Realising resilience requires long-term, strategic transformation. However, this is often forestalled by immediate, often reactive, post-disaster needs. Here, we use Natural Language Processing to compare media coverage of Cyclone Gabrielle, a significant storm affecting Aotearoa New Zealand (A-NZ) from February 12–16, 2023, with A-NZ based research on the subject. The aim is to gain new insight into the ways in which disasters are framed post-event, and over longer time scales to compare immediate and consolidated concerns and insights. Results reveal contrasting frames: local and international media frames cyclone response in relation to tasks assigned to central and local governments, global climate change dynamics, emissions reductions, infrastructure, and overall economic development. Research frames resilience in terms of community and regional adaptation, classification, measurement, and assessment. Post-disaster media coverage tends to frame resilience around readiness and response rather than emphasizing pathways to future reduction and recovery, which are present in research but lack a focus on government, governance, and pathways. Both media and research portray community effects in positive terms. With changing frequency and severity of climate related extremes, reframing coverage and proactively shifting the discourse towards enabling transformations for climate-adapted futures may be necessary to better leverage the post-disaster period for change.