Simulação chuva-vazão de uma célula de biorretenção utilizando o stormwater management model (SWMM)

Abstract

The increase in population in urban centers, evident in recent decades, has directly impacted the expansion of impermeable areas. This growth leads to significant changes in the hydrological cycle. Factors such as reduced evapotranspiration, decreased soil water infiltration capacity, increased surface runoff, and peak flows directly affect both aquatic and terrestrial ecosystems, as well as urban drainage infrastructure. In this context, Low Impact Development (LID) techniques have been widely adopted in developed countries, promoting the integration of urban and natural environments. Among these approaches, bioretention systems stand out as highly effective for urban water management, capable of reducing stormwater volume and pollutant load, and acting as buffers to mitigate runoff, thereby reducing floods and inundations. To evaluate the performance of LID techniques, including bioretention systems, hydrological models are indispensable tools. They demonstrate how stormwater moves through these storage and infiltration structures, providing a basis for their design and assessment under various scenarios. This study used the StormWater Management Model (SWMM) to evaluate the hydrological behavior of a bioretention system, specifically its storage layer, located at the Federal University of Santa Maria (UFSM) for six selected events from a series of monitored events of the structure. The simulation results revealed certain difficulties in reproducing the observed behavior of the bioretention system during modeling, as evidenced by the analysis of efficiency indicators during the calibration and validation processes. The NSE and PBIAS values ranged from -3.04 to 0.67 and -95.68% to 35.24%, respectively, during the validation process using the average of the calibrated parameters. The high number of parameters to be calibrated and factors related to uncertainties in the bioretention's drainage process (e.g., macropore formation) may have influenced these results. To address this issue, an adaptation to the model was made, which allowed the simulated results to better match the observed data, directly improving the model's performance, as evaluated by the NSE and PBIAS indicators, which during the validation procedure ranged from 0.67 to 0.97 and -20.60% to 7.86%, respectively