Prediction of total phosphorus in reservoir cascade systems

Reservoir cascade systems have attracted the attention of scientists worldwide. The present study investigates the cascade of five reservoirs (R1, R2, R3, R4, and R5) along a 192-km water channel system located in the state of Ceará, in the Brazilian semiarid region. This cascade system was implemented in 2012 to promote water availability and security to the capital of Ceará and the strategic industry and port complex of the region. However, these reservoirs have faced a progressive degradation of water quality, which has resulted in intense eutrophication and high-water treatment costs. The study evaluates the dynamics of water quality from 2013 to 2021 along this reservoir cascade (from R1 to R5). The results revealed that water quality did not improve along the cascade system, differently from previous studies on reservoirs interconnected by natural rivers. This was attributed to the low water residence time and low capacity of pollutant removal along the man-made water channel system, as well as to the high internal phosphorus loads of the reservoirs. Multiple regression models involving the explanatory variables of total phosphorus, total nitrogen, chlorophyll-a, cyanobacteria, transparency, rainfall, and volume from upstream reservoirs were obtained to determine total phosphorus concentration in downstream reservoirs, considering different combinations of reservoir pairs in the cascade and different time delays. A clear trend of R2 decline with the distance between the upstream and downstream reservoirs was observed. For example, the R2 values for the correlations adjusted between R1 and R2 (48 km), R1 and R3 (172 km), R1 and R4 (178 km), and R1 and R5 (192 km) were 0.66, 0.32, 0.22, and 0.12, respectively. On the other hand, the adoption of time delays of the order of the cumulative residence times of the reservoirs promoted a significant improvement in the R2 values. For instance, the best correlation adjusted between R1 and R5 improved from R2 = 0.12 to 0.69 by considering a time delay of 21 months. This suggests that previous data from upstream reservoirs can be used to predict current and future total phosphorus concentration in downstream reservoirs. The results from this study are important to better understand the spatiotemporal dynamics of water quality in reservoir cascade systems and thus improve water resources management, especially in drylands.

Nitrogen mass balance and uptake velocity for eutrophic reservoirs in the Brazilian semiarid region

The nitrogen (N) cycle from the catchment to the downstream reservoir is complex, particularly the quantification of N losses. However, in order to assess the nitrogen impact in a reservoir ecosystem, simplified models may be applicable regarding the TN load production and the magnitude of lake TN removal. This study presented a methodology to perform and validate a TN mass balance to further calibrate a simplified coefficient for TN losses (vf.) in 29 tropical semiarid reservoirs. The study reservoirs were highly productive ecosystems with an average total nitrogen (TN) concentration, accounting for all measurements in all reservoirs, ranging from 0.59 to 3.84 mg L-1. Regarding the production of TN load, the median values ranged from 4.35 to 2,499.43 t year-1 with median of 80.34 t year-1. The TN loads were estimated through an annual mass balance over a 24-year period. The median of the estimates was compared with reference values obtained by using the export modelling coefficient. The correlation between the median estimated and reference loads resulted in satisfactory agreement (r2 0.88) and reinforced the reliability of the mass balance alternative. From the validated TN loads, the TN uptake velocity (vf) was estimated for all reservoirs (44.9 ± 20.1 m year-1) and could be described as a general function of the water residence time. The reservoirs of the study region have demonstrated higher vf than temperate lakes and reservoirs and similar vf with Latin America/Caribbean ones. As expected, reservoirs of warmer climates tend to present intensified N loss processes compared to lakes and reservoirs of temperate regions. The methodology proposed in the present study can be used to potentially improve water quality management in tropical semiarid reservoirs.

