The South Atlantic Convergence Zone (SACZ) plays a key role in the South American monsoon system (SAMS) precipitation regime, and accounts for around 25% of the rainfall volume over Southeast Brazil between October and April, on average, with peaks of 56% in March and 41% in January. Due to its often varying position and multi-variable structure, diagnosing or quantifying SACZ episodes may lie on subjective criteria. The present study provides a climatological characterization of different SACZ types, based on their position, and differences in dynamics analysed from 19 SAMS wet periods. A cycle in SACZ configurations is identified during the rainy season: northernmost (southernmost) episodes are more likely to occur during the onset/demise (peak) months of the SAMS. Objective SACZ indices are developed taking the principal components of the dynamics of the SACZ as explanatory variables, not including precipitation. The most sensitive threshold proposed for the identification of mean SACZ type yields simultaneous true positive rate of 86%, false alarm rate of 28%, true negative rate of 72% and false negative rate of 15%, for example. The indices have potential to bring gains in predictability to SACZ-related precipitation in Southeast Brazil, apart from allowing the objective diagnosis of SACZ episodes. The indices are also correlated with the South American precipitation dipole and may be used to identify anomalous SAMS precipitation at longer time scales as well.

The Amazon rain forest covers an area of ~ 6.7 million km² of South America; nearly 60% of it is in Brazil, while the rest is shared among eight other countries. This vast extent of rain forest is a globally significant ecosystem that provides numerous ecosystem services that benefit humanity including essential climate regulation, biodiversity conservation, and hydrological stability. However, deforestation and forest degradation have led to the loss of approximately 15% of the Amazon rainforest since the 1970s, primarily driven by agricultural expansion, illegal mining, logging, and wildfires. These pressures have triggered a cascade of consequences, including biodiversity loss, disruption of cultural and ecosystem services, depletion of carbon sinks, and severe alterations to the hydrological cycle. While initially manifesting at local and regional scales, these effects increasingly pose risks to global climate stability. We simulated deforestation scenarios (15%, 50%, and 100% forest loss) using the Community Atmosphere Model (CAM 3.1) to evaluate precipitation changes and atmospheric responses. Results indicate substantial reductions in regional precipitation, hydrological disruptions affecting agricultural productivity, and an increasing risk of the Amazon transitioning from a carbon sink to a carbon source. This underscores the urgency of policy interventions, including stricter environmental regulations, trade restrictions on commodities produced illegally or in deforested areas, enhanced Indigenous land protection, and international cooperation to mitigate deforestation and promote sustainable land use. Immediate action is necessary to prevent irreversible ecological and climatic tipping points.

Flash floods and extreme rains are destructive phenomena and difficult to forecast. In 2011, the mountainous region of Rio de Janeiro state suffered one of the largest natural hazards in Brazil, affecting more than 300,000 people, leaving more than 900 dead. This article simulates this natural hazard through Quantitative Precipitation Forecasting (QPF) and streamflow forecast ensemble, using 18 combinations of parameterizations between cumulus, microphysics, surface layer, planetary boundary layer, land surface and lateral contour conditions of the Weather Research and Forecasting (WRF) Model, coupling to the Soil Moisture Accounting Procedure (SMAP) hydrological model, seeking to find the best set of parametrizations for the forecasting of extreme events in the region. The results showed rainfall and streamflow forecast were underestimated by the models, reaching an error of 57.4% to QPF and 24.6% error to streamflow, and part of these errors are related to the lack of skill of the atmospheric model in predicting the intensity and the spatial-temporal distribution of rainfall. These results bring to light the limitations of numerical weather prediction, possibly due to the lack of initiatives involving the adaptation of empirical constants, intrinsic in the parametrization models, to the specific atmospheric conditions of each region of the country.