Semi-automated extraction of Deviation Indexes (DI) from satellite Persistent Scatterers time series: tests on sedimentary volcanism and tectonically-induced motions

Abstract. We develop a methodology based on satellite Persistent Scatterers (PS) time series and aimed to calculate two indexes which are capable to depict the deviation from a deformation model defined a priori. Through a simple mathematical approach, these indexes reproduce the visual process of identification of trend deviations that is usually performed manually by the radar-interpreter, and guide the prioritization of further interpretation for those areas recording significant variations within their motion history. First tests on semi-automated extraction of the Deviation Indexes (DI) from RADARSAT-1 PS data available over Southern Italy allowed the quantification of tectonically-induced land motions which occurred in February 2005 within the town of Naro, and also the clear recognition of the precursors to mud volcano eruptions which occurred in August 2008 in the village of St. Barbara. For these areas, the information level brought by the DI increases and adds onto that of other PS parameters, such as yearly velocity, standard deviation and coherence. Factors exerting influence on the DI are critically tackled within the discussions, together with the analysis of the potentials of these indexes for monitoring and warning activities of geohazards.