A Web of Temperatures Tells the Story of El Niño ✎

Original: "Disentangling different types of El Niño episodes by evolving climate network analysis"
arXiv:1310.5494 · 2013-10-21 · arXiv non-exclusive · independent analysis · 1 min
Scientists linked points on Earth into a network based on weather similarities and found that different types of El Niño leave distinct traces in this web.

If two distant cities experience warming and cooling at the same time, an invisible thread can be drawn between them. Across the whole Earth, such threads number in the hundreds of thousands, forming a web. Each year, the web looks different. Scientists tracked how it changed from 1948 to 2009 and noticed: when El Niño intensifies in the Pacific, the web sharply contracts, but when El Niño is weak or of a different type, it remains nearly unchanged.

El Niño comes in two varieties: one warms the waters off the coast of South America, the other in the middle of the ocean. The web senses the difference: the first type makes it contract into dense clusters, while the second barely affects it.

Volcanoes also influence the web, but not immediately. After the eruption of Mount Pinatubo in 1991, the threads only rearranged several months later, because the ash and gases from the volcano spread slowly through the atmosphere, altering temperatures unevenly in different places. It turns out that the shape of the web can distinguish El Niño from a volcanic signature and even reveal which type of El Niño has arrived.

The edge density in the network is only 0.5%, meaning that out of more than 52 million possible node pairs, only about 262,000 are connected. Yet the network still captures global climate events.

A_{ij} = \Theta(W_{ij} - W^*) - \delta_{ij}
A link between two points on Earth arises when their temperature series correlate stronger than a given threshold.
T = \frac{\sum_{i,j,k} A_{ij} A_{ik} A_{jk}}{\sum_{i,j,k, j \neq k} A_{ij} A_{ik}}
Indicates how much links in the network cluster into dense groups.
Concepts
climate network ENSO seasonal cycle El Niño-Southern Oscillation El Niño La Niña Jin-Timmermann model network natural variability TAO/TRITON moorings Multivariate ENSO Index NOAA CPC recharge oscillator model interannual variability equatorial wave propagation ERSST spring predictability barrier Teleconnection climate neural networks