Quantification of the Ecological Value of Railroad Development Areas Using Logistic Regression Analysis

Abstract

According to the national railway network construction plan, Investment in railways has increased due to the need for environmentally friendly transportation, and the rail network is expanding throughout South Korea. Railway projects should be evaluated using strategic environmental impact assessments. In the “Guidelines for the Construction of Environment-friendly Railways”, seven priority headings that must be considered for railway projects are described. This guide notes that qualitative evaluation must be conducted during the survey process to reasonably predict impacts on the environment. However, quantitative evaluation with specific indicator values may also be necessary. In this study, independence analysis and logistic regression analysis were used to quantitatively evaluate railway environmental and ecological indicators. The results were used to develop a regression model reflecting seven indicators; biodiversity class, ecosystem type, vegetation conservation class, tree age class, ecological naturalness, presence of river ecosystems, and fragmented patch size. The fitness regression model showed 90.3% classification accuracy and the receiver operating curve (ROC) model fit was 88.6%. An environmental quality assessment map was prepared by classifying areas of environmental quality according to five grades. This is the first model for environmental and ecological evaluation of railway projects. Evaluation using the map showed that the railroad passes through areas with lower protection values compared to the results obtained using the national environmental evaluation map. Kappa analysis showed a low level of agreement between the two maps (kappa coefficient = 0.212). The results of this study can be applied to railway development project sites and may help to identify the best sites for the development of an environmentally friendly railway system.