Premium
Investigating the origins and significance of low‐frequency modes of climate variability
Author(s) -
Allen Myles R.,
Smith Leonard A.
Publication year - 1994
Publication title -
geophysical research letters
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.007
H-Index - 273
eISSN - 1944-8007
pISSN - 0094-8276
DOI - 10.1029/94gl00978
Subject(s) - climatology , predictability , sea surface temperature , global temperature , environmental science , pacific decadal oscillation , series (stratigraphy) , noise (video) , la niña , global change , global warming , climate change , geology , oceanography , el niño southern oscillation , physics , paleontology , quantum mechanics , artificial intelligence , computer science , image (mathematics)
An analysis of the 130‐year record of the Earth's global mean temperature reveals a significant warming trend and a residual consistent with an auto‐correlated (“red”) noise process whose predictability decays with a timescale of two years. Thus global temperatures, in isolation, do not indicate oscillations at 95% confidence against a red noise null hypothesis. Weak signals identified in the global series can, however, be traced to significant sea surface temperature oscillations in the equatorial Atlantic (period ∼10 years) and the El Niño region of the Pacific (3–5 years). No robust evidence is found in this data for interdecadal oscillations, The 10‐year Atlantic oscillation corresponds to a pattern of temperature anomalies which has been associated with interannual variations in West African rainfall and in U.S. hurricane landfall frequency.