Best management practice (BMP) implementation strategies have often overlooked the integration of both nutrient source and local environmental factors. This research developed a targeted BMP implementation approach using parcel-level nutrient load contribution alongside land use type, slope, depth to water table, stream order, and distance to watershed outlets. The Upper Kissimmee and Taylor Creek and Nubbin Slough watersheds were selected to represent contrasting land use patterns. Ten incremental scenarios, ranging from 10% to 100% of source cells' cumulative nutrient contribution, were mapped in ArcMap 10.4. Four BMPs, chosen based on technological and economic criteria, were applied to suitable land uses, and nutrient reduction parameters were incorporated into the Watershed Assessment Model to simulate watershed-scale effects. Results showed the greatest TN reduction of 1452 and 1204 tons under the 100% scenario for two watersheds, while the highest cost effectiveness occurred under the 10% scenario, with 1.33 and 1.82 tons of TN removed per million dollars. BMPs implemented in citrus groves and improved pasture contributed most to TN reduction. Marginal nutrient reduction was primarily influenced by the additional nutrient reduction and costs associated with newly added land use types, while changes in existing land use types dominated overall nutrient reduction. To maximize watershed-scale nutrient load reduction, BMPs should be prioritized in source cells with higher baseline nutrient load, greater contribution of nutrient load, and near higher-order streams.

Coastal communities throughout the (sub)tropics face increasing threats from tropical cyclones, prompting the need to develop innovative resiliency strategies. While grey infrastructure such as retaining/seawalls have long been the default solution for protecting coastal properties from destructive storm surge flooding and wave impacts, coastal wetlands offer a potential enhancement that balances shoreline protection with ecological benefits. In this study, we evaluate the outcomes and implications of a narrow mangrove belt seaward of vertical seawalls, and the impacts that they would have in an urban, developed environment at the community/neighborhood scale. We simulated eighteen hybrid designs which tested three seawall crest heights (0, 1 m, and 2 m) and six mangrove forest widths (0 - 50 m) using a coupled high-resolution hydrodynamic-wave model of a shoreline in Tampa Bay, Florida (USA). For each design, performance was evaluated under three major storm surge events that caused extensive flood damage in recent years to the local community. Our results show that just 10 m of mangrove forest can attenuate wave heights by 64%, compared to just 22% wave height attenuation over the same distance of unvegetated shoreline. However, with regards to reduction of storm surge, forest widths up to 50 m were not sufficient to reduce storm surge levels. An elevated seawall improved flood protection when the crest height exceeded the storm surge level but posed drainage issues. Hybrid designs composed of a narrow mangrove belt in front of a seawall reduced wave energy reaching inland and impacting the grey infrastructure more effectively than an unvegetated seawall. These results highlight the practicability and potential benefits of integrating nature into urban resiliency strategies.

Rivers are dynamic conduits for microplastics, yet the interplay between polymer chemistry, environmental exposure, and microbial colonization remains poorly understood. We conducted a 24 month in situ study in the Hillsborough River, Florida (USA) assessing the aging of five polymer types— low-density polyethylene (LDPE), high-density polyethylene (HDPE), polypropylene, polystyrene (PS), and polyethylene terephthalate (PET)—across five environmental layers: air, surface water, middle water, deep water, and sediment. Particles were analyzed using scanning electron microscopy, flow cytometry, Fourier transform infrared spectroscopy, and image-based shape metrics. Aging patterns were strongly layer- and polymer type-dependent: PET and PS exhibited the highest chemical weathering (carbonyl indices up to 12.7 and 4.9, respectively) and microbial colonization (average cell counts of 1.5 × 10 ^5 ± 1.2 × 10 ^5 and 1.3 × 10 ^5 ± 1.0 × 10 ^5 , respectively), while LDPE and HDPE supported substantial biofilm formation (average biomass of 4.6 ± 4.1 mg and 6.5 ± 7.6 mg, respectively) with limited photooxidation (average carbonyl indices of 0.7 ± 0.6 and 0.7 ± 0.7, respectively). Air-exposed particles underwent continuous photooxidation, surface and middle water layers promoted dense microbial biofilms (average biomass of 5.6 ± 5.2 mg and 4.1 ± 3.8 mg, respectively), and deep water and sediment showed slower colonization and lower chemical alteration (average carbonyl indices of 2.1 ± 3.0 and 2.1 ± 3.6, respectively). Mass and particle shape remained largely stable, highlighting microplastic persistence under natural riverine conditions. Computer simulations corroborated colonization trends, though sediment dynamics were underestimated, emphasizing the need for refined modeling. These results demonstrate that polymer type and environmental context govern early aging processes and suggest that targeted management strategies should account for both microplastic polymer type and riverine layer-specific transport.

