dorsal/arxiv
View SchemaThe El Nino Stochastic Oscillator
| Authors | Gerrit Burgers |
|---|---|
| Categories | |
| ArXiv ID | physics/9706007 |
| URL | https://arxiv.org/abs/physics/9706007 |
Abstract
Anomalies during an El Nino are dominated by a single, irregularly oscillating, mode. Equatorial dynamics has been linked to delayed-oscillator models of this mode. Usually, the El Nino mode is regarded as an unstable mode of the coupled atmosphere system and the irregularity is attributed to noise and possibly chaos. Here a variation on the delayed oscillator is explored. In this stochastic-oscillator view, El Nino is a stable mode excited by noise. It is shown that the autocorrelation function of the observed NINO3.4 index is that of a stochastic oscillator, within the measurement uncertainty. Decadal variations as would occur in a stochastic oscillator are shown to be comparable to those observed, only the increase in the long-term mean around 1980 is rather large. The observed dependence of the seasonal cycle on the variance and the correlation is so large that it can not be attributed to the natural variability of a stationary stochastic oscillator. So the El Ni\~{n}o stochastic-oscillator parameters must depend on the season. A forecast model based on the stochastic oscillator with a variance that depends on the season has a skill that approaches that of more comprehensive statistical models: over the period 1982-1993, the anomaly correlation is 0.65 for two-season lead forecasts.
{
"annotation_id": "bbb95b64-3fc5-4090-b24b-129ea9085cbc",
"date_created": "2026-03-02T18:01:20.662000Z",
"date_modified": "2026-03-02T18:01:20.662000Z",
"file_hash": "9a397d56e6c26f8fc7b3d529850fc08395a9a0456380380162117526982ef1b1",
"private": false,
"record": {
"abstract": "Anomalies during an El Nino are dominated by a single, irregularly\noscillating, mode. Equatorial dynamics has been linked to delayed-oscillator\nmodels of this mode. Usually, the El Nino mode is regarded as an unstable mode\nof the coupled atmosphere system and the irregularity is attributed to noise\nand possibly chaos. Here a variation on the delayed oscillator is explored. In\nthis stochastic-oscillator view, El Nino is a stable mode excited by noise. It\nis shown that the autocorrelation function of the observed NINO3.4 index is\nthat of a stochastic oscillator, within the measurement uncertainty. Decadal\nvariations as would occur in a stochastic oscillator are shown to be comparable\nto those observed, only the increase in the long-term mean around 1980 is\nrather large. The observed dependence of the seasonal cycle on the variance and\nthe correlation is so large that it can not be attributed to the natural\nvariability of a stationary stochastic oscillator. So the El Ni\\~{n}o\nstochastic-oscillator parameters must depend on the season. A forecast model\nbased on the stochastic oscillator with a variance that depends on the season\nhas a skill that approaches that of more comprehensive statistical models: over\nthe period 1982-1993, the anomaly correlation is 0.65 for two-season lead\nforecasts.",
"arxiv_id": "physics/9706007",
"authors": [
"Gerrit Burgers"
],
"categories": [
"physics.ao-ph"
],
"title": "The El Nino Stochastic Oscillator",
"url": "https://arxiv.org/abs/physics/9706007"
},
"schema_id": "dorsal/arxiv",
"source": {
"execution_id": "0a4f784c-80dd-4f33-bf1e-88d5c2ea6d69",
"id": "arXiv Dataset IDs",
"type": "Model",
"variant": "snapshot-2026-03-01",
"version": "0.1.0"
},
"user_id": 1000002
}