How El Niño and Volcanoes Rewire the Climate Web ✎

Original: "Global teleconnectivity structures of the El Niño-Southern Oscillation and large volcanic eruptions -- An evolving network perspective"
arXiv:1711.04670 · 2017-11-10 · arXiv non-exclusive · independent analysis · 1 min
Climate scientists have found that El Niño links distant corners of the Earth, while volcanoes only act close to home.

Every day, temperatures at different points on Earth fluctuate, and often these changes coincide even over great distances. Climate scientists connected such points with lines, as if stretching a web across the planet. In normal years, the web is neatly divided into sections, but when El Niño occurs in the Pacific Ocean, the lines break and new ones emerge that stretch across the entire globe.

To build such a web, the researchers used temperatures from 1948 to 2015 at 10,242 points. For each pair of points, they calculated how strongly their temperatures vary together, keeping only the strongest correlations, just 0.5% of all possible pairs. Then they measured how densely the lines cluster into knots and how far they reach.

El Niño is when waters off the coast of Peru become unusually warm. La Niña is the opposite: unusually cold.

During strong El Niño and La Niña events, the web loses its usual regional breakdown. Instead, long-distance links appear, for example between the eastern Pacific and Indonesia or West Africa. But after the 1991 eruption of Mount Pinatubo, connections became denser only in one region and did not extend far.

This way of looking at climate helps us understand when a change affects the whole planet and when it remains local. If a computer climate model reproduces the same rewiring of the web, it can be trusted more.

The network was built from only the 0.5% strongest correlations, meaning that out of more than 52 million possible point pairs in each window, only about 260,000 links remained.

c_{ij} = \frac{\langle T_i T_j \rangle}{\sqrt{\sigma_i^2 \sigma_j^2}}
Measure of linear relationship between temperature time series at nodes i and j
Q = \frac{1}{2m} \sum_{ij} \left( A_{ij} - \frac{k_i k_j}{2m} \right) \Delta_{ij}
Indicates how much the network splits into dense communities compared to a random network
Concepts
ENSO El Niño-Southern Oscillation Teleconnection El Niño La Niña climate network Jin-Timmermann model Multivariate ENSO Index spring predictability barrier natural variability equatorial wave propagation Southern Oscillation TAO/TRITON moorings seasonal cycle Warm water volume ERSST recharge oscillator model interannual variability seasonal forcing UAH satellite record