Simulação hidrológica na bacia hidrográfica do rio Pomba usando o modelo SWAT

• Autor(a) principal: Pereira, Donizete dos Reis
• Data de Publicação: 2013
• Tipo de documento: Tese
• Idioma: por
• Título da fonte: LOCUS Repositório Institucional da UFV
• Texto Completo: http://locus.ufv.br/handle/123456789/744
• topic: Modelagem hidrológica; Calibração; Validação; Hydrologic modeling; Calibration; Validation; CNPQ::CIENCIAS AGRARIAS::ENGENHARIA AGRICOLA

The hydrologic simulation is an important tool for planning and management of water resources in watersheds and can be used in the estimation of water availability, streamflow forecasting and analysis of hydrologic response of the basin to changes in the land use. Thus, this work aimed to calibrate and validate the SWAT model for the watershed of Pomba river with control section in the Astolfo Dutra gaging station; validate it upstream and downstream of the calibration sections; validate it in basins different from that used in the calibration and apply it to simulate scenarios of changes in the land use and occupation. The data used for the application of the SWAT were Digital Elevation Hydrografically Conditioning model (DEHCM), maps of soil and land, rainfall (14 monitoring stations) and flow rates (10 monitoring stations ) provided by the National Water Agency (ANA), climate data from three weather stations provided by the National Institute of Meteorology (INMET) and soils information obtained from RADAMBRASIL project. The calibration of the SWAT was performed by trial and error using the period going from 01/01/1995 to 12/31/1999 and the validation was done in the period from 01/01/2000 to 12/31/2004. The statistics used to evaluate the quality of the calibration and the validation were the coefficient of Nash and Sutcliffe (E NS) and its logarithmic version ((ENS) log) and the percentage of bias of the simulated flow rates in relation to the observed ones (PBIAS). In addition to the validation of the hydrological model for simulating continuous series of streamflow rates, was analyzed the performance of the model in simulating maximum and minimum streamflow rates associated with different return periods and minimum reference streamflow rates to grant the water use, which were made by means of simple linear regression analysis. Finally, the hydrologic responses of the watershed to three different scenarios of land uses were simulated. In scenario 1 was considered that the Permanent Preservation Areas (PPAs) of the watershed were intact; in scenarios 2 and 3 it was considered that 18.75% of the area covered with pastures was replaced by agricultural crops and eucalyptus, respectively, which was the equivalent to the replacement of the 10% of the total area of the basin. The values of ENS, (ENS) log and PBIAS obtained were 0.73, 0.58, -1.8% and 0.76, 0.78 and 5.1% for the calibration and validation periods, respectively, describing the model as appropriate and good for the hydrologic simulation of the watershed of the Pomba river. The validations for upstream and downstream sections of the calibration section resulted in satisfactory performances for basins with drainage areas greater than 700 km2 and with best rainfall representations (more monitoring rainfall station) ; for drainage areas less than 500 km2 and with low rainfall representation (few monitoring rainfall station). The validations in the sub-basins of the Novo and Xopotó rivers (Proxi-basin test) yielded satisfactory performances in sections of the Novo river that drains areas greater than 800 km2 ; in the other hand, the performance was unsatisfactory in the sub-basin of the Xopotó river, probabily due to the low rainfall representation. The simulated maximum and minimum flow rates associated to different return periods and the minimum reference low rates to grant the water use values did not differed statistically from the observed values by "t" test at 5% probability. The impacts of changes in land use on the hydrologic parameters of the Pomba river watershed yielded an average annual reduction of 13.6 mm in the flow depth and 32.6 m3 s- 1 and 2.0 m3 s-1 for the maximum and minimum streamflow rates, respectively, for scenario 1.In a similar way, for scenario 2, occurred an average annual reduction of 4.0 mm in flow depth and of 5.6 m3 s-1 and 0.6 m3 s-1 for the maximum and minimum streamflow rates, respectively. For scenario 3, it was obtained also an average annual reduction in hte flow depth and in the minimum streamflow rate of 6.5 mm and 1.5 m3 s-1, respectively, but an increase in the maximum streamflow rate of 24.0 m3 s-1. It is concluded that SWAT can be used to simulate hidrologic parameters of the Pomba river watershed in basins with drainage areas greater than 700 km2 and with a good spatial rainfall representation. The model did not performed well for basins with drainage areas less than 500 km2 and with low spatial rainfall representation.