Boundary currents transport salt and heat along shelf breaks and exhibit distinct subtidal variability. The proximity of boundary currents to the coast suggests they could influence estuarine boundary conditions. However, the influence of boundary currents on estuarine dynamics is underexplored, and a mechanistic framework for the interaction remains underdeveloped. This study explores the subtidal interaction between the Gulf Stream, a western boundary current, and the St. Lucie Estuary, Florida, USA. The direct and comparative effects of the Gulf Stream on estuarine water levels, salinity, and temperature were quantified through statistical analyses and Variance Partitioning. The influence of the Gulf Stream on subtidal estuarine mass transport was isolated using scenario analyses through process-based numerical modeling. Results show that variations in Gulf Stream transport are transmitted into the estuary through a quasi-geostrophic balance, establishing a physical pathway linking boundary current variability to subtidal estuarine dynamics. Gulf Stream transport was inversely correlated with estuarine water levels and explained more variance than river discharge and local winds. Additionally, the variations in Gulf Stream transport impacted density distribution and subtidal mass transport. Decreases in Gulf Stream transport increased average estuarine density, strengthened longitudinal density gradients, and engendered net landward mass transport. Increases in Gulf Stream transport decreased average estuarine density, weakened longitudinal density gradients, and engendered net seaward mass transport. These responses should have implications for contaminant dispersion, residence times, and sediment transport.

Predicting algal blooms in waterways receiving inflows from multiple sources is challenging since blooms and their drivers can originate from diverse sources. Models that mechanistically simulate the formation and transport of algal blooms are often computationally intensive, creating barriers to using them for daily decision- making. Given this challenge, we developed a statistical risk forecasting framework for the Caloosahatchee River and Estuary in southwest Florida, United States of America, which receives engineered water releases from the eutrophic Lake Okeechobee, as well as hydrologic inputs from the surrounding watershed. The forecasting approach considers two different hydrologic regimes (i.e., lake- versus runoff-dominated conditions) while maintaining structural simplicity such that water managers could readily apply the model for short-term decision-making. Using daily mean discharge at two United States Geological Survey stations and more than 14 years of discrete water quality sampling data with diverse temporal resolution from the South Florida Water Management District, two regression tree models were trained and tested for day-ahead bloom risk forecasting: one for conditions in which lake releases dominated, and one when watershed inputs dominated. For the model capturing lake-dominated conditions, the main bloom predictors were 30-day lagged total suspended solids at the lake and canal outlets averaged over transect residence time (R2 = 0.78 and RMSE = 6.10 mu g/L). For the model predicting watershed-dominated conditions, 30-day lagged dissolved phosphorus load, and chlorophyll-a averaged over transect residence time at the canal outlet were the most important predictors of next-day chlorophyll-a (R2 = 0.49 and RMSE = 14.50 mu g/L). Critically, algal blooms driven by Lake Okeechobee releases are potentially controllable, as they are related, at least in part, to the operation of water control structures. Thus, the strong performance of the model for lake-dominated conditions demonstrates its utility for informing release scheduling to mitigate downstream blooms.

This study explored the use of data-driven models to develop management-oriented prediction tools for algae blooms (ABs) represented by Chlorophyll-A (Chla) concentrations, using the Caloosahatchee and St. Lucie canals in Lake Okeechobee, Florida, as case studies. By comparing two modeling approaches, i.e., cascading modeling and time-lag modeling, the study aims to understand the differences in Chla dynamics between the two canals, identify the main drivers and predictors of Chla concentration in each, and develop suitable forecasting models for the canals' operation purposes. Throughout this study, both approaches demonstrated their value in improving the understanding of water quality dynamics in Lake Okeechobee canals. While some water quality parameters such as Dissolved Oxygen (DO) and Nitrate-Nitrite (NOx) were critical to ABs in the Caloosahatchee and St. Lucie canals, respectively, the effect of operation decisions on ABs was more significant on the St. Lucie than on the Caloosahatchee. From a modeling perspective, the time-lag modeling approach achieved higher predictive accuracy for Chla concentrations in both Caloosahatchee and St. Lucie canals. Particularly, at station S80 of St. Lucie canal, the XGBoost (XGB) algorithm achieved R2= 99% and RMSE = 0.001 μg/l in training, and R2= 60.1% and RMSE = 4.58 μg/l in testing. At station S79 of Caloosahatchee canal, Random Forest (RF) appeared to be the best model with R2= 85.7% and RMSE = 5.63 μg/l in training, and R2= 39% with RMSE = 10.06 μg/l in testing. In this study, the time-lag modeling approach was proven to offer decision-makers flexible tools for implementing better management strategies based solely on operation decisions and meteorological conditions.

