Monitoramento, dinâmica e modelagem hidrossedimentológica de duas bacias hidrográficas rurais com produção de leite e grãos

Abstract

The observation and use of hydrological data represents the key to the resilience and sustainability of the use of water resources and soil. Furthermore, computational modeling techniques can be used as complementary strategies to investigate water and sediment flows in rivers. In this context, the thesis addresses the monitoring of hydrosedimentological processes and the application of modeling tools in two paired rural watersheds, with drainage areas of 1.14 km² (northern watershed) and 0.67 km² (southern watershed), in the southern from Brazil. The southern watershed has 59% more riparian forest than the northern watershed and also more intensive dairy activity. The drainages of the two watersheds are connected to the Passo Real water reservoir, one of the largest reservoirs in southern Brazil, with great importance for energy generation and supply. The first chapter of this thesis deals with the investigation of hydrological and sedimentological processes at the scale of rainfall events in the two paired watersheds. We investigated the effect of riparian vegetation in the two watersheds on surface hydrology and sediment transport. In addition, we investigated the connectivity aspects of these watershed, representative of regional conditions, with the Passo Real reservoir. The southern watershed, even with a smaller size, presented higher surface runoff and peak flow for most of the observed events, even with similar land uses to the northern watershed. On the other hand, the total production of specific sediments was higher for the northern watershed. This behavior can be explained due to the practice of minimum soil disturbance, which means greater disturbance, promoting greater production of sediments. The second chapter addresses the identification of sediment sources. Therefore, knowing the main origin of sediments is valuable information for decision making. The methods for determining the optical emission tracer properties of inductively coupled plasma and X-ray fluorescence confirmed that roads presented the greatest contributions from suspended sediment sources (between 84.3% and 89.2%), followed by crops and drainages, for the two watersheds. The third chapter describes the unprecedented and challenging efforts to apply LISEM to simulate the flow and production of sediments in the study area. We simulated a robust set of various rainfall events, divided into calibration and validation steps. The model satisfactorily represented the hydrology and sediment production of the simulated events for the calibration step. On the other hand, in the validation stage, the model did not adequately represent the processes, except for some events. This result is attributed to the uncertainties related to how the calibrated variables represent the variables observed in the field and, mainly, to the effect of spatial variability on the representation of hydrological processes. Strategies for watershed-scale monitoring, despite being a major challenge in environmental sciences in emerging countries, are also crucial investments to identify changes in hydrological systems, especially in the face of the effects of climate and ecosystem changes.