Since approximately 2012, Northeast Brazil dramatically suffers from the harshest drought in the recent history, with serious consequences on water resources, anthropogenic uses and ecosystem services. Among else, the hydropower production, irrigation agriculture, water supply and net cage aquaculture are the principal uses adopted in the Itaparica Reservoir, located in the Sub-Middle São Francisco River. Due to the often uncontrolled water withdrawals, the climate and land use change effects, the water quantity and quality in the reservoir has deteriorated, leading to socio-economic and environmental problems. E.g., phenomena such as harmful algae blooms (HAB) in the lentic areas are attributed to the high fluctuations of the water levels in the reservoir (up to maximum 5 m per year), due to hydropower production. Moreover, the newly built and highly argued water diversion project will withdraw water from Itaparica via two channels, the eastern one located in one of the major branches of the reservoir: the Icó-Mandantes Bay, focus of this study.
Two (depth-averaged) and three-dimensional hydrodynamic and
transport models have been implemented using the open TELEMAC-MASCARET
modeling system (2D, 3D). The aim was to provide a pioneer tool for
the region, capable to simulate different (combined) climate-, issues-
and stakeholders-oriented scenarios and, thus, to support water
management and decision-making. In this work: (1) high-resolution
unstructured grids for low and high water levels (respectively LWL and
HWL) have been set up, to assess their impact on hydraulics, tracer
transport and exchange processes between the bay and the reservoir
main stream; (2) an alternative approach was implemented to estimate
the water residence times of the bay’s complex system; (3) nutrient
emissions (e.g. phosphorus) from a net cage aquaculture system were
investigated on half-year cycles; moreover, (4) the impacts on the
flow field of the eastern channel of the water diversion; (5) the
effects of a flash flood combined with tracer transport from an
intermittent tributary and finally (6) the 3D effects induced by
moderate or extreme winds as well as by heating of the water surface
have been assessed.
The findings showed that (1) the dynamics of the bay and the reservoir main stream followed different velocity regimes (at least one order of magnitude higher for the latter, i.e. range of 10-2 to 10-1 m/s for reference conditions); (2) the bay’s water residence times were estimated to be higher than six months (about two months for the reservoir), higher for HWL and high discharges, compared to LWL and low discharges; (3) a relevant increase of phosphorus due to a small fish production of 130 t/y was observed, higher for LWL on the short term and for HWL on the long term; (4) the eastern diversion channel did not influence significantly the hydrodynamics of the bay; although, it is important to monitor constantly water quality parameters, especially during rainy periods after prolonged droughts; (5) during such events, the nutrient inputs from the tributary and the nearby drainage systems overflows will affect the water withdrawals (irrigation, water supply); (6) a windstorm increased the flow velocities (at least one order of magnitude, i.e. up to 10-1 m/s) without altering significantly the flow circulation patterns; this occurred substantially for the heating scenario, which had in contrast a lower effect on velocities.
The main implications for water management derived from the findings summarized above are outlined hereafter. (1) It is not advisable to increase the discharges and the water levels in the reservoir to stimulate water exchange processes, because it could and increase the risk of development of HAB; (2) given to the low exchange rates between the reservoir main stream and the bay, it is suggested not to install an aquaculture system inside the bay or at least to ensure sufficient water depth beneath the cages, in order to allow translocation and dilution of organic material and avoid an extreme increase of sediments; (3) the withdrawals for drinking water and irrigation agriculture should stop working during flash floods from the intermittent tributaries, as well as during windstorms, and at least three days afterwards; (4) monitoring the water quality in the eastern diversion channel is of vital importance, due to the low water depth and the high evaporation rates; (5) a heating of the water surface would likely increase the risk of development of HAB in the shallow areas, so that further assessments with a water quality module are needed to support advanced remediation measures; (6) the 3D model proves to be a necessary tool to identify high risk contamination areas, e.g. for installation of new aquaculture systems, capable of additionally taking into account wind and heating effects. In conclusion, the complex water system investigated urges of adaptive and differentiated measures to the continuously changing natural conditions and anthropogenic impacts. An efficient communication and collaboration is needed between the water users, managers and researchers, e.g. to discuss the feasibility of the proposed operation measures, such as the inversion of the water flow withdrawn by the eastern diversion channel, in the case of alarming nutrient overloads and high amount of algae in the shallow stagnant areas.
In future work, the existent models should be coupled with a water quality module to address some of the still open research questions, such as focusing on (1) the risk of HAB development, mainly on their inoculation in the lentic bay areas and their interaction with the reservoir main stream, (2) the impact of HAB on the withdrawals for drinking water or irrigation agriculture and (3) the adaptation of hydroelectric production to reduce water level fluctuations, in order to minimize the introduction of nutrients from the desiccated soils in the shallow areas and, thus, the greenhouse gases (GHG) emissions as well. Moreover, external forces such as wind, heating and cooling processes should be always included in the modelling, since they influence indeed the hydraulics of water bodies such as the Icó-Mandantes Bay.