How much could sea level rise cost Australia?

Under RCP4.5, a moderate emissions pathway, the study projects that 267,500 properties and two million hectares of land will be affected by 2100, with present-value damages of $855.1 billion ($274.3 billion from property and $580.7 billion from land use). Under RCP8.5, a high-emissions pathway, that rises to 719,600 properties and 3.7 million hectares, with damages reaching $1.91 trillion ($705.7 billion from property and $1.20 trillion from land use). The researchers express average annual damages as a proportion of GDP. Across their sensitivity cases, these annual damages are equivalent to around 1.05–1.39% of Australian GDP under RCP4.5 and 2.26–3.02% under RCP8.5.

These are substantial numbers, but the national totals are only part of the value of the study. The modelling distinguishes consequences  by state, SA2 region and land use category, showing how the same national-scale climate process produces very different financial consequences depending on local elevation, land use and property values.

Moving beyond exposure mapping

Much coastal risk analysis available for Australia has focused on physical exposure: which areas, properties, or infrastructure intersect a future inundation extent and how that changes under different climate scenarios. State coastal hazard maps and public tools such as Coastal Risk Australia do this well. What has been missing is a consistent valuation layer on top of that physical picture, one that converts inundated hectares into dollar losses a council can weigh against its other capital and risk planning.

Kompas and colleagues add that layer. Their model combines a bathtub-style inundation approach, using Kirezci et al.'s global tide-surge projections and Geoscience Australia's digital elevation data, with property and land use data drawn from state government sources and real estate listings across the country. 

The 88 underlying land-use classes are aggregated into ten categories: residential; commercial; industrial and manufacturing; quarries; agriculture; infrastructure; education and public facilities; parks and outdoor areas; reserves; and unidentified private-sector land.

The study's economic analysis is primarily a direct-loss valuation. It estimates the value of affected property and land rather than modelling wider economic disruption such as business interruption, infrastructure service loss, supply-chain effects or recovery. Its contribution is therefore to translate physical exposure into spatially differentiated financial consequence, rather than to estimate the full economic consequences of sea-level rise.

This distinction matters when interpreting one of the study's headline findings: approximately 63% of estimated financial losses are classified as land-value losses, compared with 37% from property. That 63% should not be interpreted only as agricultural or infrastructure loss. It represents affected land value across the study's broad set of land-use categories.

The geography of consequence matters

The study finds substantial variation both between and within states. Under the base case and RCP8.5, Queensland, New South Wales and Western Australia carry the largest shares of total economic loss, at 27.5, 22.7 and 21.3 percent respectively. Victoria, South Australia, the Northern Territory and Tasmania follow.

Property damage is only part of the picture. Land use losses, covering agriculture, infrastructure and reserves, make up 63 percent of total economic damage nationally, a useful corrective for anyone who assumes coastal risk is mainly a property story. Western Australia, New South Wales and Queensland account for the largest shares of land use loss, driven largely by agricultural land in low-lying coastal catchments. The Northern Territory stands out on a different measure. As a share of gross state product, it is the most exposed jurisdiction in the country, with annual losses equivalent to 3.7 percent of GSP under RCP4.5 and 5.9 percent under RCP8.5, even though its absolute dollar figures are smaller than the larger states.

There is also a difference between where the greatest area is exposed and where the greatest financial loss occurs. Differences in land use, property concentration and value alter the consequence associated with a given area of inundation. This is an important consequence of moving beyond exposure. Physical extent alone is a poor proxy for the significance of an event.

The estimates depend on some important assumptions

The headline values should be interpreted as scenario-based estimates rather than predictions. Several assumptions materially affect the results.

First, the inundation modelling uses a bathtub approach. This provides nationally consistent coverage but does not reproduce all of the hydrodynamic processes that determine local coastal flooding.

Second, Australia does not yet have nationally consistent depth-damage functions. Some damage and valuation relationships are therefore derived from more detailed studies in Victoria, New South Wales and Queensland and applied more broadly.

Third, the present-value results are sensitive to the discount rate. The study uses 5% for its base case and tests rates of 3% and 7%. The resulting differences are substantial.

The assessment also excludes or incompletely represents several potentially important consequences, including environmental assets, wetlands, cultural heritage, coastal erosion and parts of the infrastructure system.

These limitations define the scope of the numbers. The study provides a consistent national estimate of direct financial loss, rather than a complete estimate of the social and economic consequences of coastal flooding.

What does this mean for decision-making?

The study's main lesson is methodological. Hazard maps tell us where something may happen. Decisions generally require another step: estimating what happens to the things we care about as a result.

Kompas et al. demonstrate the value of moving from exposure to direct financial consequence. That allows governments and other decision makers to compare locations, land uses and scenarios using a consistent measure, and to identify areas where the potential consequences warrant closer investigation.

For many local decisions, another step is then required: understanding how physical damage affects infrastructure services, access, business activity, recovery and different population groups.

A council deciding where development should occur, an infrastructure provider prioritising adaptation investment, or a government evaluating retreat or protection options may need to understand consequences such as:

• whether communities lose access to essential services

• whether infrastructure failures create cascading effects across other networks

• how long disruption persists and how recovery changes over time

• which businesses, households or population groups bear the consequences

• how different interventions change those outcomes.

That is the decision problem Resilience Explorer® is designed around. It brings hazard information together with assets, infrastructure systems, communities and consequence modelling so that organisations can examine how hazards propagate through the systems they are responsible for and what that means for particular decisions.

The Australian study provides a useful national-scale demonstration of the first part of this process: exposure becomes more decision-relevant when it is translated into consequence. The type of consequence being modelled, however, matters.

Frequently asked questions

How much could sea-level rise and storm surge cost Australia?

The study estimates present-value direct financial losses of $855.1 billion to 2100 and $1.91 trillion under RCP8.5 in its base case. These estimates cover property and land-value losses and exclude several other categories of consequence.

How many Australian properties could be affected?

Approximately 267,500 properties are projected to be affected by 2100 under RCP4.5 and 719,600 under RCP8.5.

Are these predictions of future losses?

They are scenario-based model estimates. The results depend on assumptions about sea-level rise, inundation, asset and land values, damage relationships and discount rates. They are best interpreted as estimates of potential direct financial consequences under the modelled scenarios.

Does most of the estimated loss come from buildings?

No. Around 63% of the study's estimated financial loss is associated with land value and 37% with property. The land-use component spans ten categories, including residential and commercial land, agriculture, infrastructure, public facilities, parks and reserves.

Does the study estimate the full economic impact of sea-level rise?

No. It primarily estimates direct losses to property and land value. It does not model wider economic effects such as business interruption, supply-chain disruption, infrastructure service loss, labour impacts or recovery dynamics.

Why does the study matter for adaptation planning?

It demonstrates how nationally consistent hazard information can be translated into spatially differentiated financial consequences. This provides a stronger basis for comparing regions and scenarios and identifying where more detailed consequence and adaptation analysis is warranted.