The monitoring of surface-water quality followed by water-quality modeling and analysis are essential for generating effective strategies in surface-water-resource management. However, worldwide, particularly in developing countries, water-quality studies are limited due to the lack of a complete and reliable dataset of surface-water-quality variables. In this context, several statistical and machine-learning models were assessed for imputing water-quality data at six monitoring stations located in the Santa Lucía Chico river (Uruguay), a mixed lotic and lentic river system. The challenge of this study is represented by the high percentage of missing data (between 50% and 70%) and the high temporal and spatial variability that characterizes the water-quality variables. The competing algorithms implement univariate and multivariate imputation methods (inverse distance weighting (IDW), Random Forest Regressor (RFR), Ridge (R), Bayesian Ridge (BR), AdaBoost (AB), Hubber Regressor (HR), Support Vector Regressor (SVR) and K-nearest neighbors Regressor (KNNR)). According to the results, more than 76% of the imputation outcomes are considered “satisfactory” (NSE > 0.45). The imputation performance shows better results at the monitoring stations located inside the reservoir than those positioned along the mainstream. IDW was the model with the best imputation results, followed by RFR, HR and SVR. The approach proposed in this study is expected to aid water-resource researchers and managers in augmenting water-quality datasets and overcoming the missing data issue to increase the number of future studies related to the water-quality matter.

The monitoring of surface-water quality followed by water-quality modeling and analysis is essential for generating effective strategies in water-resource management. However, worldwide, particularly in developing countries, water-quality studies are limited due to the lack of a complete and reliable dataset of surface-water-quality variables. In this context, several statistical and machine-learning models were assessed for imputing water-quality data at six monitoring stations located in the Santa Lucía Chico river (Uruguay), a mixed lotic and lentic river system. The challenge of this study is represented by the high percentage of missing data (between 50% and 70%) and the high temporal and spatial variability that characterizes the water-quality variables. The competing algorithms implemented belonged to both univariate and multivariate imputation methods (inverse distance weighting (IDW), Random Forest Regressor (RFR), Ridge (R), Bayesian Ridge (BR), AdaBoost (AB), Hubber Regressor (HR), Support Vector Regressor (SVR), and K-nearest neighbors Regressor (KNNR)). According to the results, more than 76% of the imputation outcomes are considered satisfactory (NSE > 0.45). The imputation performance shows better results at the monitoring stations located inside the reservoir than the ones positioned along the mainstream. IDW was the most chosen model for data imputation.

Changes in land use/land cover (LULC) directly or indirectly affect water quality in watercourses and impoundments. Sustainable management strategies aimed to enhance ecosystem health and community well-being require an accurate water-quality evaluation. This study looks into the correlation between temporal changes in LULC, represented by selected landscape variables (land cover area and proportion, patch density, Euclidean nearest-neighbor distance, mean shape index, and Shannon index), and water quality variables (nitrate, total phosphorus, and total suspended solids) at catchment scale. To compare the watershed-size influence, this analysis was performed at two different spatial scales represented by two Uruguayan basins of different sizes, San Salvador (3,118 km2) and Del Tala (160 km2). Partial Least Squares and Random Forest unsupervised machine-learning models were employed for this analysis. By exploiting a non-model-biased method based on game theory (SHAP), the LULC characteristics were quantified and ranked based on their level of importance in the water-quality evaluation. The main outcomes of this study proved that patch density is one of the most influencing metrics in both watersheds and for both models. Agricultural land use is the most critical one at both catchments and agricultural with a forage crop land uses are the most important ones for both algorithms. Furthermore, it is possible to state that the adopted techniques are valuable tools that can provide an adequate overview of the water‐quality behavior in space and time and the correlations between water-quality variables and LULC.