Climate change makes the future more uncertain, but robust modelling can offer a pathway to clarity. While no one can predict exactly how flood risk will evolve, we can examine a range of physically plausible futures. JBA’s updated Australia Inland Flood Model does just that – offering a scientifically grounded lens through which insurers, reinsurers, and governments can stress test their exposure. Here, we describe how this updated model uses physical climate storylines to explore changes in flood hazard and losses across Australia.
A changing climate and growing need
Flooding is already Australia’s costliest natural hazard. Recent events have prompted a rethink in flood insurance. At the start of the 21st century, many Australian home insurance policies excluded or only partially covered flooding. Major events such as the 2011 Brisbane floods increased demand for full and consistent coverage. A government review in 2012 led to initiatives that made flood insurance more accessible, but affordability challenges remain in high-risk areas.
With a changing climate, the impacts of flooding are expected to grow. Insurers face tough questions: Where will floods occur more often? Will they be more severe? How will losses change across an already marginal and varied environment?
To answer these questions, models must move beyond historical data alone. They need to reflect the range of potential climate futures and do so in a way that is accessible and useful to decision-makers. JBA’s latest catastrophe model for Australia, enabled on the Oasis platform, provides that capability.
A new approach: Physical climate storylines
At the heart of this update are 15 climate change event sets that simulate flood hazard under warming scenarios from 2 °C to 4 °C, in 0.5 °C increments. These event sets are built on three physical climate storylines, derived from structurally different global climate models.
These physical storylines describe how rainfall and river flow extremes might evolve in a warming world – without incorporating changes in exposure, vulnerability, or adaptive capacity. They are designed to isolate the physical hazard component of future flood risk. Furthermore, they are neither forecasts nor probabilities, but internally consistent representations of how the climate system could respond to global warming.
As shown by Figure 1, the three physical climate storylines lead to markedly different patterns in future rainfall and river flow extremes across Australia. This diversity allows users to explore how their risk profile responds under contrasting plausible futures.
For example, the climate-driven trends in extreme rainfall and river flows span increases and decreases on parts of the East Coast. These divergent outcomes reflect real uncertainty in the direction and magnitude of hazard change – particularly in major population centres. This diversity allows users to explore how their risk profile might shift under contrasting but credible futures.
Summarising a complex future
While Figure 1 highlights differences in the hazard across physical storylines, Figure 2 shows how these changes translate into economic risk. For the different physical storylines, it presents trends in average annual loss (AAL) from the present day (baseline) and across three warming levels (2°C, 3°C, 4°C) for both surface water and river flood, and their combined effect. The values are normalised to a baseline of 100 to make trends easier to compare.
Notably, surface water flood losses increase steadily with warming in all physical storylines, indicating a consistent response at the portfolio level, even though this does not imply certainty about the precise scale of change. In contrast, river flood losses vary more significantly between the physical storylines, depending on how precipitation and evaporation trends interact regionally. For the test portfolio used in this assessment, river flood accounts for roughly twice the loss of surface water flood, meaning that these differences materially affect the total combined risk borne by insurers.
The result: total losses under warming vary substantially across the storylines. This underscores the value of the multi-model approach, not just in acknowledging uncertainty, but in capturing a range of potential financial exposure that insurers and reinsurers must consider.
Why a multi-model approach matters
Future flood risk is not just uncertain – it is deeply dependent on complex physical processes that respond differently to global warming. Relying on a single projection of risks obscures this complexity. At the same time a naive averaging across different GCMs ignores the fact that they are not independent samples of the climate system.
By incorporating three physical climate storylines, our model highlights regions where both the direction and magnitude of change are uncertain. It also reveals areas where flood risk increases consistently across storylines, providing greater confidence in the signal of change.
This multi-model approach supports transparent scenario analysis aligned with expectations from the Australian Prudential Regulation Authority (APRA) and other regulators, helping users communicate risk with nuance and rigour.
Insights for practice
The model enables users to explore how changes in hazard translate into changes in risk and loss. This supports underwriting and pricing strategies that incorporate climate-adjusted hazard, as well as climate scenario testing for exposure management. It also informs strategic reinsurance placement and capital allocation.
These insights feed directly into regulatory compliance and disclosure efforts – both under APRA and international frameworks – by providing traceable and defensible evidence for decision-making.
Additionally, the model can be used by governments to explore sensible locations for building new defences or improving existing ones, and to support "build-back" programmes akin to Flood Re in the UK.
Conclusion: Planning for a range of futures
Climate models cannot predict the future – but they can help us understand the boundaries of what’s possible. By embedding physical climate storylines into the JBA Australia Inland Flood Model, we are equipping users with the tools to explore a range of plausible futures.
It’s not about certainty. It’s about clarity amid uncertainty.