Objectives Natural disasters have a significant impact on mental health. Data collected from the population offer a unique opportunity for post-disaster monitoring to help identify psychological support needs. The aim of this study is: 1) to identify psychopathological aspects for the county of Prêcheur at risk from lahars (volcanic lava), and 2) to phenotype psychopathological aspects from data collected from the population. Method We applied an artificial intelligence (AI) assisted psycho-phenotyping method to data from 40 people over a 20-month period, to extract psychopathological and psychiatric aspects linked to traumatic natural hazards. These were then compared with the results of psychometric tests measuring overall mental health and post-traumatic stress. Results Rumination and negativation were among the most important psychopathological aspects identified. In addition, we noted the presence of re-experiencing and avoidance as core psychiatric dimensions over time. Among these, cognitive avoidance and emotional avoidance were found and seem to have emerged after the disaster. Conclusion We have proposed a new syndromic surveillance approach for mental health based on digital data that can support conventional approaches by providing additional useful information in the context of a disaster. Further studies are needed to better control bias, identify associations with valid instruments, and explore computational methods for continuous adjustment of the AI-analysis model.