Modelagem hidrológica de pequenas bacias hidrográficas a partir de dados de elevação obtidos por lidar

Abstract

The hydrographic basins reflect the occupation of their areas in both the environmental quality dynamics and their drainage network, being the characterization and understanding of their dynamics is a key part for planning and management of drainage systems in an efficient way. For more than two decades the hydrological simulation has being benefited from the enhancement of Geographic Information Systems (GIS), including the increase in computational capacity and greater availability of terrain information, including LiDAR technology. In order to simulate the behavior of watersheds in the face of rainy events, rainfallflow transformation models are used, among them, the most used by the technical community is the method developed by the Soil Conservation Service (SCS), due to its ease of application, parameters and relation with the physical characteristics of the river basin. However, the method is known to consider the basin in a concentrated way, not exploiting significantly the variability of its physical characteristics, and influencing the simulation of hydrological processes that are naturally variable in time and space. Because of this, several studies seek to incorporate this variability into hydrological models. In the present research, the quality of the data derived from the LiDAR technology is evaluated in comparison with data from the SRTM mission, through the extraction of drainage networks and delimitation of watersheds. In addition, a distributed approach is used for the well-known Soil Conservation Service (SCS) flow-through method where the method is applied at the pixel level to better account for watershed variability. The application of the distributed approach of the method was done through computational routines developed in Fortran, which provided the travel time, runoff depth and application of HUT per pixel. At the same time, the antecedent soil moisture condition was determined for each rainfall event, as well as the initial abstraction rate (λ), which in the concentrated approach indicates the portion of precipitation that does not generate runoff, being the value of 20% considered too high by the scientific literature of the subject. The evaluation of the digital elevation models and the distributed approach were applied in the Saint Hilaire sub-basin, located in the municipalities of Porto Alegre and Viamão, due to the availability of observed rainfall-flow data, and digital terrain and elevation models. Working in five different spatial resolutions, the results of the extraction of the drainage networks of the digital models presented the best results when compared to the reference drainage network and those obtained by the lower spatial resolutions, however, the hydrographic basins from the high resolution models represent the basin determined as a reference better. In the distributed approach of the SCS method, the runoff volume and maximum flow rate of the hydrograms were underestimated, while the peak time of the events was anticipated. The adjusted values of the initial abstraction rate (λ) resulted in values that were predominantly lower than those proposed by the traditional method, in addition, the spatial resolution of the digital model exerted little influence on the value of λ, although it represented more satisfactorily the hydrographs when higher spatial resolution is used.