Arias ME, Tarabih OM, Hasani SS, James RT, Kaplan DA. 2025. Water quantity and quality modeling of Lake Okeechobee, Florida: a systematic and critical literature review. Lake Reserv Manage. XXX-XXX.Lake Okeechobee is the largest lake in the southeast United States. It is a critical water resource that supports South Florida's water supply, flood control, and diverse aquatic ecosystems. Lake Okeechobee's management is constrained by balancing in-lake storage capacity for flood protection, water quality requirements for flows to the Everglades, and increasingly deleterious discharges to Florida's coasts. Research has improved our understanding and prediction of Lake Okeechobee's physical, chemical, and biological properties, often with the aid of computer models. This article is an overview of scientific literature related to water quantity and quality modeling of Lake Okeechobee and its contributing watersheds. We reviewed 115 studies published since 1980. Lake water quality and nutrient load reduction from the contributing watersheds were the focus of 45 and 34 of these papers, respectively. These modeling studies improved our understanding of the Lake Okeechobee environment on several fronts. They supported field and lab research on sediment-water interactions that found internal loads were as important as external loads in affecting water quality. Models also verified the concept that sediment phosphorus settling/resuspension could be enhanced by shallow depth, flocculent mud sediments, and wind-driven waves. Additionally, modeling has improved our understanding of algae growth limitation, a dynamic multifaceted phenomenon that does not align with single-nutrient limitation typical of deeper, colder lakes. Looking ahead, current limitations and future directions for how lake modeling could support management actions include the incorporation of machine/deep learning in forecasting, along with a deeper understanding of both short- and long-term environmental changes.

Study region: Irrawaddy River, Myanmar, Southeast Asia Study focus: Detection of streamflow changes serves as a critical foundation for water resources management. The data-scarce Irrawaddy is ideally positioned for this study, as investigations of regional hydrology are lacking, despite being one of the largest rivers in Southeast Asia. Hence, this work achieves an unprecedented and detailed analysis of the Irrawaddy streamflow variability from 2001 to 2023 due to climate, dams, and land cover change. The assessment integrated trend analyses with diagnostic modeling using the Variable Infiltration Capacity with reservoir module (VIC-Res) for two sets of scenarios: a set with and without dam operations, and another without dams using fixed 2001 and 2023 land cover. New hydrological insights for the region: Streamflow significantly decreased in the lower Irrawaddy. Moreover, the Irrawaddy streamflow variability was mainly attributed to climate for two proven reasons. First, both land cover change and dam operations contributed marginally to mean seasonal and annual streamflow changes (around +/- 2 %) due to small land cover change basin-wide (within +/- 5 %) and that all dams are on low-order tributaries with the degree of regulation of only 2 %. Compared to the effect of land cover change on streamflow, dams exerted higher flow changes in magnitude and variability. Second, precipitation and streamflow maintained a statistically significant linear correlation (p < 0.01) and exhibited similar decreasing tendency, despite no significance detected for precipitation. This study enhances understanding of the Irrawaddy's contemporary hydrology to support basin-scale water resources management.

In constructed wetland ecosystems, physical, chemical, and biological processes work together to weather microplastics over time. In this study, the aging of five microplastic polymer types (low-density polyethylene, high-density polyethylene, polypropylene, polystyrene [PS], and polyethylene terephthalate [PET]) was contextualized in four habitats within a wastewater constructed (treatment) wetland over 18 months. This is the first study to analyze the weathering and microorganism colonization of different types of microplastics as they age within a constructed wetland. A multistep analysis approach was employed to assess microplastic aging. Microorganisms quickly and abundantly colonized all polymer types, with the highest biomass and cell counts being found on particles at the wetland cell outlet. Overall, the wetland cell outlet showed the highest metrics of weathering, and PET and PS were found to be the most weathered within this system. No complete disintegration was found in this study, underscoring the durability of microplastics in aquatic ecosystems. As microplastics abundantly flow through wastewater treatment streams and receiving constructed wetlands have demonstrated strong retention potential, it is important that environmental managers better contextualize the aging abilities of these systems.

