Investigating teleconnection patterns associated with the rainy season of the northern northeast Brazil using a hidden Markov model

Abstract

The interannual variability of the northern northeast Brazil (NNB) rainfall is directly affected by the dynamics of large-scale systems through atmospheric circulation. In this context, a hidden Markov model (HMM) was used to describe the daily rainfall occurrence and intensity at four meteorological stations of the Metropolitan Region of Fortaleza (MRF) in the NNB, and the global teleconnection patterns that influence precipitation regimes during the rainy season (February, March, April and May) from 1975 to 2013. An HMM with 4 states was set up and interpreted as: very rainy (1), rainy (2), less rainy (3), and dry (4) weather conditions. They agree satisfactorily with the interannual variability of the rainy season in this region. State 2 is the only one which showed a statistical trend, indicating a probable decrease of precipitation occurrence in the rainy season. Results also show that the meteorological weather associated with both states 2 and 4 is strongly related to the El Nino-Southern Oscillation (ENSO), the North Atlantic tripole and are intrinsically conditioned to the large-scale atmospheric teleconnections of the Northern Hemisphere. All those mechanisms modulate the shift of the Intertropical Convergence Zone (ITCZ) southward (state 2) or northward (state 4), thus affecting the precipitation occurrence over the NNB. State 1 is distinguished by the influence of the Atlantic Meridional Mode (AMM) and the ITCZ displacement further south, while state 3 is identified by the absence of teleconnection patterns.