Upper-Level Mediterranean Oscillation index and seasonal variability of rainfall and temperature

The need for early seasonal forecasts stimulates continuous research in climate teleconnections. The large variability of the Mediterranean climate presents a greater difficulty in predicting climate anomalies. This article reviews teleconnection indices commonly used for the Mediterranean basin and...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Theoretical and applied climatology 2019-02, Vol.135 (3-4), p.1059-1077
Hauptverfasser: Redolat, Dario, Monjo, Robert, Lopez-Bustins, Joan A., Martin-Vide, Javier
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:The need for early seasonal forecasts stimulates continuous research in climate teleconnections. The large variability of the Mediterranean climate presents a greater difficulty in predicting climate anomalies. This article reviews teleconnection indices commonly used for the Mediterranean basin and explores possible extensions of one of them, the Mediterranean Oscillation index (MOi). In particular, the anomalies of the geopotential height field at 500 hPa are analyzed using segmentation of the Mediterranean basin in seven spatial windows: three at eastern and four at western. That is, different versions of an Upper-Level Mediterranean Oscillation index (ULMOi) were calculated, and monthly and annual variability of precipitation and temperature were analyzed for 53 observatories from 1951 to 2015. Best versions were selected according to the Pearson correlation, its related p value, and two measures of standardized error. The combination of the Balearic Sea and Libya/Egypt windows was the best for precipitation and temperature, respectively. The ULMOi showed the highest predictive ability in combination with the Atlantic Multidecadal Oscillation index (AMOi) for the annual temperature throughout the Mediterranean basin. The best model built from the indices presented a final mean error between 15 and 25% in annual precipitation for most of the studied area.
ISSN:0177-798X
1434-4483
DOI:10.1007/s00704-018-2424-6