Open source software with a flexible modelling engine for processing geospatial data for risk analysis

A generic computer monitor with a screenshot of the RiskScape website.


GNS Science, and NIWA are both Crown Research Institutes. GNS Science’s purpose is to understand natural Earth system processes and resources, and to translate these into economic, environmental and social benefits. NIWA’s mission is to conduct leading environmental science to enable the sustainable management of natural resources for New Zealand and the planet.

GNS Science and NIWA had used their own development teams to build an app and release the RiskScape® project, which has been around since 2004. This version of RiskScape contributed to multiple research projects and informed many natural hazard risk decisions over the years. However, it did have its limitations being built on technologies from the early 2000s and including hard-coded data making the system inflexible.


Catalyst was already doing work at GeoNet so the RiskScape® Governance Group asked us to do an independent review of the RiskScape® software. We reviewed both the code and what was being done governance-wise. One of Catalyst’s suggestions was to use a third party development team, and maximise the opportunities to use the expertise and skill set that a specialist development partner brings. GNS and NIWA went to market for the software upgrade project and Catalyst won the tender.

RiskScape® is open source software with a flexible modelling engine for processing geospatial data for risk analysis.

It is highly customisable and can process multiple hazards and their relationships. For example, how might flood depth affect a timber-framed house on a concrete base, compared with a brick house on a timber base, and what is the additional impact of the flow rate. To make these calculations data can be input from a raster image that shows the depth of flooding, and vector data containing building footprint and construction information. Then you can ask, what are the adverse consequences for houses exposed to flooding?

Scientists could run these calculations longhand by using data and writing their own scripts, but RiskScape® uses advanced geospatial processing functionality with native support for multiple data types to assist natural hazards experts with their risk analysis. Scientists can either define their models themselves using the RiskScape® expression language, Java and Python, or they can use a step by step wizard that takes them through a more guided process.

One of the things Catalyst did at the beginning was take everyone on the project through our Agile and Product Owners Fundamentals training. This gave us an opportunity to think about how we were going to work together, and we were able to practise that. It gave GNS and NIWA an opportunity to learn about how we work, and we could learn each other’s jargon to help the project run smoother. The Agile methodology is a great fit with scientific methods. In both you have an idea, implement the idea, see what impact it has, and then make changes.

After development started about three years ago, it became clear other organisations and people were interested in using the software outside of NIWA and GNS. EQC approached GNS to come on board as a funder. EQC wanted to replace its existing modelling tool, Minerva, with RiskScape®


Catalyst worked really closely with the teams at NIWA and GNS on this project. Our team loves working with the scientists as they have hard problems to solve and they don’t tell us how to solve them; they describe the problem and the team at Catalyst figures it out, which made it a really collaborative project.

The RiskScape® software was released under AGPL so it stayed as available as possible for research. AGPL requires distribution of any code that you change, even if people are just using it on their server. AGPL sets the tone that people can take software and use it but must keep sharing it forward. This was particularly important when EQC came on the project as a funder, as they want to improve our resilience to natural hazards, by ensuring everyone has access to information that will help them make decisions.

RiskScape® has many potential uses. It has been utilised recently for land-use planning decision-making. Regional council’s contract GNS and NIWA to find out if it’s safe to build; will a location and buildings be exposed to natural hazards like landslides or flooding?

The Earthquake Commission (EQC) uses RiskScape® to analyse New Zealand’s residential building stock’s exposure to earthquakes. Understanding natural hazards and their potential impact is vital for New Zealand’s physical and economic security. Loss modelling helps EQC understand the impact of a disaster by providing realistic estimates of likely damage and loss. It is used for post-event modelling to provide near-real-time loss estimates and the spread of damage. This allows the insurance and emergency response sector to quickly understand the scale and spread of damage after an event. It is also used for modelling potential natural disaster scenarios to plan how to respond and build community resilience, and setting the EQC levy rate.

Recently this sophisticated risk-modelling tool was used when the Hunga Tonga–Hunga Ha‘apai volcano erupted in Tonga in January 2022, sending ash into the upper atmosphere and triggering a destructive tsunami. Scientists were able to feed in data about the volcanic ash fall, and model the impacts of three hazards: the ash fall, tsunami, and sonic boom, so when communications were re-established with Tonga they were able to provide that data. This also helped inform aid agencies around the world to gauge what level of aid they needed to send to Tonga.

“RiskScape’s® flexible design allows many sectors, such as land-use planners, asset managers, emergency managers and insurance providers, to better manage the significant risks from natural hazards that Aotearoa faces now and into the future.” Richard Woods, GNS Science Project Lead.

Catalyst has also been able to contribute to other open source projects as part of this work. For example our team found bugs in some of the libraries we use (such as GeoTools) so we were able to fix those and contribute the code upstream. Sometimes there were changes we made that couldn’t go upstream, and in that case we published the codebases on the RiskScape® site in the documentation for people to download and use.

Catalyst continues to work on enhancements and more feature developments for RiskScape® as part of our ongoing relationship with NIWA and GNS Science. While RiskScape® is highly technical science software for a specific user group, the things it’s being used for have a big impact on all New Zealanders.