Evaluation of anthropogenic sinkhole susceptibility in the city of Naples (Italy) using a presence only algorithm

Giuseppe Bausilio¹, Mariagiulia Annibali Corona¹, Diego Di Martire¹, Mariano Di Napoli², Luigi Guerriero¹, Rita Tufano¹ & Domenico Calcaterra^1
1Department of Earth, Environmental and Resources Sciences, University of Naples Federico II, Naples, Italy.
²Department of Earth, Environmental and Life Sciences, University of Genoa, Genoa, Italy.
Corresponding author e-mail: giuseppe.bausilio@unina.it

Sinkholes are widespread phenomena on Earth and represent a significant hazard to the population residing in areas affected by such events. They consist of vertical depressions with a three-dimensional funnel shape, often triggered by dissolution or erosion of geological materials. To define sinkholes susceptibility for the city of Naples (Italy), a presence-only Machine Learning technique, namely Maximum Entropy, has been applied to the study area. The result obtained consists of a sinkhole susceptibility map that highlights a relatively important extension (4.9%) of the most susceptible class which consolidates the necessity of studying such phenomena in an effort to help the government officials for selecting and applying mitigation measures and adaptation plan.

sinkhole, susceptibility mapping, machine learning, MaxEnt.