Comparing stormwater quality and watershed typologies across the United States: A machine learning approach

Salty chemical cocktails as water quality signatures: Longitudinal trends and breakpoints along different U.S. streams

Along urban streams and rivers, various processes, including road salt application, sewage leaks, and weathering of the built environment, contribute to novel chemical cocktails made up of metals, salts, nutrients, and organic matter. In order to track the impacts of urbanization and management strategies on water quality, we conducted longitudinal stream synoptic (LSS) monitoring in nine watersheds in five major metropolitan areas of the U.S. During each LSS monitoring survey, 10-53 sites were sampled along the flowpath of streams as they flowed along rural to urban gradients. Results demonstrated that major ions derived from salts (Ca2+, Mg2+, Na+, and K+) and correlated elements (e.g. Sr2+, N, Cu) formed 'salty chemical cocktails' that increased along rural to urban flowpaths. Salty chemical cocktails explained 46.1% of the overall variability in geochemistry among streams and showed distinct typologies, trends, and transitions along flowpaths through metropolitan regions. Multiple linear regression predicted 62.9% of the variance in the salty chemical cocktails using the six following significant drivers (p < 0.05): percent urban land, wastewater treatment plant discharge, mean annual precipitation, percent silicic residual material, percent volcanic material, and percent carbonate residual material. Mean annual precipitation and percent urban area were the most important in the regression, explaining 29.6% and 13.0% of the variance. Different pollution sources (wastewater, road salt, urban runoff) in streams were tracked downstream based on salty chemical cocktails. Streams flowing through stream-floodplain restoration projects and conservation areas with extensive riparian forest buffers did not show longitudinal increases in salty chemical cocktails, suggesting that there could be attenuation via conservation and restoration. Salinization represents a common urban water quality signature and longitudinal patterns of distinct chemical cocktails and ionic mixtures have the potential to track the sources, fate, and transport of different point and nonpoint pollution sources along streams across different regions.

Parameterization of nutrients and sediment build-up/wash-off processes for simulating stormwater quality from specific land uses

Urbanization changes land cover through the expansion of impermeable surfaces, leading to a significant rise in runoff, sediment, and nutrient loading. The quality of stormwater is related to land use and is highly variable. Currently, stormwater is predominantly described through watershed models that rely minimally, if at all, on field monitoring data. The simple event mean concentration (EMC) wash-off approach by land use is a common method for estimating urban runoff loads. However, a major drawback of the EMC approach is it assumes concentration remains constant across events for a specific land use. Build-up/wash-off equations have been formulated to consider variations in concentration between events. However, several equation parameters are challenging to estimate, making them difficult to use. We conducted a monitoring and modeling study and investigated the impact of land use on stormwater quantity and quality and optimized and investigated the build-up/wash-off parameters for three homogenous urban land uses to estimate nutrients (nitrogen and phosphorus) and sediment loads. Stormwater from commercial, medium-density residential, and transportation land uses was sampled using automatic samplers during storm events, and water quality was characterized for a variety of them for 14 months. Analysis of stormwater samples included assessments for total nitrogen, total phosphorus, and total suspended solids. Results showed that medium-density residential land use had the highest median total nitrogen and total phosphorus event mean concentrations and commercial had the highest median total suspended solids EMCs. Water quality parameters (or build-up/wash-off parameters) exhibited significant variation between land uses, confirming that land use is a key determinant of stormwater quality. The median particle size for each land use was less than 150 μm, indicating that the most common particle size in stormwater was a very fine sand or smaller. This small size should be considered by stakeholders in the design of stormwater treatment systems.

Urban stormwater quality: A review of methods for continuous field monitoring

Urban stormwater is contaminated by a wide range of substances whose concentrations vary greatly between locations, as well as between and during rain events. This literature review evaluates advantages and limitations of current methods for using continuous water quality monitoring for stormwater characterization and control. High-temporal-resolution measurements have been used to improve the understanding of stormwater quality dynamics and pollutant pathways, facilitate the performance evaluation of stormwater control measures and improve operation of the urban drainage system with real-time control. However, most sensors used to study stormwater were developed for either centralized water treatment or natural water contexts and adaptation is necessary. At present, the primary application of interest in stormwater - characterization of pollutant concentrations - can only be achieved through the use of indirect measurements with site-specific relationships of pollutants to basic physical-chemical parameters. In addition, various problems arise in the field context, associated with intermittent or variable flow rates, the accumulation of debris and sediment, adverse conditions for electrical equipment and human factors. Obtaining reliable continuous stormwater quality data requires the adoption of best practices, including the calibration and regular maintenance of sensors, verification of data and accounting for the considerable uncertainties in data; however, the literature review showed that improvement is needed among the scientific community in implementing and documenting these practices.

