On the opposite relation between extreme precipitation over west Amazon and southeastern Brazil: observations and model simulations

摘要

During the austral summer of 2014 and 2015, a severe drought occurred in southeastern Brazil at the same time when flooding conditions were registered in the state of Acre in the western Amazonia of Brazil. This study aimed the identification of the atmospheric and oceanic large-scale characteristics and regional features associated with these conditions, and verification of the Center of Weather Forecasting and Climate Studies/National Institute of Space Research (CPTEC/INPE) atmospheric global circulation model ability in reproducing the observed features. The state of Acre was chosen as a pilot area of the PULSE-Brazil project - a web platform tool that contains climate, ecological and health data for Brazil. The observational data and model results were analysed for the climatological period of 1981-2010. Correlation analyses between precipitation over west Amazon and atmospheric and oceanic variables indicated the main patterns in both data sets. Extreme wet and dry Januaries over west Amazon were selected to explore the regional and large-scale associated features. The general results of climate analyses indicated an opposite relation of precipitation between western Amazon and southeastern Brazil, in January, besides the typical precipitation dipole between the south Atlantic convergence zone and southeastern South America. This configuration is related to anomalous humidity flux at low levels and associated anomalous sea level pressure over southeastern South America induced by subsidence. The role of convection anomalies over the Pacific Ocean on South America anomalies is discussed. The wet and dry cases in Acre region display opposite atmospheric anomalies over South America that are linked to wavetrains over the Pacific Ocean, likely related to the opposite conditions of Indonesia-Pacific tropical convection. Similar patterns of the dry cases were observed during January of 2014 and 2015. The model reproduced some observed atmospheric patterns related to precipitation extremes and the results are discussed in terms of regional and large-scale climate variability.