A grey relational model for soil erosion vulnerability assessment in subwatershed of lesser Himalayan region

Abstract

Assessment of the vulnerability of a watershed to soil erosion is essential for developing management strategies for a watershed. Different approaches are in practice to assess soil erosion vulnerability but lack of data and accessibility of the watersheds in rugged mountainous terrain puts restrictions on using quantitative techniques to assess vulnerability. Hence, multi criteria based qualitative methods using soil, morphological parameters and land use as criteria are being used. The observations from these components generally differ in measurement scale and sample size. The sample size may also not be large enough for using standard quantitative techniques for ranking the subwatersheds. This study addresses the limitation of heterogeneous scales of measurement and small sample sizes to assess the vulnerability using Grey Relational Analysis (GRA). So far, GRA has been used limitedly in some studies to determine few forest functions. There is an absolute lack of information about applying this method in many domains of natural resource management, including watersheds. In the present study, GRA is applied to assess the vulnerability of a watershed in lesser Himalayas to soil erosion by assessing that of subwatersheds constituting the watershed using information derived from remote sensing and field data. Parameters pertaining to Geomorphology, soil erodibility and land use were identified and classified into benefit and defect type based on their contribution towards erosion. Three different components comprising of 21 variables and 21 subwatersheds were taken, and a 21 × 21 matrix was constructed, and data was normalised to make the scale uniform. Out of the total 21 subwatersheds, the subwatersheds most prone to soil erosion were because of their morphometric parameters like low elongation ratio and ruggedness number. The least prone was due to its high value morphometric parameters despite a low presence of forest land use and steep slope. The results indicate that forest area (69.48%) alone is not sufficient for a subwatershed being less vulnerable to soil erosion neither high slope makes a subwatershed highly vulnerable. The suggested method helps to analyse the combined influence of the various components on erosion vulnerability and ranks the different subwatersheds according to their vulnerability towards soil erosion. The study demonstrates that GRA can effectively rank the constituent subwatersheds of a watershed according to their vulnerability to soil erosion, thereby aiding the decision process for watershed conservation and formulation of management strategies.