Chlorophyll-a prediction in tropical reservoirs as a function of hydroclimatic variability and water quality

The study goal was to determine spatiotemporal variations in chlorophyll-a (Chl-a) concentration using models that combine hydroclimatic and nutrient variables in 150 tropical reservoirs in Brazil. The investigation of seasonal variability indicated that Chl-a varied in response to changes in total nitrogen (TN), total phosphorus (TP), volume (V), and daily precipitation (P). Therefore, an empirical model for Chl-a prediction based on the product of TN, TP, and normalized functions of V and P was proposed, but their individual exponents as well as a general multiplicative factor were adjusted by linear regression for each reservoir. The fitted relationships were capable of representing algal temporal dynamics and blooms, with an average coefficient of determination of R2 = 0.70. The results revealed that nutrients yielded better predictability of Chl-a than hydroclimatic variables. Chl-a blooms presented seasonal and interannual variability, being more frequent in periods of high precipitation and low volume. The equations demonstrate different Chl-a responses to the parameters. In general, Chl-a was positively related to TN and/or TP. However, in some cases (22%), high nutrient concentrations reduced Chl-a, which was attributed to limited phytoplankton growth driven by light deficiency due to increased turbidity. In 49% of the models, precipitation intensified Chl-a levels, which was related to increases in the nutrient concentration from external sources in rural watersheds. Contrastingly, 51% of the reservoirs faced a decrease in Chl-a with precipitation, which can be explained by the opposite effect of dilution of nutrient concentration at the reservoir inlet in urban watersheds. In terms of volume, in 67% of the reservoirs, water level reduction promoted an increase in Chl-a as a response to higher nutrient concentration. In the other cases, Chl-a decreased with lower water levels due to wind-induced destratification of the water column, which potentially decreased the internal nutrient release from bottom sediment. Finally, applying the model to the two largest studied reservoirs showed greater sensitivity of Chl-a to changes in water use classes regarding variations in TN, followed by TP, V, and P.

Uncovering the influence of hydrological and climate variables in chlorophyll-A concentration in tropical reservoirs with machine learning

Climate variability and change, associated with increasing water demands, can have significant implications for water availability. In the Brazilian semi-arid, eutrophication in reservoirs raises the risk of water scarcity. The reservoirs have also a high seasonal and annual variability of water level and volume, which can have important effects on chlorophyll-a concentration (Chla). Assessing the influence of climate and hydrological variability on phytoplankton growth can be important to find strategies to achieve water security in tropical regions with similar problems. This study explores the potential of machine learning models to predict Chla in reservoirs and to understand their relationship with hydrological and climate variables. The model is based mainly on satellite data, which makes the methodology useful for data-scarce regions. Tree-based ensemble methods had the best performances among six machine learning methods and one parametric model. This performance can be considered satisfactory as classical empirical relationships between Chla and phosphorus may not hold for tropical reservoirs. Water volume and the mix-layer depth are inversely related to Chla, while mean surface temperature, water level, and surface solar radiation have direct relationships with Chla. These findings provide insights on how seasonal climate prediction and reservoir operation might influence water quality in regions supplied by superficial reservoirs.

Internal phosphorus loading and its driving factors in the dry period of Brazilian semiarid reservoirs

This study investigated the relationships between physical, limnological and climatic drivers with the internal total phosphorus (TP) loading produced over the dry period in 30 water supply reservoirs of the Brazilian semiarid. Improvements in the understanding of sedimentary TP fluxes in reservoirs of dryland regions are pressing as they usually have serious water quality related issues, remaining mostly eutrophic especially under frequent drought events. Gross daily fluxes and net seasonal average release rates were calculated from mass balance and regression equations considering water and sediment TP concentrations, anoxic duration, water temperature and fish contribution. Additionally, the ratio of wind speed to reservoir volume was proposed as a new surrogate and then applied as explanatory variable to predictive models. The results indicated TP release rates higher than reported for non-semiarid lakes/reservoirs with average gross fluxes ranging from 17.64 to 35.99 mg m^-2 day^-1. This may be attributed to the enriched sediments (1029.49 ± 552.49 mg kg^-1) allied with warmer water temperature, high trophic state, and prolonged anoxic periods (average duration of about 60 days). The average release rates were negatively correlated with water transparency and water depth, and positively correlated with Chl-a, wind speed and trophic state. The release rates increased across the trophic gradient (p < 0.05), about 10-fold higher under hypertrophic conditions than in oligotrophic ones. As anoxia is linked with eutrophication, phosphorus release is more likely in eutrophic ecosystems. Regarding the new surrogate, a strong predictive ability for TP release (R²: 0.26-0.93) was observed. Similarly, the proposed models presented a physically consistent behavior with a stabilizing releasing pattern suggesting the achievement of equilibrium in nutrient exchange between sediment-water interface. This research advanced by combining and proposing methods to assess and quantify sedimentary fluxes in data-scarce regions balancing accuracy and transferability, in order to be replicable to other dryland environments globally.

