Several droughts and floods in Amazonia and Northeast Brazil have occurred in recent years and projections from Intergovernmental Panel on Climate Change indicate an increase of these extreme events. El Niño Southern Oscillation (ENSO) is one of the phenomena associated with extreme rainfall events in the Amazon. However, recent studies have indicated that the basic response of ENSO is dependent on the Madden-Julian Oscillation (MJO) phase. Hence, this study analyses the MJO influence on precipitation extreme events over northern South America in El Niño and La Niña years. Extreme precipitation events over northern South America for the rainy season (December-May) were obtained through a composite analysis of the combinations of ENSO and MJO phases. Most of the dry extreme events occurred during El Niño periods, while wet extreme events were more recurrent during La Niña or neutral years. However, the results showed that the MJO convection could enhance or weaken the basic response of ENSO on extreme precipitation events. Moreover, dry/wet extreme events over both Amazon and Northeast Brazil are favoured when MJO convection over Indonesia is enhanced (MJO phases 4 and 5)/suppressed (MJO phase 2). Additionally, the interannual variability of the extreme events showed an increasing linear trend for dry extreme events and a decreasing linear trend for wet extreme events. The results presented here contribute to a better understanding of the climate variability and will be helpful for the forecast of ENSO effects on extreme events over northern South America. KEY WORDS extreme events; ENSO and MJO interaction; northern South America

No abstract

The present precipitation and temperature patterns and expected future changes (2073–2098) in Africa are investigated using the Hadley Centre Global Environmental Model 2‐Earth System (HadGEM2‐ES) under the fifth phase of the Coupled Model Intercomparison Project (CMIP5) protocols for historical and future emission scenarios simulations. In a CMIP5 multimodel analysis, the annual cycles of temperature and precipitation simulated by HadGEM2‐ES were very close to the multimodel ensemble mean. HadGEM2‐ES temperature simulation compares well with the National Center for Atmospheric Research (NCAR) reanalysis over the 1979–2004 periods, except for a summer overestimation in Central Africa, and a winter underestimation in tropical West Africa. The precipitation simulation compared well with the Global Precipitation Climatology Project (GPCP) data from 1979 to 2004 over the entire Africa, except in the Intertropical Convergence Zone (ITCZ), where the model fails to capture adequately the transition phase of the monsoon circulation. The dry regimes over Northern Africa as well as the wetter regime occurring over Central Africa, which is mainly regulated by the ITCZ displacement, and during the austral summer of Southern Africa, are also fairly reproduced by the HadGEM2‐ES model. The model projects for the end of the 21st century a rainy South Africa, a change of the flood/drought cycle in the Tropics and a warming over the whole continent, varying from 3 to 7 °C. HadGEM2‐ES performance for Nigeria shows good reproduction of precipitation seasonal cycles for some locations, outside the ITCZ. However, the comparison with in situ measurement in Ilorin and Lagos shows the model is not being able to reproduce the precipitation annual cycle. Future projections for Nigeria exhibit warming everywhere and an enhancement of precipitation, especially in the northern part of the country.

No abstract

Historical simulations (present climate) and projections under RCP8.5 scenario (future climate) by HadGEM2-ES of temperature and precipitation are analyzed during the four seasons in South America. Projections of precipitation are discussed in terms of atmospheric circulation. The South Atlantic Convergence Zone (SACZ) and the Pacific South America (PSA) patterns are analyzed in simulations of present climate and in future climate projections. The model shows small systematic errors over South America, larger close to the northern South American coast in DJF and MAM. The seasonal variability of precipitation, temperature and wind fields is very well reproduced, mainly the summer/winter differences. The SACZ and the Intertropical Convergence Zone (ITCZ) are well simulated. The good model performance to reproduce the precipitation, temperature and wind fields, in the present climate, gives confidence in the projection results subject to the future scenarios. Changes from the present time to the future indicate increased precipitation over southern and southeastern Brazil and areas nearby and the tropical western South American coast. Reduced precipitation is projected over eastern Amazonia, northern South America and southern Chile. The changes are related to changes in the low level wind flow over the tropical North Atlantic, which reduces the advection of moisture to the continent and also to the increased low level flow over central South America southwards, which increases the humidity in the southern regions. The upper level flow changes are also consistent with the precipitation changes. There is a weakening of the Bolivian High and a strengthening of the subtropical jet over the continent. The SACZ dipole pattern is well simulated and in the future projections the southern center anomalies are more intense than in the present time. The PSA1 and PSA2 patterns are well represented in the present climate, but in the future projection only one dominant mode is identified as the typical teleconnection over the Pacific and South America.

