Ecological assessment of landfills with multivariate analysis — A feasibility study

Methods of multivariate data analysis are applied for ecological monitoring of the landfills. Traditional waste management approach often fails to reveal the specific areas within a landfill due to the complexity of geometrical configuration and variety of degradation processes. The present investigation is based on three different man-caused formations in the Samara region: illegal dump Bezenchuk, poorly run landfill Otradny, and a modern well-run landfill Kinel. Refuse samples are obtained by means of step-by-step drilling. Each sample is characterized by several measured variables such as depth, temperature, humidity, ash content, volumetric weight, and pH. The samples also have a variety of evaluated properties such as age, belonging to a stratum or lens, topsoil, etc., determined by the traditional methods of landfill exploration. Both measured and evaluated data are subjected to PCA and PLS analysis. Chemometric methods give possibility to explore the structure of a landfill, to reveal their specific areas, and to predict evaluated properties using measured data only.