Phosphorus mass balance and input load estimation from the wet and dry periods in tropical semiarid reservoirs

The dynamics of total phosphorus (TP) in 18 strategic reservoirs of the high-density reservoir network of the Brazilian semiarid was evaluated during the wet and dry periods for the past 12 years. Seasonal overlying concentrations presented no significant differences for about 90% of the reservoirs (p>0.05). This was attributed to a trade-off between the hydrological/limnological processes occurring in the two seasons. Then, a transient complete-mix mass balance model was applied with particular adaptations for the tropical semiarid reservoirs to estimate the TP load for each season. Because of the relatively well-mixed conditions and high hypolimnetic dissolved oxygen concentrations during the wet season, the wet load was assumed to represent the external TP load. On the other hand, because of the absence of reservoir inflow during the dry season, phosphorus release under anoxic sediment conditions and wind-induced resuspension under shallow water depths, the dry load was assumed to reflect the internal TP load. The maximum external loads were related to peak inflows, notably after drought periods. Consistently, the largest internal loads were obtained during the drought periods, when the reservoirs were shallower and more prone to phosphorus release and resuspension. By comparing the impact of the two input load types, the wet period load was predominant in 72% of the reservoirs. The areal phosphorus loads ranged from 0.66 to 7.29 gP m² year^-1, which were consistent with the literature, despite the very high density of reservoirs. Finally, power-law curves including data for all studied reservoirs were adjusted between the dry period load and volume, dry and wet period loads, wet period load and inflow, and total load and catchment area, resulting in satisfactory R² (0.84-0.98).

Modeling flow-related phosphorus inputs to tropical semiarid reservoirs

Hydrological data and total phosphorus (TP) concentration at reservoirs' outlet were combined in a transient complete-mix model to obtain mean input loads and inlet concentration-flow relationships. This approach was designed to investigate the issue of phosphorus pollution in semiarid regions with intermittent rivers. The methodology was applied for twenty reservoirs in the State of Ceará, Brazilian semiarid. The modeled TP loads correlated well (R² = 0.74) with reference loads estimated from environmental inventories, with only 10% of underestimated results. The average input loads per unit area of the catchments ranged from about 4 to 40 kg km^-2 yr^-1, which were considerably lower than the national average of about 500 kg km^-2 yr^-1. This was attributed to lower precipitation indexes, intermittent river regime and a high-density reservoir network, peculiar of the Brazilian semiarid. Meanwhile, the input load per unit area of a small and highly populated urban catchment, with higher precipitation indexes and deficient sanitation was substantially higher (2626 kg km^-2 yr^-1). Moreover, the fitted TP concentration-flow relationships directly reflected different TP input sources: strong u-shaped behavior marked the curves of highly non-point source dominated catchments, whereas a dilution pattern prevailed in those with significant point source inputs. The model validation with measured riverine TP concentration reached a NSE of 0.63. However, peak values in TP concentration during low flow rates sensitively affected the fitting of the models. In spite of non-point source dominance in the catchments, some relationships presented a slight signal of this use type. The variation range of the fitting parameters in comparison with other studies, as well the expected behavior of the curves in light of land use characteristics, strongly support the methodology applied in this study. The proposed approach will potentially help address the TP issue in tropical semiarid regions. Furthermore, the paper presents a simple way to deal with the challenging lack of monitored data in such environments.