Multi-criteria decision analysis in environmental sciences: Ten years of applications and trends

Abstract

Decision-making in environmental projects requires consideration of trade-offs between socio-political, environmental, and economic impacts and is often complicated by various stakeholder views. Multi-criteria decision analysis (MCDA) emerged as a formal methodology to face available technical information and stakeholder values to support decisions in many fields and can be especially valuable in environmental decision making. This study reviews environmental applications of MCDA. Over 300 papers published between 2000 and 2009 reporting MCDA applications in the environmental field were identified through a series of queries in the Web of Science database. The papers were classified by their environmental application area, decision or intervention type. In addition, the papers were also classified by the MCDA methods used in the analysis (analytic hierarchy process, multi-attribute utility theory, and outranking). The results suggest that there is a significant growth in environmental applications of MCDA over the last decade across all environmental application areas. Multiple MCDA tools have been successfully used for environmental applications. Even though the use of the specific methods and tools varies in different application areas and geographic regions, our review of a few papers where several methods were used in parallel with the same problem indicates that recommended course of action does not vary significantly with the method applied.

Introduction

Environmental decisions are often complex and draw upon multidisciplinary knowledge bases which incorporate natural, physical social sciences, politics, and ethics. Environmental decision makers rely on many experimental tests, computational models, and tools to assess human health and ecological risks associated with environmental stressors and the impact of remedial and abatement strategies on risk reduction. However, applying these tools is becoming increasingly difficult for three reasons. First, there are many emerging risks (e.g., climate change, nanotechnology, etc.) for which information is not available and decisions should be made under significant uncertainty. Second, for many traditional stressors and situations multiple lines of evidence regarding the same measure (e.g., risk) are available, but they may point to different management alternatives. Finally, stakeholders, who may have vested interest in specific courses of action, are gaining increased access to all available information and, given the data uncertainty, can justify often opposing courses of action. As such, integration the heterogeneous and uncertain information demands a systematic and understandable framework to organize the technical information and requires expert judgment.

Multi-criteria decision analysis (MCDA) provides a systematic methodology to combine these inputs with cost/benefit information and stakeholder views to rank project alternatives. MCDA is used to discover and quantify decision maker and stakeholder considerations about various (mostly) non-monetary factors in order to compare alternative courses of action. There are numerous approaches that all fall under the umbrella of MCDA, each involving different protocols for eliciting inputs, structures to represent them, algorithms to combine them, and processes to interpret and use formal results in actual advising or decision making contexts. Linkov et al., 2004, Linkov et al., 2006 reviewed the application of MCDA in environmental management focusing on management of contaminated sites published in 1992–2002. MCDA techniques have been applied to optimize policy selection in the remediation of contaminated sites, the reduction of contaminants entering aquatic ecosystems, the optimization of water and coastal resources, and the management of other resources. In some of these studies, the researchers have explicitly taken into account the opinions of local community groups and other stakeholders through focus groups, surveys, and other techniques and formally integrated these opinions into the decision process. Many papers concluded that the application of MCDA methods provides a significant improvement in the decision process and public acceptance of the suggested remedial or abatement policy.

The last decade brought not only an increased interest in the application of formalized decision-analytical tools, but also better structured and complete databases. The goal of this paper is to review the recent literature and identify current trends in MCDA applications to environmental management, and to discuss the possible reasons that the various methods are gaining attention in different applications. In 2003, our search strategy was to find MCDA applications by all means, including searching gray literature. The current challenge was to limit the number of papers for the analysis to a manageable size but still objectively represent the state of the applications. Similar to Linkov et al. (2004), we consider three main MCDA approaches: MAUT, Outranking (including PROMETHEE and ELECTRE), and AHP. These approaches share common mathematical elements, i.e., values for alternatives are assigned for a number of dimensions, and then multiplied by weights and finally combined to produce a total score. The approaches differ significantly in the details of how values are assigned and combined, meaning that the processes have different information – and knowledge – requirements and the calculated scores have different mathematical properties and thus slightly different meanings. Practitioners often view one of the various approaches as most appropriate due to the priority they place on its relative strengths and weaknesses. (e.g., Figueira et al., 2005, Belton and Stewart, 2002).