Análise da representatividade de diferentes hipóteses de modelagem matemática para definição de manchas de inundações em áreas urbanas: estudo de caso da cidade de Governador Valadares-MG

Abstract

This research objectifies the analysis of computational hydraulic models for floodplain representation in urban areas,with great importance for delimitation of naturally floodablezones. It aimsto provide elements forplanning occupation and the support for alert systems, in cities that are already implanted in the floodplains of great rivers, with frequentflooding risk. It searchesthe peculiarities identification of this kind of modeling, using the critical analysis of the existing data and the obtained results. The researchalso aims at the identification of Manning roughness coefficients associates to different kinds of urban occupation obtained according to the calibration of four different scenarios of one-dimensional models. The study was developed inGovernadorValadarescity, in Minas Gerais, which owns data of bathymetry and topography, as well as of several gageddata of flooding periods registered, with calibration developed in steadyflow simulation to posterior comparisonswithunsteadyflows simulationsresults.To elaborate the four different kinds of models were used representation of the urban area in a database GIS (Geographic Information System), representing the buildings installed in the flood plain, in different levels of details.The software HEC-RAS 4.1, of HydrologicEngineeringCenter, was used to the modeling, simulations and calibrations. For the data topologic obtainment and for floodplain delineation it used the software ArcGIS, developed by ESRI, and HEC-GeoRAS, developed jointly by ESRI and HydrologicEngineeringCenter. The coefficients calibrations of Manning were accomplished with observed data ofeleven series of gaged flows and heights and data obtained ofthe developed flowcurve (CPRM, 2004), in seven gagedsections. For the other analyses werealso used fivehydrographs of recordedevents, and theirrespective peak flows. There were proceededasensibility analyses for a wide flows band, to verify the limits validity of the adopted parameters, along the flood main channel and floodplain. The comparative evaluation of floodplain delineation simulated with effective observed data allows discerning the characteristics of precision, representativity and practicability of the four modeling hypotheses, to define the models representation degree. Finally, it relates over about modelingspractical aspects, limitations, potentialities and critical analysis of the data.