Bridging Traditional and Conventional Knowledge for Soil Classification in Landslide-Prone Areas Using Exploratory Factor Analysis

Sahil Sankhyan¹KALA VENKATA UDAY¹Varun Dutt^2*

• ¹Indian Institute of Technology Mandi, Mandi, Himachal Pradesh, India
• ²Computing and Electrical Engineering; Humanities and Social Sciences, Indian Institute of Technology Mandi, Mandi, Himachal Pradesh, India

Understanding the soil's propensity to movement and landslides requires understanding its categorization. Texts from ancient India include important information about the color, texture, smell, and taste of soil. Until now, soils have been categorized based on their susceptibility to movements based on contemporary geotechnical features, yet the traditional knowledge (TK) perspective on Indian soil categorization has not been thoroughly examined. This research attempts to identify characteristics derived from TK and categorize soils in landslide-prone locations according to these characteristics, while integrating these traditional qualitative indicators with quantitative geotechnical data to develop a unified classification approach. A questionnaire was created for this study to categorize soils according to TK characteristics. 32 questions about TK soil properties such as color, texture, smell, and taste were included in the questionnaire. Based on the questionnaire, landslide experts assessed 55 landslide sites. Soil samples from the surveyed landslide places were gathered, and laboratory tests were used to examine them for current geotechnical features. Separate exploratory factor analyses (EFAs) were performed on a dataset with TK characteristics, a dataset with geotechnical features, and a combined dataset including both TK characteristics and geotechnical features, enabling an integrated analysis of traditional and conventional soil parameters. Three major soil categories, earthy, coarse-grained, and fine-grained soils, were identified by EFAs, with retained factors explaining 65.17%, 68.32%, and 71.46% of the variance within the TK, geotechnical, and combined datasets, respectively. Additionally, the discipline and educational attainment of the experts significantly influenced fine-grained, coarse-grained, and earthy soil types. We discuss the ramifications of using TK-based soil categorizations in Himalayan regions that are prone to landslides.