Processamento digital de imagens para determinação do coeficiente de Manning na região da Barragem de São José do Jacuípe - BA.

Simulations of hydrological models constitute a powerful ferment for evaluation and environmental management, aiding in decision making. The scale of complexity of these models requires special attention, especially when working with surface runoff of open channels, since in this context a variable appears, which is not well worked will imply in incoherent results, this is the coefficient of Manning. The Manning coefficient corresponds to the opposite force to the flow, the determination of this coefficient is preponderant to indicate the resistance of the flow in the channel, so it is not a simplistic task since there are no exact methods for its determination. The present paper had the objective of applying digital image processing PDI - color composition and supervised classification, using the Maximum Likelihood algorithm - Maxvers Landsat Image 08, TM + sensor of February of 2015 with spatial resolution of 30 meters 16bits spectral resolution, cartographic projection WGS84, 24Sul spindle with the intention of mapping the Manning classes to the municipality of São José do Jacuípe, especially in the area between the São José do Jacuípe Dam and the Urban area of this city. The choices of the classes were according to (DE JONG et al., 2003; CAMPOS, 2011). Based on the evaluation of the accuracy of the results, field work and visual identification, this technique proved adequate for this mapping, and it is advisable to use it in future work.