We provide a comprehensive analysis of the Holocene climate and vegetation changes over South America through numerical simulations. Holocene climate for several periods (8 ka, 6 ka, 4 ka, 2 ka, and present) were simulated by an atmospheric general circulation model, forced with orbital parameters, CO2 concentrations, and sea surface temperature (SST), while the analysis of the biome distributions was made with a potential vegetation model (PVM). Compared with the present climate, our four simulated periods of the Holocene were characterized by reduced South Atlantic Convergence Zone intensity and weaker South American Monsoon System (SAMS). The model simulated conditions drier than present over most of South America and gradual strengthening of SAMS toward the present. The Northeast Brazil was wetter because of southward migration of the intertropical convergence zone (ITCZ). Moreover, SST conditions were the main forcing for the climate changes during the mid Holocene inducing larger austral summer southward ITCZ migration. PVM paleovegetation projections are shown to be consistent with paleodata proxies which suggest fluctuations between biomes, despite the fact that ages of dry/wet indicators are not synchronous over large areas of the Amazonian ecosystem. Holocene PVM simulations show distinct retreat in Amazonian forest biome in all four simulated periods. In 6 ka, present caatinga vegetation in Northeastern Brazil was replaced with savanna or dense shrubland. The simulations also suggest the existence of rainforest in western Amazonia and the expansion of savanna and seasonal forest in the eastern Amazon, with shifts in plant community compositions and fragmentation located mostly in ecotone regions. Moreover, our PVM results show that during the Holocene, the Amazonian tropical forest was smaller in area than today, although western Amazonia persisted as a tropical forest throughout the Holocene.

Abstract. The climate in the Amazon region is particularly sensitive to surface processes and properties such as heat fluxes and vegetation coverage. Rainfall is a key expression of the land surface–atmosphere interactions in the region due to its strong dependence on forest transpiration. While a large number of past studies have shown the impacts of large-scale deforestation on annual rainfall, studies on the isolated effects of elevated atmospheric CO2 concentrations (eCO2) on canopy transpiration and rainfall are scarcer. Here, for the first time, we systematically compare the plant physiological effects of eCO2 and deforestation on Amazon rainfall. We use the CPTEC Brazilian Atmospheric Model (BAM) with dynamic vegetation under a 1.5×CO2 experiment and a 100 % substitution of the forest by pasture grasslands, with all other conditions held similar between the two scenarios. We find that both scenarios result in equivalent average annual rainfall reductions (Physiology: −257 mm, −12 %; Deforestation: −183 mm, −9 %) that are above the observed Amazon rainfall interannual variability of 5 %. The rainfall decreases predicted in the two scenarios are linked to a reduction of approximately 20 % in canopy transpiration but for different reasons: the eCO2-driven reduction of stomatal conductance drives the change in the Physiology experiment, and the smaller leaf area index of pasturelands (−72 % compared to tropical forest) causes the result in the Deforestation experiment. The Walker circulation is modified in the two scenarios: in Physiology due to a humidity-enriched free troposphere with decreased deep convection due to the heightening of a drier and warmer (+2.1 ∘C) boundary layer, and in Deforestation due to enhanced convection over the Andes and a subsidence branch over the eastern Amazon without considerable changes in temperature (−0.2 ∘C in 2 m air temperature and +0.4 ∘C in surface temperature). But again, these changes occur through different mechanisms: strengthened west winds from the Pacific and reduced easterlies entering the basin affect the Physiology experiment, and strongly increased easterlies influence the result of the Deforestation experiment. Although our results for the Deforestation scenario agree with the results of previous observational and modelling studies, the lack of direct field-based ecosystem-level experimental evidence regarding the effect of eCO2 on moisture fluxes in tropical forests confers a considerable level of uncertainty to any projections of the physiological effect of eCO2 on Amazon rainfall. Furthermore, our results highlight the responsibilities of both Amazonian and non-Amazonian countries to mitigate potential future climatic change and its impacts in the region, driven either by local deforestation or global CO2 emissions.

The reliability of predictions from climate and weather models is linked to an adequate representation of the land surface processes. To evaluate performance and to improve predictions, land surface models are calibrated against observed data. Despite an extensive literature describing methods of land surface model calibration, few studies have applied a calibration method for semiarid natural vegetation, especially for the semiarid northeast of Brazil, which presents caatinga as its natural vegetation. Caatinga is a highly dynamic ecosystem with the physics at the land surface-atmosphere interface still poorly understood. Therefore, in this study a multiobjective hierarchical method, which provides means to estimate optimal values of the model parameters through calibration, is evaluated. This method is applied to caatinga by using the Integrated Biosphere Simulator (IBIS). Results demonstrated that the calibrated set of vegetation parameters produced a considerably different energy balance from the default parameters. In general, the model was able to simulate the partition of the available energy into sensible and latent heat fluxes when the calibrated parameters were used. The IBIS model was not able to capture short-term, intense changes in latent heat flux from a dry condition to a wetter condition, however, even when the new set of calibrated parameters was used. Therefore, the parameter optimization may not be sufficient if processes are missing or misrepresented. This study is one of the first to understand the physics at the land surface-atmosphere interface in the caatinga ecosystem and to evaluate the ability of the IBIS model to represent the biophysical interactions in this important ecosystem.