THE INDIAN OCEAN DIPOLE AND THE SOUTH AMERICAN

THE INDIAN OCEAN DIPOLE AND THE SOUTH AMERICAN MONSOON SYSTEM Anita Drumond and Tércio Ambrizzi University of São Paulo, 2007 anitadru@model. iag. usp. br; ambrizzi@model. iag. usp. br

Post Doc developed during 2006 “The Role of the Indian Ocean Dipole on the onset of the SAMS”

The Onset of the Brazilian Rainy Season • starts in the Austral Spring (Marengo et al. , 2001 and references therein) • the process is partially associated with the propagation of transient systems over the continent (Gan et al. , 2004) • the process can be modulated by the surface temperature over Southeastern Amazonia (Fu et al. , 1999) • The IOD plays a role over the surface air temperature in the subtropical South America during the Austral Spring. This occurs through the propagation of a stationary Rossby wave train emanating from the Indian anomalous convection towards the subtropical Pacific and Atlantic Oceans (Saji et al. , 2005)

Positive Dipole Mode Indian Ocean Dipole (IOD) (Saji et al. , 1999) An air-sea coupled phenomenon observed in the Equatorial Indian ocean. It is characterized by an anomalous zonal dipolar structure of several parameters as a SST, OLR and sea surface high pressure reaching their peak during the end of the Austral Winter and Spring. Negative Dipole Mode www. jamstec. go. jp/frcgc/research/d 1/iod

It is known that. . . • the subtropical air surface temperature can be modulated by the IOD during the Austral Spring Question Can the IOD extremes modulate the onset of the Brazilian rainy season?

Data and Methodology • Chen et al. (2002) precipitation (2. 5º) • NCEP/NCAR Reanalysis (2. 5º) • Smith e Reynolds (2004) SST (2º) • Climatology: 1950 -2000 • Analysis: 1950 a 2000 • The linear tendency, the inter decadal oscillation and the ENSO signal were removed from the indexes in order to isolate de IOD signal.

IOD extreme events: Saji and Yamagata (2003) multivariate analysis • Oceanic Index – Dipole Mode Index (DMI) – difference between the normalized SST anomalies of the Western Indian Ocean (WI) (60º-80ºE, 10ºS-10ºN) and the Eastern Indian Ocean (EI) (90º-110ºE, 10ºS-0º). • Atmospheric Index – normalized surface zonal wind anomalies over the Equatorial Indian Ocean (Ueq) (70º-90ºE, 5ºS-5ºN). • IOD extreme event - DMI and Ueq exceed 0. 5 s in amplitude for at least 3 months. Moreover, WI and EI must also have opposite signals and exceed 0, 5 s in amplitude for at least 3 months.

(σ) SON DMI (Ocean Index) Year

SON DMI absolute value (s) IOD EXTREME EVENTS El Niño La Niña Neutral + IOD 9 events - IOD 11 events 1961, 1963, 1967, 1972, 1977, 1982, 1983, 1994, 1997 1958, 1960, 1964, 1971, 1974, 1975, 1989, 1992, 1993, 1996, 1998

Anomalous patterns observed during opposite phases of IOD

Correlation between SON DMI and SST T test – 95% SON anomalous SST composite ( • C) + IOD - IOD T test – 90%

Anomalous Precipitation Composites for SON (mm/day) DOI + DOI - Teste T c/ 90%

Monthly anomalous precipitation composite (mm/day) + IOD - IOD T test - 90%

Correlation between DMI and precipitation anomalies SON AUG OCT NOV T test – 95%

LAG Correlation between DMI and OCTOBER precipitation anomalies JUL AUG SEP LAG Correlation between DMI and NOVEMBER precipitation anomalies JUL AUG SEP OCT T test – 95%

SON composite of anomalous vertically integrated moisture flux (x 102 g. cm-1. s-1) and its divergence (mm/day) + IOD - IOD T test - 90%

SON 200 h. Pa anomalous zonally asymmetric component (x 106 m 2/s) + IOD - IOD SH: AC C; NH: C T test - 90% AC

The combined role of ENSO and IOD during the Austral Spring

9 positive IOD events selected for the period 19502000: • 6 occurred in El Niño events (63, 72, 77, 82, 94, 97) • 2 in neutral years (61, 67) • none of them in La Niña episodes 11 negative IOD episodes selected: • 3 were observed in El Niño events (58, 92, 93) • 2 in neutral years (60, 96) • 4 during La Niña episodes (64, 71, 74, 75)

Neutral ENSO / Positive IOD composites SON Precipitation anomalies (mm/day) SON anomalous vertically integrated moisture flux (x 102 g. cm-1. s-1) and its divergence (mm/day) SON 200 h. Pa anomalous zonally asymmetric component of stream function (x 106 m 2/s) anomalies

Neutral ENSO / Negative IOD composites SON Precipitation anomalies (mm/day) SON anomalous vertically integrated moisture flux (x 102 g. cm-1. s-1) and its divergence (mm/day) SON 200 h. Pa anomalous zonally asymmetric component of stream function (x 106 m 2/s) anomalies

El Niño / Positive IOD composites SON Precipitation anomalies (mm/day) SON anomalous vertically integrated moisture flux (x 102 g. cm-1. s-1) and its divergence (mm/day) SON 200 h. Pa anomalous zonally asymmetric component of stream function (x 106 m 2/s) anomalies

El Niño / Negative IOD composites SON Precipitation anomalies (mm/day) SON anomalous vertically integrated moisture flux (x 102 g. cm-1. s-1) and its divergence (mm/day) SON 200 h. Pa anomalous zonally asymmetric component of stream function (x 106 m 2/s) anomalies

La Niña / Negative IOD composites SON Precipitation anomalies (mm/day) SON anomalous vertically integrated moisture flux (x 102 g. cm-1. s-1) and its divergence (mm/day) SON 200 h. Pa anomalous zonally asymmetric component of stream function (x 106 m 2/s) anomalies

Discussion • + IOD: dipole presents enhanced precipitation in the subtropics and drought in the Eastern tropical South America during Austral Spring; anomalous anticyclonic moisture flux over Southeastern Brazil transporting moisture from Amazon towards the La Plata Basin; upper level anomalous wave train propagation from the Indian ocean towards South America. • IOD extreme events composites also suggest the inversion of the main anomalous patterns associated with both IOD phases. However, the anomalies are more intense in the + IOD composites.

Discussion • Anomalous patterns are more evident in October and November. Lag correlation suggests that the SST observed in the Indian Ocean since July presents some relationship with the precipitation observed in OCT and NOV. • The role of ENSO on the South American precipitation prevails during the occurrence of the negative IOD phase. However, the impacts of the positive IOD events are quite independent of the ENSO phase.

Climate Studies Group