Why the Ocean Turns Off the Lights in America ✛ ✎

Original: "Spatio-temporal Patterns between ENSO and Weather-related Power Outages in the Continental United States"
arXiv:2406.14474 · 2024-06-20 · arXiv non-exclusive · independent analysis · 1 min
When the Pacific Ocean turns colder than usual, power outages become more frequent in the United States.

The Pacific Ocean acts like a conductor, orchestrating weather across the United States. When ocean waters turn colder than usual, La Niña kicks in, and lights go out more often across America. Researchers analyzed 24 years of data and spotted a clear pattern: the ocean shifts the weather, and the weather wreaks havoc on power lines and transformers.

La Niña is the cold phase of the largest climate oscillation in the tropical Pacific. Waters off the coast of South America become unusually cold, which alters weather patterns worldwide.

The ocean issues its commands in three ways. In the western U.S., La Niña brings cold and heavy snowfall; in the Southeast, it brings heat; and in the central states, it brings downpours. Snow topples trees onto power lines, heat overloads air conditioners, and downpours flood substations. So outages become more frequent. The connection isn't immediate: the ocean first changes the weather, and only months later does the power grid start to stumble.

By the end of the century, things could get much worse. If greenhouse gas emissions remain high, by 2081–2100, heat-related outages in the eastern U.S. could increase nearly 26-fold. A power grid built for the old climate would find itself in a new one. But now, with ocean forecasts in hand, energy planners can prepare in advance for tough seasons.

In spring 2018, a storm in California brought heavy snowfall and cold, leaving 10,898 customers without power and causing a loss of 38 MW of demand. And in March 2023, a series of tornadoes with extreme precipitation blacked out about one million customers across more than ten states.

x(t) = a + bt + \epsilon(t)
The original series is decomposed into a trend and a random residual
x'(m,n) = \hat{x}(m,n) - \frac{1}{24}\sum_{n=1}^{24}\hat{x}(m,n)
Deviation from the average for a given month over 24 years
R = \frac{\sum_{t=1}^{T-k}[\tilde{x}(t+k)-\bar{x}][\tilde{y}(t)-\bar{y}]}{\sqrt{\sum_{t=1}^{T}[\tilde{x}(t)-\bar{x}]^2\sum_{t=1}^{T}[\tilde{y}(t)-\bar{y}]^2}}
Measure of the association between two series when one is shifted relative to the other by k months
Concepts
ENSO La Niña spring predictability barrier El Niño-Southern Oscillation Southern Oscillation recharge oscillator model El Niño Multivariate ENSO Index Jin-Timmermann model NOAA CPC ERSST Westerly wind burst TAO/TRITON moorings equatorial wave propagation seasonal cycle seasonal forcing Indian Ocean capacitor effect interannual variability Warm water volume North Equatorial Countercurrent