Ferramenta computacional para incorporar a trajetória de tormentas na modelagem hidrológica com o Hidropixel

Abstract

Floods rank among the most recurrent environmental disasters in Brazil, producing substantial human and economic impacts. Their severity depends not only on total rainfall depth but also on the spatial organization and movement of precipitation over the watershed. Because rainfall systems rarely behave in a stationary manner – particularly in medium- to large-sized basins –

the assessment of distinct storm trajectories becomes essential for understanding critical flood- generation scenarios. Storm direction, orientation, and translation velocity may synchronize or

desynchronize runoff generation and propagation processes, thereby modifying peak discharge, time to peak, and overall hydrograph shape. Within this context, this study developed a computational tool integrated into QGIS to generate moving spatiotemporal precipitation fields and incorporate them into hydrological modeling using the Hidropixel–Distributed Linear Reservoirs (Hidropixel-DLR) framework. The system combines a Python-based graphical user interface with computational routines implemented in Fortran 90 to reduce computational cost. It enables explicit control of storm translation through predefined trajectories, velocity, and radius of influence, while also allowing the simulation of spatially uniform or spatially distributed rainfall patterns. The tool was applied to the upper Tapacurá River basin (Pernambuco State, Brazil), delineated up to the Vitória de Santo Antão stream gauge (39170000), encompassing an area of approximately 262 km2. Upstream–downstream (TMJ) and downstream–upstream (TJM) storm trajectories were evaluated under different storm influence areas. The results demonstrated a clear influence of storm trajectory on hydrological response. Under spatially uniform rainfall conditions, the TMJ trajectory produced a peak discharge of 933.0 m3/s at 21.0 h, whereas the TJM trajectory resulted in 263.87 m3/s at 28.75 h, representing a 3.5-fold increase in peak flow and an earlier time to peak in the upstream– downstream direction. When spatially and temporally variable rainfall was considered, peak discharges reached 307.04 m3/s (19.75 h) for TMJ and 156.43 m3/s (20.0 h) for TJM, preserving the directional bias while modifying hydrograph structure. Furthermore, increasing the storm radius from 4 km to 8 km substantially amplified peak discharges (from 307.04 to 954.42 m3/s in TMJ – approximately threefold – and from 156.43 to 640.27 m3/s in TJM), partially attenuating the directional bias. Overall, the proposed tool enhances the physical representation of the rainfall–runoff process, providing support for hydrological studies focused on extreme events and flood risk management.