BRAGA, A. C. F. M.; http://lattes.cnpq.br/6275517069065092; BRAGA, Ana Cláudia Fernandes Medeiros.
Resumen:
Hydrological models are traditionally used for representing and simulating, at basin
scale, the processes of runoff generation and propagation over the river network. The majority
of those models only considers the land surface processes of the hydrological cycle, using
rainfall and temperature as input data. This characteristic of the hydrological models can be a
limiting factor, particularly when studying the impacts of climate variability and change on
water resources. In this work, the soil-vegetation-atmosphere transfer scheme (SVAT) of the
atmospheric model BRAMS was used to produce flow in hydrographic basins. To this
purpuse, was inserted into SVAT of the atmospheric model the calculation of runoff and was
used a scheme of flow propagation to calculate discharge in hydrographic basins. Thus,
attempts were made to maintain a consistency of processes, with the calculation of the
terrestrial and atmospheric phases of water cycle by a single model. Four experiments were
run for four events and analysed in various basins to evaluate the generated flow and
sensitivities tests were made. Results show that, in most part of the basins, there was an
overestimation of simulated flow. The parameterization reduced the simulated volume at rates
between 10 and 40%, compared to the standard experiment, but still maintaining the flow
overestimated. This overestimation was caused in part by rainfall and the soil umidity flux.
The flow was very sensitive to the deepening of the soil layers, and succeeded in bringing
together the hydrographs observed and simulated in several basins. Some basins, however,
already in the standard experiment showed promising results and other showed greater
influence of precipitation than other factors.