Harmful algal blooms have large impacts on aquatic ecosystem and human health. Nutrient enrichment, in combination with warm water temperatures, high sunlight availability, and low water turbulence, have been proven to be major factors driving algal blooms. In this study, lake eutrophication processes, including phytoplankton production and nutrient cycling, were simulated and coupled with a reservoir operations model to optimize multi‐criteria lake operation goals. The main objective of this study was thus to design reservoir operations that would minimize phosphorus (P), nitrate‐nitrogen (NOx), and phytoplankton loads to downstream water bodies, while meeting other societal water resource demands in eutrophic lakes. We used an open‐source, multi‐objective evolutionary algorithm framework with four optimization objectives (minimizing P, NOx, and phytoplankton loads and water demand deficits), assessing each constituent separately and in combination. In addition, different optimization scenarios associated with each objective were investigated. To effectively demonstrate our findings, we implemented our approach in Lake Okeechobee, the largest subtropical lake in the US. We identified multiple opportunities to reduce downstream loads while minimizing impacts on water demand deficits. Notably, considering combined load objectives yielded substantial reductions in summertime P, NOx, and phytoplankton exports by up to 73%, 82%, and 73%, respectively, with minimal increases in water demand deficits. This supports the idea that alternative operational strategies could provide an effective and economical reservoir management strategy for balancing downstream water quality and societal water resource needs.

We introduce a significantly enhanced version of AquaNutriOpt, now equipped with advanced mathematical optimization capabilities absent in its initial release (Khanal et al., 2024). AquaNutriOpt II is a user-friendly, free, open-source Python tool designed to address the complex challenge of optimizing nutrient management for controlling harmful algal blooms. In this latest version, users gain the flexibility to incorporate multiple time periods into their analyses. Moreover, they can now optimize the management of two nutrients concurrently (primarily phosphorus and nitrogen) through an innovative multi-objective optimization framework. Building upon its predecessor, AquaNutriOpt II continues to streamline the identification of optimal Best Management Practices (BMPs) and Treatment Technologies (TTs), including determining the most suitable locations for implementation while considering budgetary constraints. To showcase the efficacy and advantages of AquaNutriOpt II, we apply it to a real-world case study centered on Lake Okeechobee in Florida, USA.

Densely populated coastlines around the world are exposed to hazards such as storm surges from tropical cyclones. Coastal wetlands have the potential to reduce flooding during tropical cyclones by attenuating storm surge and wave energy. While previous research has demonstrated attenuation over tens of kilometers, the benefits over narrow coastal fringes remain largely unquantified. In this study, we used a 2-D high resolution model to quantify the value of coastal wetlands for flood mitigation and assess their potential role during extreme weather events. To provide context, we evaluated attenuation in Charlotte Harbor, Florida, during Hurricane Ian (2022), which made landfall as a Category 4 hurricane, costing over 150 human lives and over USD112 billion in damages. The model, developed in Delft3D Flexible Mesh and coupled with the SWAN wave model, features an unstructured grid and an explicit representation of frictional forces caused by vegetation structure (canopy height, diameter, and density). We compared inundation depth and water velocity across four cross-shore transects under vegetated and non-vegetated scenarios and assessed structural damage costs. Results show storm surge water levels were reduced 6–17

Urbanization, driven by population growth, alters watershed hydrology and nutrient runoff. However, the complex interplay between urbanization and nutrients in regional watersheds remains an open question. This study assessed how urbanization affects streamflow, total nitrogen (TN), and total phosphorus (TP) loads in six diverse Florida watersheds covering an area of 10,600 km2. This was carried out by introducing 2070 land use/land cover (LULC) projections to a watershed hydrology/water quality model. We investigated how different levels of urban density, as a proxy for urbanization patterns, affect streamflow and nutrient variability. Results indicate that urban land could increase from 14% to 27% in 2070. This expansion could lead to monthly streamflow increases of 0%-36%, based on watershed and urbanization patterns. Future TP loads could change by -8% to +140%, with decreases attributed to LULC transitions from high-use fertilizer agriculture to low/medium density residential classes. Projected TN loads are more consistent, with simulated changes of -1% to +26%. Among LULC transitions, the largest increases in TP and TN are caused by potential urbanization of freshwater wetlands. This study provides knowledge relevant to regions undergoing similar urbanization trends, enabling managers to make better land development plans with water quality considerations. It also contributes a detailed modeling framework that can be adopted even with the use of different LULC datasets and software.

