Application of catastrophe theory to spatial analysis of groundwater potential in a sub-humid tropical region: a hybrid approach

Abstract

Geospatial techniques and Multi-Criteria Decision Analysis (MCDA) play a crucial role in the planning and management of land and water resources. GIS-based MCDA technique ‘Catastrophe theory’ has been recently proposed for evaluating groundwater potential. However, the major limitation of ‘Catastrophe theory’ is that only quantitative factors/thematic layers can be used for assessing groundwater potential, though qualitative factors are equally important. To overcome this inherent limitation, a novel GIS-based MCDA approach named ‘Hybrid Catastrophe’ technique is proposed in this study. The ‘Hybrid Catastrophe’ technique integrates the original ‘Catastrophe theory’ with the ‘Analytic Hierarchy Process (AHP)’ to take into account both qualitative and quantitative thematic layers for assessing groundwater potential, thereby improving the reliability and versatility of the original Catastrophe technique. The applicability of ‘Hybrid Catastrophe’ technique is demonstrated through a case study wherein 8 influential thematic layers (both quantitative and qualitative) were considered for assessing groundwater potential. The four quantitative layers were assigned weights based on the ‘Catastrophe theory’ and the remaining four qualitative layers were assigned weights based on the ‘AHP theory’. These thematic layers were integrated in GIS to delineate groundwater potential zones. The ‘Hybrid Catastrophe’ technique yields four groundwater potential zones in the study area: (i) ‘very good’ (covering 16% of the study area), (ii) ‘good’ (54%), (iii) ‘moderate’ (29%) and (iv) ‘poor’ (1%) and its accuracy was found to be 77% that is reasonably high. The proposed ‘Hybrid Catastrophe’ technique is versatile and it can be successfully applied to other parts of the world for evaluating groundwater potential at diverse spatial scales irrespective of agro-climatic, hydrologic and hydrogeologic conditions.