In recent years, the availability of global model analyses has increased our understanding of the South American tropospheric warm season circulations and particularly their monsoon-like features. Additionally, simulations from simplified models have permitted to comprehend the basic characteristics of the circulation and its forcings. The current knowledge of the moisture sources and transports at regional scales is much less advanced. There is some limited information that suggests the existence of a low-level jet similar to that over the Great Plains of the United States. However, in the case of South America, there is a lack of a comprehensive description of this central element of the atmospheric water cycle. The reasons are a limited observational network and the mesoscale nature of the low-level jet.
In this study we examine regional characteristics of the warm season circulation and specifically moisture transports at lower levels, using the National Centers for Environmental Prediction (NCEP) mesoscale Eta model, that was developed primarily for forecasting heavy precipitation events and other mesoscale phenomena. The model was applied to a domain that covers all South America and parts of the adjoining oceans with a horizontal resolution of 80 km and 38 vertical levels. At these resolutions, the Eta coordinate is expected to depict adequately the abrupt barrier effect of the Andes mountains. This version performs a static run using the NCEP global reanalyses as initial fields and boundary conditions. A series of short term forecasts (up to 36 hours) for November of 1997 were used to construct the basic data set used in this study.
The low-level jet produced in the model appears to be topographically bounded and diurnally modulated by the thermal forcing, and its maximum intensity is observed at about 03 UTC (near local midnight). The low-level jet has a related region of vertically integrated moisture flux convergence in northeastern Argentina/southern Brazil, where precipitation has a local maximum.
A comparison with satellite estimates shows that the model reproduces all major areas of precipitation, although it tends to produce unrealistic rain near northeastern Brazil. This is a problem that affects many other models as well, and does not appear to affect the forecast fields at other regions