Spatiotemporal Modeling of Water Quality Trends in a Coastal Wildlife Refuge: A Statistical Approach to Ecological Risk and Resource Management

The research presents a spatiotemporal statistical modeling and analysis of variations in water quality parameters, such as turbidity, pH, dissolved oxygen (DO), salinity, and temperature etc., in the Back Bay National Wildlife Refuge, a coastal ecosystem in Virginia Beach, United States. Based on longitudinal biweekly monitoring data from designated sites, we employed quantitative methodologies including time series decomposition, correlation analysis, ANOVA, and seasonal diagnostics to analyze trends and relationships among the key parameters. Our findings revealed a significant decline in DO levels post-1997, with episodic recovery at select locations, reflecting both climatic shifts and potential local interventions. Notably, spatial analyses demonstrated substantial differences in water quality across various sites, with two sites (the Bay Area and Site D) exhibiting the highest levels of instability, indicative of localized anthropogenic stressors such as land use change or pollution discharge. Besides, the observed statistical correlations among water quality parameters reveal complex interdependencies shaped by environmental and anthropogenic influences. The identification of statistical anomalies underscores the importance of localized monitoring and adaptive regulatory strategies. The results emphasize the importance of continuous, spatially resolved monitoring and adaptive water management strategies to enhance community engagement and ensure sustainable resource stewardship, offering actionable insight for environmental planners and environmental agencies aiming to preserve aquatic ecosystem integrity and foster long-term public trust in water governance.