Machine learning-guided determination of Acinetobacter density in waterbodies receiving municipal and hospital wastewater effluents

A smart artificial intelligent system (SAIS) for Acinetobacter density (AD) enumeration in waterbodies represents an invaluable strategy for avoidance of repetitive, laborious, and time-consuming routines associated with its determination. This study aimed to predict AD in waterbodies using machine learning (ML). AD and physicochemical variables (PVs) data from three rivers monitored via standard protocols in a year-long study were fitted to 18 ML algorithms. The models' performance was assayed using regression metrics. The average pH, EC, TDS, salinity, temperature, TSS, TBS, DO, BOD, and AD was 7.76 ± 0.02, 218.66 ± 4.76 µS/cm, 110.53 ± 2.36 mg/L, 0.10 ± 0.00 PSU, 17.29 ± 0.21 °C, 80.17 ± 5.09 mg/L, 87.51 ± 5.41 NTU, 8.82 ± 0.04 mg/L, 4.00 ± 0.10 mg/L, and 3.19 ± 0.03 log CFU/100 mL respectively. While the contributions of PVs differed in values, AD predicted value by XGB [3.1792 (1.1040-4.5828)] and Cubist [3.1736 (1.1012-4.5300)] outshined other algorithms. Also, XGB (MSE = 0.0059, RMSE = 0.0770; R² = 0.9912; MAD = 0.0440) and Cubist (MSE = 0.0117, RMSE = 0.1081, R² = 0.9827; MAD = 0.0437) ranked first and second respectively, in predicting AD. Temperature was the most important feature in predicting AD and ranked first by 10/18 ML-algorithms accounting for 43.00-83.30% mean dropout RMSE loss after 1000 permutations. The two models' partial dependence and residual diagnostics sensitivity revealed their efficient AD prognosticating accuracies in waterbodies. In conclusion, a fully developed XGB/Cubist/XGB-Cubist ensemble/web SAIS app for AD monitoring in waterbodies could be deployed to shorten turnaround time in deciding microbiological quality of waterbodies for irrigation and other purposes.

Using Random Forest, a machine learning approach to predict nitrogen, phosphorus, and sediment event mean concentrations in urban runoff

Estimating pollutant loads from developed watersheds is vitally important to reduce nonpoint source pollution from urban areas, as a key tool in meeting water quality goals is the implementation of Stormwater Control Measures (SCMs). SCMs are selected and sized based on influent pollutant loads. A common method used to estimate pollutant loads in urban runoff is the Event Mean Concentration (EMC) method. In this study, we develop and apply data-driven models using Random Forest (RF), a machine learning approach, to predict Total Nitrogen (TN), Total Phosphorus (TP), Total Suspended Solids (TSS), and Ortho-Phosphorus (Ortho-P) EMCs in urban runoff. The parameters considered in this study were climatological characteristics (i.e., Antecedent Dry Period or ADP, Precipitation Depth or P, Duration or D, and Intensity or I) and catchment characteristics including land use-related parameters including Imperviousness or Imp, Saturated Hydraulic Conductivity or K_sat, and Available Water Capacity or AWC), and site-specific parameters including Slope (S), and Catchment Size (A). Stormwater quality data for this study were obtained from the National Stormwater Quality Database (NSQD), which is the largest repository of stormwater quality data in the U.S. Results demonstrate that land use-related characteristics (i.e., Imp, K_sat, and AWC) were the most effective variables for predicting all EMCs. For TP, TSS, and Ortho-P, site-specific characteristics (S and A) had a greater effect than climatological characteristics (i.e., ADP, P, D, and I). However, for TN, climatological characteristics had a greater effect than site-specific characteristics (S and A). In addition, for TN, TP, and TSS, precipitation characteristics (P, D, and I) were found to be more effective parameters for estimating EMCs than ADP. This study highlights the most influential parameters affecting EMCs which can be used by stakeholders and SCMs designers to improve estimates of nutrients and sediment EMCs. The selection and design of the highest performing SCMs is essential in achieving effective treatment of stormwater, attaining water quality goals, and protecting downstream waterbodies.