To provide a better subseasonal-to-seasonal (S2S) hydrological forecast, it is essential to investigate the factors that control streamflow prediction at time scales beyond that of traditional weather forecasts. Using a hydrological forecast framework built around NASA's Catchment-CN land model and GEOS S2S forecast meteorology, this study examines the predictive skill of subseasonal (similar to 30 days) streamflow in Southeast Asia and shows how that skill may be improved in combination with satellite-based rainfall information in areas for which the rain-gauge measurements are particularly poor. Initialized at four different times of a year, the prediction skill along the Irrawaddy River in Myanmar was significantly improved, going from no skill up to a correlation coefficient R of 0.65 during the wet season and up to 0.55 during the following transitional period by introducing Integrated Multi-satellitE Retrievals for GPM (IMERG) satellite-based precipitation into our land initialization methodology. The streamflow forecast skill along the Mekong River was reasonably high (R of 0.6-0.7) during the dry season before and after the utilization of IMERG data, and the wet-season forecast skill modestly increased up to R of 0.8. The accurate land initialization is found to contribute dominantly to the predictive skill of subseasonal streamflow; however, low rainfall forecast skill occasionally offsets the positive contribution from the land initialization. Our findings suggest an alternative way to enhance S2S hydrological forecasting in other large river basins where rain gauge information is limited and illustrate the need for a careful application of forecast rainfall to hydrological prediction during the transitional seasons.

This paper introduces a model of constructed wetlands (CWs) for on-site landfill leachate treatment, emphasizing the efficacy of zeolite and biochar as adsorbent media to enhance pollutant removal. While zeolite improves nitrification, biochar enhances denitrification and organic matter removal. The model employed a series of continuously stirred reactors and was calibrated and validated using data from 2 parallel CW mesocosms. Each mesocosm incorporated a 250-L vertical subsurface flow CW (V-CW) followed by a 440-L horizontal subsurface flow CW (H-CW). One was an unamended control with gravel media and the other was amended with zeolite in V-CW and biochar in H-CW. The model considered hourly changes in water storage, temperature, and transformations of nitrogen species, chemical oxygen demand (COD), and dissolved oxygen. Model robustness was demonstrated by low normalized root mean square error values (<0.3) for effluent ammonium nitrogen and COD concentrations. The model, validated and subject to a sensitivity analysis, accurately predicted pollutant removal efficiency (<13.3% error for unamended, <3% for adsorbent-amended), making it a valuable tool for evaluating full-scale CW dynamic performance under variable conditions.

The influences that weathering processes may have on plastic settling have not been well documented or quantified. In this study, the physical characteristics and settling velocities of environmentally weathered micro- and macroplastic fibers collected in the Mekong River in Southeast Asia were measured in a laboratory setting for the first time. Size and shape ranges were larger than previous studies in which lab-generated particles were employed for experimentation. Measurements were used to evaluate several current drag equations. Out of four tested drag equation candidates, the Zhang & Choi (2022) equation was best at predicting the behaviors of weathered particles with an average error of 18% and root-mean-square error of 2.4%. The impacts of settling on estimating plastic loads were demonstrated using a plastic transport model parametrized for the Mekong directly upstream of Phnom Penh. Results suggest that a much larger fraction of plastic particles move toward the bottom of the water column than estimated with the conventional settling velocity method (97% vs 1%). These findings provide essential information on the settling characteristics of environmentally weathered microplastics, which will subsequently allow transport models to better represent the behaviors of plastics found in the aquatic environment.

