Various studies have identifi ed that between 2012 and 2017, Brazil's semiarid region suffered severe drought. However, few studies have analyzed whether this drought also affected the eastern coastal region of Northeast Brazil (ENEB). Therefore, the objective of this work is to identify rainfall anomalies in these regions and verify the hydrometeorological impact on reservoirs in the 2012-2017 interval. For this purpose, we used precipitation data and atmospheric variables in the period from 1981 to 2017 to investigate the rainy season and associated dynamic patterns, as well as the consequences of these mechanisms on the variation of the water parameters of important reservoirs. The results indicated that rain events in the ENEB during 2012-2017 presented similar climatological behavior, without the characteristic of a drought event as observed in the semiarid region. The meteorological analyses showed that the combination of convergence with moisture over the Atlantic Ocean possibly favored greater frequency of shallow convective rainfall in ENEB, an important factor to explain the absence of generalized negative anomalies in the region. As a consequence, the reservoirs did not suffer from water collapse, unlike in the semiarid region.

This study evaluates the projected changes in the atmospheric water budget and precipitation under the RCP8.5 scenario over two CORDEX-CORE domains: South America (SAM) and Europe (EUR). An ensemble of ive twenty-irst century projections with the Regional Climate Model version 4 (RegCM4) and their driving Global Climate Models (GCMs) are analyzed in terms of the atmospheric water budget terms (precipitation, P; evapotranspiration, ET; and moisture lux convergence, C). Special focus is on four subregions: Amazon (AMZ), La Plata basin (LPB), Mid-Europe (ME) and Eastern Europe (EA). The precipitation change signal in SAM presents a dipole pattern, i.e. drier conditions in AMZ and wetter conditions in LPB. Over the two European regions a seasonality is evident, with an increase of ~ 25% in precipitation for DJF and a decrease of ~ 35% in JJA. The atmospheric water budget drivers of precipitation change vary by region and season. For example, in DJF the main drivers are related to the large-scale moisture lux convergence, while in JJA over the AMZ atmospheric moisture lux convergence plays only a minor role and local processes dominate. For JJA in the GCMs the high values of the residual term do not allow us to assess which mechanisms drive the precipitation change signal over the AMZ and LPB, respectively. Same conclusions are found for the RegCM4 JJA simulations over the LPB and EA. This points to the importance of the spatial resolution of climate simulations and the role of parameterization schemes in climate models. Our work illustrates the usefulness of analyzing regional water budgets for a better understanding of precipitation change patterns around our globe.

High‐resolution (dx = 25 km) simulations of the regional climate model RegCM4.7 coupled with the Community Land Model version 4.5 (CLM4.5) under low (RCP2.6) and high (RCP8.5) emissions scenarios, which interact with the Fifth Assessment Report of the Intergovernmental Panel on Climate Change (AR5‐IPCC), were carried out over tropical South America (TSA). These simulations were produced through boundary conditions from simulations driven by the general circulation model HadGEM2‐ES. With the goal of verifying the added value (AV) of RegCM4.7, reproducing in an adequate and coherent way the regional aspects of the historical period (1986–2005), as well as evaluating the regional aspects simulated by the model in the scope of the change projections for the far‐future (2080–2099). For this study, the climate in the TSA was characterized based on the variables of precipitation and near‐surface air temperature. For the evaluation of the spatial–temporal representation, frequency and distribution of the regional and global simulation, the high‐resolution observational dataset of the Climate Research Unit version ts4.02 (CRU) was used. Although some differences and biases still persist, RegCM4.7 presents AV in the spatial representation of precipitation and temperature over the northeast region of Brazil and part of the Andes, especially in winter. However, it does not adequately represent precipitation over the Amazon Basin, especially in summer. Results for future projections indicate that the more refined simulation of RegCM4.7 improves the projected change patterns of the coarser resolution simulation of HadGEM2‐ES and even modifies the precipitation signal in some cases. Both models project increase temperature with greater magnitude for RCP8.5. RegCM4.7 presents a much more refined and realistic spatial distribution. HadGEM2‐ES simulates the major aspects of climate enough to consider forcing RegCM4.7 to generate simulations with better performance and more realistic projections.

Resumo Este estudo investigou com observações e modelagem usando dados diários de Radiação de Onda Longa (ROL) a fase positiva favorável à atividade convectiva associada à variabilidade atmosférica intrassazonal de baixa frequência conhecida com Oscilação de Madden-Julian (OMJ) entre a região do Indico-Pacífico (INDI-PAC) e Nordeste do Brasil (NEB) entre os meses de novembro a maio no período de 1982 a 2013. O modelo usado foi global Ocean-Land-Atmosphere Model - OLAMV3.3, e para as observações foram usados dados observados por satélites gerados pelo Climate Prediction Center/National Ocean Atmospheric Admistration. O objetivo foi verificar o tempo do deslocamento da banda zonal dessas oscilações de convecção desde seu pulso inicial, região INDI-PAC até chegar sobre o NEB. Aplicou-se o método de filtragem de dados conhecido como Filtro de Lanczos na banda de 20-70 dias com a finalidade de eliminar oscilações altas (escalas sinóticas) e baixas freqüências (anuais ou mais). Características espectrais usando análises de Ondeletas mostraram que essas oscilações entre 20-70 dias, no período de novembro a maio, têm um espectro de energia máximo nos dados observados centrado no período de 8-16 pêntadas, final de dezembro e início de abril. O OLAM mostrou um adiantamento desse pico de máxima energia na região INDI-PAC e uma extensão na data de chegada no NEB, significando um atraso, sendo essa data para final de abril e maio.