IAG calls for rigour with natural hazard data

IAG has called for greater resourcing of federal agencies with large data holdings on weather and geological risk to inject more scientific rigour into the models they produce.

The insurer’s senior manager of natural perils, Mark Leplastrier, told a senate committee late last week that better data - both historical and current - could help a range of sectors predict and prepare for predicted impacts of climate change.

Insurers rely on a variety of public and proprietary datasets to assess risk.

For example, a flood database commissioned by the Insurance Council of Australia back in 2012 enables insurers to easily check if a property is within the boundaries of a flood plain. The data comes mostly from local councils.

"The ability to actually release that information into the public domain has made a huge difference in insurers being able to access the best information and having a common source of truth as well," Leplastrier said.

“The last thing we'd want is misaligned risk messages in the community, and, if we can all work off the best information, that's really important.”

Leplastrier was keen to see a similar effort around the risk of tropical cyclones to Australia.

He noted that while the Bureau of Meteorology (BoM) maintained the Australian tropical cyclone database, “every man and his dog is basically picking up that dataset and trying to make sense of it to understand cyclone risk".

“If we could put much more scientific effort into establishing a much better source of truth there, we'll start being able to have aligned views of risk and what to do about it,” he said.

Leplastrier said Geoscience Australia had made some effort in the area with its tropical cyclone risk model (TCRM), which “digests historical cyclone tracks and … has the ability to produce a nice hazard surface or risk model you'd use in insurance or planning".

“That's set up as a community environment,” he said.

“Researchers can contribute to the model and improve on it. That might be an interesting thing to take forward, if we can actually really focus scientific attention into that model.”

The Insurance Council of Australia’s general manager of policy risk and disaster, Karl Sullivan, similarly said the public release of a PSMA-built dataset called Geoscape could provide huge benefits to industry.

“Geoscape is essentially the building footprint of every building in Australia and some attributes about them—things that the insurance industry would dearly like to know about every property, because then you don't have to ask your customer,” Sullivan said.

“That sort of thing being made freely available, or more freely available, would spark a great deal of innovation not just in insurance but in the building market, the planning area - it's almost impossible to whiteboard out all of the things that could occur from making that sort of thing become national information architecture.”

There is precedent in PSMA datasets being made public, given PSMA itself is owned by Australia’s nine Commonwealth, state and territory governments.

Two years ago, the geocoded national address file (GNAF) dataset was made public.

“What resulted was a huge amount of innovation not just from the insurance industry, who were suddenly able to access it more freely, but from small start-up companies, innovators and app developers, who were suddenly able to really think outside the box,” Sullivan said.

GNAF is also being improved by users such as NBN Co, which has been one of the most publicly affected by errors in the dataset.