Constructed wetlands (CWs) are well-proven ecologically engineered systems that remove nutrients, pathogens, and other contaminants from wastewater. The inherent role of CWs as wastewater polishing systems makes them hotspots for microplastic accumulation. To investigate the transport and characteristics of microplastic pollution in full-scale surface flow municipal CWs, environmental microplastic samples were collected from the Se7en/Seven Wetlands in Lakeland (Florida), one of the largest wastewater CWs in the United States (650 hectares). Both water and sediment samples were collected at eight control stations in each of the seven wetland cells and at six locations in the influent distribution channel to investigate the linkage between water and sediments in microplastic fate and transport. This study found that the Se7en Wetlands has a near complete retention of microplastics greater than 300 mu m. Concentrations at each sampling site in the CW ranged from 0 to 17 particles/m3 in water and 0-1217 particles/kg dry weight in sediment. The highest concentrations of particles were found in sediment samples, especially those found in locations of reduced flow, signaling settling occurring as water velocities decrease. Microplastics were also analyzed for size, shape indicators, and polymer type to inform how their properties change as they travel through the system. Particles were extracted visually and analyzed using ImageJ and Raman spectroscopy. Polyethylene and polypropylene were the most abundant polymer types with the highest proportion of particles being fragments between 1 and 5 mm in size. Shape indicators (particle area, circularity, and solidity) in water decreased through the treatment stream, whereas sediment samples saw no clear trend. Overall, this study finds that large CWs retain a vast amount of microplastics. It sheds light on how the dynamics of natural processes affect the fate and transport of microplastics as well as the importance of evaluating multimedia concentrations in CWs.

Excessive nutrient loads represent a major source of water pollution globally. Best Management Practices (BMPs) are typically implemented in agricultural and urban landscapes to reduce nutrient losses at individual parcels. However, the effectiveness of BMPs in reducing nutrient loads at regional watershed domains remains an open question. This study aims to evaluate the effectiveness of parcel-based agricultural and urban BMP implementations in reducing watershed-wide phosphorus (P) and nitrogen (N) loads. The Lake Okeechobee watershed in Florida (USA), covering 10,600 km2 of mixed agricultural and urban land-use/land-cover, is used as a case study. Hydrological and nutrient transport processes were simulated using the Watershed Assessment Model (WAM). The model was calibrated and validated for river discharge and nutrients for each of the six subwatersheds, with R2 values ranging from 0.43 to 0.80 for flows and 0.24 to 0.85 for nutrients. Four what-if scenarios were simulated representing different regional BMP implementations: No BMPs implemented, current conditions (1032 km2 area influenced), maximum potential scenario (5923 km2 area influenced), and optimal BMP placement. Simulations indicated that currently implemented BMPs could be reducing P loads from 482 to 468 tons/year (3%) and N loads from 4384 to 3796 tons/year (13%). Implementation of all potential BMPs could reduce P to 307 tons/year (36%) and N to 3036 tons/year (31%). Dividing the watershed into subwatersheds illustrated that BMPs in Taylor Creek-Nubbin Slough (50% agriculture) have the potential to reduce P by 55%, and in Upper Kissimmee (23.5% urban) BMPs have the potential to reduce N by 42%. In addition, optimal BMP spatial distributions in each subwatershed could reduce P loads and N loads by 10% and 4%, respectively, providing a cost-effective solution to nutrient pollution mitigation. Our simulations suggest that other strategies would be needed to further reduce P and N, especially in subwatersheds that have low BMPs reduction potential.

Freshwater lakes worldwide suffer from eutrophication caused by excessive nutrient loads, particularly nitrogen (N) and phosphorus (P) from wastewater and runoff, affecting aquatic life and public health. Using a large (1800 km2) subtropical lake as an example (Lake Okeechobee, Florida, USA), this study aims to (1) predict key water quality parameters using machine learning (ML) algorithms based on easily measurable variables, (2) identify spatial patterns of these parameters, and (3) determine environmental drivers influencing turbidity levels. The study employs four ML algorithms-Extreme Gradient Boosting (XGB), Light Gradient-Boosting Machine (LGBM), Support Vector Regression (SVR), and Random Forests (RFs)-to predict total phosphorus (TP), total nitrogen (TN), nitrate + nitrite (NOx-N), and turbidity, via station-specific and lake-wide modeling approaches. The station-specific models uncover spatial patterns, while the lake-wide models support operational decision-making. Results indicated that lake stage (water level), water temperature, and, most notably, turbidity were the main nutrient predictors, with XGB demonstrating superior prediction performance. Spatial analysis using K-means clustering identified three distinct lake regions based on nutrient levels and turbidity. Due to its importance, SHapley Additive exPlanations (SHAP) were employed to identify and quantify environmental factors affecting turbidity. Inflows and lake stage were found as primary drivers of turbidity near lake inlets, while wind speed and air temperature affected turbidity in the middle of the lake. This research advances the understanding of lake water quality dynamics, emphasizing the importance of frequent monitoring of turbidity and its environmental drivers for enhanced management and future mitigation efforts.