Coastal tourism management increasingly requires monitoring approaches capable of translating large hydrochemical datasets into interpretable indicators of recreational seawater conditions. However, many conventional water quality indices were developed primarily for ecological or regulatory assessment and are not specifically adapted for tourism-oriented coastal monitoring or autonomous high-frequency observations. This study developed and applied the Tourism-Oriented Seawater Quality Index (SeaWQI-T), a simplified framework integrating turbidity, dissolved oxygen, oxidation-reduction potential (ORP), and pH into a single tourism-oriented seawater quality indicator. The framework was evaluated using 102,988 measurements collected at 10-second intervals during a summer 2025 monitoring campaign across six stations in the Gulf of Vlorë, Albania, using a multiparameter probe mounted on an unmanned surface vehicle (USV). Mean SeaWQI-T values ranged from 57.7 to 77.6, revealing a persistent offshore-nearshore spatial gradient, with the offshore station consistently classified as Good and the remaining stations classified as Fair or Moderate. Statistical and multivariate analyses confirmed significant spatial differences among stations (p < 0.001), while principal component analysis explained 84.9% of total hydrochemical variability within the first two components. Sensitivity analysis further demonstrated that the dominant spatial gradient remained stable under alternative weighting schemes. Overall, the results demonstrate that SeaWQI-T can condense high-frequency autonomous monitoring observations into a simplified but scientifically interpretable framework for tourism-oriented coastal environmental assessment and management.
Coastal tourism depends strongly on the quality of nearshore seawater experienced directly by visitors, yet high-frequency environmental monitoring often produces large datasets that are difficult to interpret for tourism-oriented management. This study applies a tourism-oriented seawater quality index (SeaWQI-T) to summarize high-frequency physical and chemical measurements collected during a summer monitoring campaign in the Gulf of Vlorë, Albania. The index combines weighted subindices of turbidity, dissolved oxygen, oxidation reduction potential (ORP), and pH to represent environmental conditions associated with water clarity and perceived cleanliness in recreational coastal waters. The analysis is based on 102,988 measurements recorded between June and August 2025 at six monitoring stations, collected at 10-second intervals using a multiparameter probe mounted on an unmanned surface vehicle. The results reveal a clear spatial gradient in seawater conditions across the gulf: the offshore station, characterized by stronger open-sea influence, consistently achieved the highest index values and was classified as Good, whereas the remaining stations remained within the Fair category throughout the monitoring period. Statistical tests and multivariate analyses confirmed strong station-dependent differences in hydrochemical variables and identified distinct environmental regimes separating nearshore and offshore waters. The limited day-to-day variability observed during the study period indicates that, under relatively stable summer conditions, the index primarily served to distinguish spatial differences among monitoring locations rather than short-term temporal changes. Overall, the results demonstrate how high-frequency coastal observations can be condensed into a concise tourism-relevant indicator while preserving the underlying environmental information for detailed analysis. The study also demonstrates how autonomous surface vehicles combined with simplified water quality indices can support tourism-relevant environmental monitoring in coastal regions where routine observation programs remain limited.
Recreational coastal waters are increasingly affected by urbanization, maritime activities, and tourism, creating a need for predictive tools that support proactive water quality management. This study proposes an explainable machine learning framework for forecasting near-future changes in the Tourism-Oriented Seawater Quality Index (SeaWQI-T) using high-frequency seawater monitoring data collected in the Gulf of Vlorë, Albania. A summer monitoring campaign (June–August 2025) produced 102,988 physicochemical observations from six monitoring stations using an unmanned surface vehicle equipped with a Horiba U53 multiparameter sonde. Following quality control and temporal aggregation, the data were used to formulate a multiclass forecasting problem (decrease, stable, or increase), and Logistic Regression, Random Forest, and Extreme Gradient Boosting (XGBoost) models were evaluated across multiple forecasting horizons. XGBoost achieved the best validation performance, while Random Forest demonstrated superior generalization on the independent test dataset and provided the most stable explainability results. SHapley Additive exPlanations (SHAP) identified SeaWQI-T dynamics, turbidity, dissolved oxygen, and oxidation–reduction potential as the most influential predictors. The proposed framework demonstrates that integrating explainable machine learning with autonomous high-frequency monitoring can provide accurate, interpretable forecasts to support intelligent coastal recreation management and sustainable tourism planning.