(a) Analytical framework focusing on the dynamics produced by the (local) interplay of natural hazards and vulnerabilities under (global) environmental change and socioeconomic trends. (b) Flood risk example at the local scale: explanatory model based on Di Baldassarre et al. (2013) emphasizing hypothetical feedbacks between five key variables that are assumed to influence each other and change gradually overtime (thin arrows), while being abruptly altered by the sudden occurrence of flooding (thick arrows). Dashed arrows indicate control mechanisms: wealth influences how flood exposure can potentially change overtime and also determines whether levees can be built or not, while levees reduce the frequency of flooding. (c) Hypothetical wealth trajectories in relation to disaster occurrences: bouncing back, forward or collapsing after a major disaster. (d) Ranges of availability of systematic time series across decades in the study of flood risk dynamics. The fuzzy classification highlights the limited availability of data to carry out empirical studies about socionatural interactions.