Groundwater vulnerability maps derived from a time-dependent method using satellite scatterometer data

Abstract

Introducing the time variable in groundwater vulnerability assessment is an innovative approach to study the evolution of contamination by non-point sources and to forecast future trends. This requires a determination of the relationship between temporal changes in groundwater contamination and in land use. Such effort will enable breakthrough advances in mapping hazardous areas, and in assessing the efficacy of land-use planning for groundwater protection. Through a Bayesian spatial statistical approach, time-dependent vulnerability maps are derived by using hydrogeological variables together with three different time-dependent datasets: population density, high-resolution urban survey, and satellite QuikSCAT (QSCAT) data processed with the innovative dense sampling method (DSM). This approach is demonstrated extensively over the Po Plain in Lombardy region (northern Italy). Calibrated and validated maps show physically consistent relations between the hydrogeological variables and nitrate trends. The results indicate that changes of urban nitrate sources are strongly related to groundwater deterioration. Among the different datasets, QSCAT-DSM is proven to be the most efficient dataset to represent urban nitrate sources of contamination, with major advantages: a worldwide coverage, a continuous decadal data collection, and an adequate resolution without spatial gaps. This study presents a successful approach that, for the first time, allows the inclusion of the time dimension in groundwater vulnerability assessment by using innovative satellite remote sensing data for quantitative statistical analyses of groundwater quality changes.