Analysis of Rainfall Variability Over Subtropical South America Considering a Complex Network Perspective
Abstract
Subtropical South America presents large rainfall variability from interannual to interdecadal time scales that has significant agricultural, environmental, energy and economic impacts. Several studies have shown that the tropical Pacific, Atlantic and Indian oceans influence rainfall variability in the subtropics through different mechanisms. Moreover, these oceans interact among them forcing sea surface temperatures in remote basins through atmospheric and oceanic teleconnections. However, it is not clear to date how the SST anomalies in different basins interact to induce rainfall anomalies. In this study we consider the interaction among oceans and their influence on rainfall over subtropical South America from a complex network perspective. The network is defined with indices that characterize the regions of the tropical oceans that influence rainfall, that is, Nino3.4 (El Niño/Southern Oscillation (ENSO), the Tropical North Atlantic (TNA, (60W-30W;10N-30N)), the Indian Ocean Dipole (IOD, (50E-70E;10S-10N)-(90E-110E;10S-0N)) and the Equatorial Atlantic (ATL3, (3N- 3S;170W-120W)) as well as an index that characterizes precipitation over different regions of the subtropics. We investigate the collective behavior of the network focusing on the detection of synchronization events in order to know more about the dynamics that characterize their interaction. We use two different measures of synchronization, that is the mean distance and the root mean square of the cross-correlation. For comparison we consider the Hadley Center Sea Surface Temperature (HadSST), the Extended Reconstructed Sea Surface Temperature (ERSSTv3b) and the Precipitation Climatology datasets from the Global Precipitation Climatological Center (GPCCv5) in the period 1901-2006. The synchronization is characterized by interdecadal variability in most of the seasons considered during 20th century. There are differences between the results of the two datasets (HadSST and ERSSTv3b) particularly in the first half of the 20th century, leading to different periods of synchronization. As an example, considering the Niño 3.4 index in austral spring and the Indian Ocean Dipole, Tropical North Atlantic and a rainfall index for northern Argentina for the months from November to January (DJF), the HadSST dataset shows two statistically significant synchronization events. The first of them occurs during the 1920s and it is not observed in ERSSTv3b dataset, while the second one starts in mid-90s and lasts for a decade being also present in the ERSSTv3b dataset.
- Publication:
-
AGU Spring Meeting Abstracts
- Pub Date:
- May 2013
- Bibcode:
- 2013AGUSMGC23A..03M
- Keywords:
-
- 1616 GLOBAL CHANGE / Climate variability;
- 3309 ATMOSPHERIC PROCESSES / Climatology;
- 1620 GLOBAL CHANGE / Climate dynamics