercurrent natural variability seasonal cycle monsoon onset