Spatial analysis and multi-criteria decision making for regional-scale geothermal favorability map

•We use spatial distribution analysis for geothermal conceptual definition.•We use spatial association analysis geothermal conceptual definition.•We determine optimum cutoff distance for each evidential map layer.•We determine importance of each evidential map layer by calculation of weight.•We compare four prediction model for generation geothermal prospective map. Fry analysis and weights of evidence were employed to study the spatial distribution and spatial association between known occurrences of geothermal resources and publicly available geoscience data sets at regional-scale. These analyses support a regional-scale conceptual model of geological, geochemical and geophysical interaction by calculating the optimum cutoff distance and weight of each evidence feature. Spatial association analysis indicated the geochemical and geophysical data play more important roles than geological data as evidence layers to explore geothermal resources. Integration of spatial evidential data indicates how these layers interacted to form the geothermal resources. Boolean index overlay, Boolean index overlay with OR operation, multi-class index overlay and fuzzy logic prediction models were applied and compared to construct prospective maps. Prediction rate estimator showed the fuzzy logic modeling resulted in the most reliable and accurate prediction with prediction rate about 26 in the high-favorite areas.