A method for the remote sensing identification of uncontrolled landfills: formulation and validation

Abstract

The identification of uncontrolled landfills is a central environmental problem in all developed and developing countries, where several illegal waste deposits exist as a result of rapid industrial growth over the past century. Remote sensing can potentially provide crucial information for the identification of contaminated sites, but surprisingly there is a marked lack of rigorously validated approaches. In this paper we introduce and validate a method that uses remotely sensed information and a geographic information system (GIS) to identify unknown landfills over large areas. The method is applied to a study area located in NE Italy (part of the Venice lagoon watershed) using IKONOS satellite data. Soil contamination effects on the radiometric properties of vegetation, calibrated using spectral signatures of stressed vegetation from known illegal landfill sites, were used to define numerous candidate sites that are most likely to host waste materials. Distributed geographical information, such as the position of the road network, the population density, and historical aerial photographs, have then been used to select the most likely contaminated sites among the candidates identified through remote sensing. The importance of the integration of GIS and remote sensing is highlighted and represents a key instrument for environmental management and for the spatially distributed characterization of possible uncontrolled landfill sites.