Extreme weather events, Climate change, Media framing, Narrative roles, Floods

Extreme weather events are increasingly treated as pivotal moments for climate engagement. Yet attribution does not occur in a vacuum. Extreme weather events invite multiple plausible policy-relevant culprits: the hazard as an extreme phenomenon, governance or infrastructure failures, land-use decisions, and anthropogenic climate change. Who is cast as villain, victim, or hero shapes both interpretation and which responses appear warranted. Yet, current research overlooks these rich narratives of accountability. In addition, media narratives and public interpretation may construct these accountability narratives differently, yet research examines media discourse, public interpretation, and behavioural responses in isolation. This leaves unresolved how extreme weather events impact engagement with climate change. We address this gap by analysing the May 2023 floods in Emilia-Romagna, Italy, combining a systematic newspaper text analysis with a representative post- event survey fielded nine months later. Using a shared narrative character-role framework (villains, victims, heroes), we compare role attributions in newspapers and residents' accounts, distinguishing flooded from non-flooded municipalities. Newspapers overwhelmingly cast the extreme weather event as the villain and rarely assign villain roles to climate change, whereas residents distribute blame across policy-relevant culprits. Flood-exposed residents assigned villain status to climate change significantly more often than non-exposed residents, even after controlling for primary information sources. This subjective attribution of climate change as villain - rather than mere flood exposure - was the dominant predictor of self-reported mitigation behaviour, providing a concrete mechanism linking experience to climate action. These findings suggest that post-disaster media narratives systematically misrepresent the accountability structures that affected populations construct, and that this misalignment may constrain the translation of extreme weather experience into sustained support for climate mitigation.