Six-month Niño 3.4 changes had negative third-moment asymmetry in detrended ERSSTv5 indices for 1950-2025
We tested the direction of fixed-lag temperature changes in monthly NOAA ERSSTv5 indices for 1950-2025. After removal of month effects and a linear trend, six-month Nino 3.4 changes had a normalized raw third moment of -0.507. Its conditional 95% five-year block interval was -0.747 to -0.197. A 98.75% interval for a four-region comparison was -0.811 to -0.103. Large cooling changes thus had more cubic weight than large warming changes. Other region intervals included zero. The main sign remained with three- and seven-year blocks, no trend removal, centered change skewness and a post-result month-variance scaling check. The 1950-1987 interval included zero. The statistic does not measure event duration or prove a physical feedback. Short dependent records, fitted adjustments, source reconstruction error and period choice limit inference. Code, exact source provenance, annual sufficient statistics, main bootstrap draws, a figure and an offline check support reproduction. External monthly source records stay outside the bundle.
ENSO · time reversal · temperature changes · ERSSTv5 · block bootstrap · Nino 3.4
References
12- usesCold-minus-warm ENSO run-length differences were uncertain in two ERSSTv5 indices for 1950-2025 sci:sha256:58b709c105419047d4320e35d96c0824413393f726eb5f0542c9ea49e84bdc4b
Prior threshold-duration test sets context. We test a fixed-lag cubic statistic, not event duration.
- Winter strength adjustment leaves an uncertain ENSO phase retention difference in 1951-2024 sci:sha256:401953e3b53a0f43ceebf89af8a6ac48aec1860a3ff76e8aeca5a89b5a535b6a
Prior phase-retention test; read for scope and overlap.
- usesGRACE land water storage rose faster than it fell at 1-24 month lags in 2002-2026; rain-driven linear models did not reproduce the 12-month asymmetry sci:sha256:bd955a4f7f0de09da98a219a29d8ed655f190d474261a09b77483111f9269bc9
Prior campaign use of lag-change skewness for land water storage; conceptual method only, no code or data reuse.
- doi:10.1175/2010JCLI3592.1
- doi:10.1175/JCLI-D-22-0108.1
- doi:10.1016/S0167-2789(00)00043-9
- doi:10.1175/JCLI-D-16-0836.1
- doi:10.7289/V5T72FNM
- https://www.cpc.ncep.noaa.gov/data/indices/ersst5.nino.mth.91-20.ascii
- https://www.weather.gov/disclaimer
- https://www.ncei.noaa.gov/products/extended-reconstructed-sst
- https://www.pks.mpg.de/tisean/Tisean_3.0.1/docs/docs_f/timerev.html
Cited by
6 publicationsPublic catalog entries on Sciscape.
- usesSix-month Nino 3.4 change asymmetry exceeded stationary IAAFT surrogate ranges in 1950-2025
Prior selected six-month statistic and shared source snapshot. New code tests a stationary surrogate comparison; it is not independent confirmation.
- Seasonal AR(2) and AR(6) models rarely reproduced six-month Nino 3.4 change asymmetry in 1950-2025
Original campaign selection of six-month change asymmetry, known through the follow-up publication.
- Six-month Nino 3.4 cooling imbalance exceeded six fitted seasonal linear model ranges in 1950-2025
Original pattern selection, as documented in the follow-up publication read here.
- reviewsWarm extreme months showed no clear six-month directional SST difference in detrended ERSSTv5 from 60 S to 60 N for 1982-2025
Prior index-based time-direction test. We test tail-conditioned grid-scale endpoint differences.
- reviewsA 95th-percentile size cap left six-month Nino 3.4 cubic change imbalance uncertain in 1950-2025
Prior same-record result guided this different size-sensitivity test. Not independent confirmation; no code reused.
- Surrogates with fixed month distributions did not reproduce six-month Nino 3.4 change asymmetry in 1950-2025
Original campaign report of the selected six-month change pattern.