This is an observational study of the large-scale moisture transport over South America, with some analyses on its relation to subtropical rainfall. The concept of aerial rivers is proposed as a framework: it is an analogy between the main pathways of moisture flow in the atmosphere and surface rivers. Opposite to surface rivers, aerial rivers gain (lose) water through evaporation (precipitation). The magnitude of the vertically integrated moisture transport is discharge, and precipitable water is like the mass of the liquid column-multiplied by an equivalent speed it gives discharge. Trade wind flow into Amazonia, and the north/northwesterly flow to the subtropics, east of the Andes, are aerial rivers. Aerial lakes are the sections of a moisture pathway where the flow slows down and broadens, because of diffluence, and becomes deeper, with higher precipitable water. This is the case over Amazonia, downstream of the trade wind confluence. In the dry season, moisture from the aerial lake is transported northeastward, but weaker flow over southern Amazonia heads southward toward the subtropics. Southern Amazonia appears as a source of moisture to this flow. Aerial river discharge to the subtropics is comparable to that of the Amazon River. The variations of the amount of moisture coming from Amazonia have an important effect over the variability of discharge. Correlations between the flow from Amazonia and subtropical rainfall are not strong. However, some months within the set of dry seasons observed showed a strong increase (decrease) occurring together with an important increase (decrease) in subtropical rainfall.

December–March climatologies of precipitation and vertically integrated water vapor transport were analyzed and compared to find the main paths by which moisture is fed to high-rainfall regions in the Southern Hemisphere in this season. The southern tropics (20°S–0°) exhibit high rainfall and receive ample moisture from the northern trades, except in the eastern Pacific and the Atlantic Oceans. This interhemispheric flow is particularly important for Amazonian rainfall, establishing the North Atlantic as the main source of moisture for the forest during its main rainy season. In the subtropics the rainfall distribution is very heterogeneous. The meridional average of precipitation between 35° and 25°S is well modulated by the meridional water vapor transport through the 25°S latitude circle, being greater where this transport is from the north and smaller where it is from the south. In South America, to the east of the Andes, the moisture that fuels precipitation between 20° and 30°S comes from both the tropical South and North Atlantic Oceans whereas between 30° and 40°S it comes mostly from the North Atlantic after passing over the Amazonian rain forest. The meridional transport (across 25°S) curve exhibits a double peak over South America and the adjacent Atlantic, which is closely reproduced in the mean rainfall curve. This corresponds to two local maxima in the two-dimensional field of meridional transport: the moisture corridor from Amazonia into the continental subtropics and the moisture flow coming from the southern tropical Atlantic into the subtropical portion of the South Atlantic convergence zone. These two narrow pathways of intense moisture flow could be suitably called “aerial rivers.” Their longitudinal positions are well defined. The yearly deviations from climatology for moisture flow and rainfall correlate well (0.75) for the continental peak but not for the oceanic peak (0.23). The structure of two maxima is produced by the effect of transients in the time scale of days.

No abstract

South American subtropics east of the Andes exhibit a region of intense climatological frontogenesis in equivalent potential temperature (EPT) in the December to March season, mostly produced by deformation of the wind field. The goal of this paper is to investigate the large scale features associated with intense and weak frontogenesis by deformation (FGD) in EPT in the region where it attains its climatological maximum. This can be approximately delimited by 32–42° S and 66–69° W, which is small enough as to contain only one synoptic perturbation at a time. The spatial average of the positive values of frontogenesis at 850 hPa over the whole region (DFG+) is used to represent the strength of the perturbation. ECMWF ERA-40 reanalysis data set is used to calculate DFG+ at six hour intervals for 21 seasons (1981–2002). Compositing analysis is carried out for strong (above the 0.75 quantile) and weak (below the 0.25 quantile) events. For strong events the geopotential field at 850 hPa exhibits the North Argentinean Low (NAL), a transient trough and the Low Pressure Tongue East of the Andes (LPT). Upon comparison with the composite field of FGD it can be observed that FGD exhibits a strong maximum over the Argentinean Col (AC) which separates the NAL and the trough. These features are absent in the weak frontogenesis composite, which exhibits a stronger South Pacific Subtropical High close to the continent. At 250 hPa the strong FGD composite exhibits a trough over the Andes with a wind speed maximum to its east. Both of these features are associated with the deepening of the NAL in the literature. These are not present in the weak FGD composites. Strong events show an intense quasi meridional corridor of water vapor transport from the Amazon to the subtropics that encounters westerly flow in the neighborhood of the AC. This is absent in weak events. A preliminary analysis of precipitation is carried out using the GPCP daily data set. An intense precipitation nucleus appears slightly northeast of the AC, with maximum intensity in the day that follows the strong events. Weak events exhibit a drying of the subtropics instead, between one and three days after the events. Higher precipitation over the oceanic South Atlantic Convergence Zone can be also observed. Analogous composites were constructed for the presence and absence of both the AC and the LPT, showing similar characteristics to the strong and weak FGD event composites respectively, but with lower intensities. This shows that by selecting strong FGD events, intense NAL and LPT events are also singled out

No abstract