Introducing the coupled stepwise areal constraining and Mahalanobis distance: a promising MCDM-based probabilistic model for landfill site selection

• Published in: Environmental science and pollution research international Vol. 27; no. 20; pp. 24954 - 24966
• Main Authors: Davoudi Moghaddam, Davoud, Haghizadeh, Ali, Tahmasebipour, Naser, Zeinivand, Hossein
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.07.2020; Springer Nature B.V
• Subjects: Algorithms; Aquatic Pollution; Atmospheric Protection/Air Quality Control/Air Pollution; Constraining; Decision making; Earth and Environmental Science; Ecotoxicology; Environment; Environmental Chemistry; Environmental factors; Environmental Health; Environmental science; Garbage; Goodness of fit; Landfill; Landfills; Modelling; multi-criteria decision making; Multiple criterion; municipal solid waste; Pattern recognition; Pollution sources; prediction; Probabilistic models; Research Article; river water; Rivers; Site selection; Spatial dependencies; Statistical analysis; Waste disposal sites; Waste Water Technology; Water Management; Water pollution; Water Pollution Control; watersheds; Probabilistic modeling; Success rate curve; Prediction rate curve; Multi-criteria decision-making; Goodness of fit
• ISSN: 0944-1344; 1614-7499; 1614-7499
• DOI: 10.1007/s11356-020-08746-9

This study sets out to propose a new ensemble of probabilistic spatial modeling and multi-criteria decision-making comprised of stepwise areal constraining and Mahalanobis distance algorithms in order to assess areal suitability for landfilling. The Ardak watershed was selected as the study area due to encountering several cases of open garbage dumps and uncontrolled landfills which are one of the main sources of river water pollution in the upstream of the Ardak dam. The results revealed that the proposed algorithm successfully assists in inventory-irrespective probabilistic modeling of landfill siting which is mainly indebted to the role of areal constraining in providing training and validation samples for the Mahalanobis distance model. The latter also showed a robust pattern recognition results from which a discernible differentiation of the area was attained while the spatial dependencies between the environmental factors were taken into account. Mahalanobis distance also gave an outstanding performance in terms of goodness of fit (area under the success rate 89.367) and prediction power (area under the success rate 89.252). Based on a five-point scale classification scheme, about 2.7% and 2.6% of the study area, respectively, have high and very high suitability for landfilling, while the remaining area is shared between very low-to-moderate suitability classes. According to the current trail of literature regarding landfill site selection which mostly relies on mere areal filtering, a probabilistic model would give invaluable inferences regarding the pattern of suitability/susceptibility of the area of interest and causative role of the influential factors. Graphical Abstract