The influence of location on water quality perceptions across a geographic and socioeconomic gradient in Appalachia

Understanding perceptions of water quality held by residents is critical to address gaps in public awareness and knowledge and may provide insight into what defines communities that are more/less resilient to changing water conditions locally. We sought to identify spatial patterns of water quality perceptions gathered in a survey of SouthernWest Virginia (WV) residents during spring/summer 2018. Using over 500 survey responses across 15 counties, we calculated spatial autocorrelation metrics and modeled the relationships between overall water quality perceptions and county-level socioeconomic endpoints, such as poverty rate, per capita income, and education level. We identified significant differences across counties labeled as socioeconomically "transitional", "at-risk", and "distressed", as it pertained to responses for water quality perceptions, education level, and income level. We also found significant positive relationships between overall water quality perceptions, elevation, and income level. We calculated an empirical semivariogram and fit an exponential model to explain a significant autocorrelation pattern within a range of 104.2 km. Using that semivariance function, we created a kriging interpolation surface across the study area to identify significant clusters of water quality perceptions. This work highlights the influence of location on water quality perceptions within SouthernWest Virginia, but the analytical framework should be considered in further research, when samples are spread across large areas with varying socioeconomics.