Interannual variability of Tropical Atlantic and its influence on extreme precipitation events: Focus on the Amazon Basin

Abstract

Evalúa la influencia de la variabilidad atlántica en la precipitación de la Cuenca Amazónica. Este estudio proporciona un análisis detallado, en la escala de tiempo interanual, de la influencia del Pacífico tropical y el Océano Atlántico en el régimen de precipitaciones sobre América del Sur, con un enfoque en la Cuenca Amazónica. Para ello, se utilizan datos mensuales de temperatura de la superficie del mar para determinar los principales modos de variabilidad en tres regiones atlánticas: Atlántico norte tropical (TNA), Atlántico sur tropical (TSA) y Atlántico sur subtropical (STSA). Además, se analizó la interacción Pacífico y Atlántico mediante correlaciones móviles y ondeletas cruzadas para el período 1880-2017. Descubrimos que las regiones del Atlántico se han estado acoplando y desacoplando en períodos específicos. Asimismo, los océanos Pacífico y Atlántico se han mostrado en fase o en fase opuesta en diferentes períodos de tiempo y frecuencias. Se utilizaron datos de precipitaciones para identificar las regiones influenciadas por la TSM del Pacífico y el Atlántico. Además, el flujo de humedad, divergencia y estructura vertical sobre América del Sur y los océanos adyacentes muestran que las condiciones anómalas de la temperatura de la superficie del mar en estas regiones oceánicas modifican las regiones de origen y sumidero de la humedad atmosférica en todas las estaciones.

This study provides a detailed analysis, in the interannual time-scale, of the influence of the tropical Pacific and the Atlantic Ocean in the precipitation regime over South America, with a focus on the Amazon Basin. For this purpose, first, monthly sea surface temperature data is used to determine the principal modes of variability in three Atlantic regions: Tropical North Atlantic (TNA), Tropical South Atlantic (TSA), and Subtropical South Atlantic (STSA). These patterns are obtained from Empirical Orthogonal Function (EOF), where the first mode (PC1) of each region explains the interannual variability in high percentage. The PC1 time series also shows SST peaks related to extreme precipitation events in the Amazon Basin. In addition, the Pacific and Atlantic interaction was analyzed by moving correlations and cross-wavelets for the 1880-2017 period, using Pacific indices (E and C index) and the PC1 of each Atlantic region. We found that the Atlantic regions have been coupling and decoupling in specific periods. Likewise, the Pacific and the Atlantic oceans have been shown in-phase or in opposite phase in different periods of time and frequencies. Precipitation from satellite and combined observed-satellite for the 1940-2010 period were used to identify regions influenced by the Pacific and Atlantic SST. The results suggest that different El Niño Southern Oscillation (ENSO) flavors mainly impact precipitation during the onset (October-November) and peak (December-February) of the South America Monsoon System (SAMS). We have also shown that anomalous conditions over the Tropical North and South Atlantic Ocean impact precipitation regimes during the demise (March-April) and the absence of the SAMS (May-September), while the Subtropical South Atlantic impacts in precipitation during the peak and the absence of the SAMS. Additionally, moisture flux, its divergence, and vertical structure over South America and the adjoining oceans were obtained for different conditions of tropical oceans, through the oceanic indices. Results show that anomalous conditions of sea surface temperature in these oceanic regions modify the source and sink regions for atmospheric moisture in all seasons. A reduction in the moisture inflow, mainly in warm tropical Atlantic conditions, could lead to extremely dry periods in central-southern Amazon and the northern La Plata Basin. This effect, indeed, could be even more extreme in tropical South America when it is preceded by the El Niño event. However, contrasted impacts are observed during El Niño years, with rainfall deficit in tropical South America, east of the Andes, and rainfall excess over southeastern South America and northwest of the Andes (northern Peru, coastal Ecuador, and Colombia). These findings provide an analysis that will help to improve adaptation and risk management policies related to impacts of climate variability and change.