This paper provides the outcomes of a work consisting in modeling and learning some earthquakes data collected during the Mayotte seismovolcanic crisis of 2018-2021. We highlight the performance of some process data models in order to illustrate the spatial and temporal dynamic. Unsupervised clustering method, spatial pattern analysis, spatial density estimation through spatial marked point process; time series and spatio-temporal models are efficient tools that we studied in this paper to look for the spatial and temporal variation of such spatial data mainly driven by the detected underwater volcano around Mayotte called Fani Maoré . The dynamic of the magnitude and depth events of the Fani Maoré with the use of the above mentionned models seems to perform the data. We present a discussion thoughout the presentation of the obtained results together with the limit of this study and some forthcoming projects and modeling developments.