Today’s Nat Cat Modeling: Wildfire, Earthquake, and Flood Insights

5 January 2026 / 7 min read

From bushfires in Australia to earthquakes in Mexico and floods across Southeast Asia, the catastrophe risk landscape is evolving. Verisk’s latest models use up-to-date data to help insurers better quantify and manage complex natural hazards.

Increasing global natural catastrophe losses are pressuring insurers and reinsurers to sharpen their understanding of risk. At the recent Verisk Insurance Conference (VIC) in London, experts from Verisk’s Catastrophe & Risk Solutions division introduced new or updated models that reflect this reality, providing a more detailed view of how natural hazards unfold and how losses accumulate.

Verisk’s latest research brings new insights for three peril regions: bushfires in Australia, earthquakes in Mexico, and flooding and typhoon-related losses across Southeast Asia.

As climate change and urban growth reshape the global risk landscape, Verisk continues to advance catastrophe modeling to help insurers and reinsurers better understand and manage natural hazards. Drawing on richer data, improved physics-based methods, and localized insights, Verisk’s latest suite of models offers a clearer, more detailed view of how events such as bushfires, earthquakes, floods, and typhoons unfold—and how they affect portfolios worldwide.

From enhanced bushfire modeling in Australia to a reimagined earthquake framework for Mexico and newly developed flood and typhoon models across Southeast Asia, these updates represent a significant advancement in understanding catastrophe risk. Each model reflects years of scientific research and industry collaboration, providing users with a more granular view of industry exposure, refined hazard modules, and the ability to capture catastrophic risks that often drive loss.

A deeper understanding of bushfire risk in Australia

Bushfires continue to be a serious concern in Australia, and as demonstrated in 2019 – 20, they can be catastrophic and costly. Verisk’s updated Australia bushfire model, released in June 2025, captures the nature of bushfires across the continent through a holistic enhancement of the previous modeling framework.

This work led to an entirely overhauled catalog that now supports 100,000-year simulations. “We developed a high-resolution and physics-driven hazard module and thoroughly updated the vulnerability module, including adding several secondary risk characteristics,” Bernhard Reinhardt, Assistant Vice President, Catastrophe & Risk Solutions at Verisk, told the audience at VIC London. These updates draw on an expanded set of historical data, with thousands of additional observed fires and insurance claims since the last release in 2017. The new catalog also incorporates six major historical events, including the devastating 2019-2020 bushfires that caused multibillion-dollar losses across Australia which can be used for benchmarking.

The updated model divides Australia into more than 70 bioregions, capturing local vegetation, climate, and fire statistics - key factors that directly affect potential losses and risk assessment. It also considers large-scale climate drivers, such as the El Niño Southern Oscillation, which can significantly influence weather patterns across Oceania and, in turn, the bushfire risk.

Refinements to fire-spread dynamics make the model even more relevant for insurers and (re)insurers. “If we think back to what happened in California in early 2025,” said Reinhardt, “the catastrophic fires there demonstrated how winds can accelerate spread and challenge suppression efforts. Our model explicitly accounts for the weather development throughout the event, giving insurers a more realistic view of potential claims scenarios.”

The model also reflects growing resilience efforts by individuals and authorities. It supports address-level mitigation features, such as wall and roof construction materials to reflect those efforts. Therefore, investing in capturing such granular data can significantly enhance the accuracy of underwriting and risk assessment, which enables both insurers and reinsurers to more precisely quantify exposure, differentiate risk profiles, and make more data-driven decisions.

Reassessing earthquake risk in Mexico

Mexico’s position atop three major tectonic plates make it among the most seismically active regions in the world. Verisk’s updated Mexico earthquake model , also released in June, reflects 15 years of new scientific data and engineering knowledge, along with a fundamental rebuild of country wide exposure assumptions.

“We started with a comprehensive update of our industry exposure database,” said Reinhardt. For example, the new version now explicitly captures certain types of high value exposures like large industrial facilities and considers market specific practices like coinsurance structures.

“We have also created a differentiated view of which of the exposures are actually insured,” he continued. “Given it’s typically the better-built or engineered structures that are insured, this provides a much more realistic view of the risk—especially important in Mexico, where detailed exposure data is relatively limited.”

The model’s hazard component has been completely redeveloped to give users a more accurate and insightful view of earthquake risk. New stochastic catalogs, now integrate the latest scientific understanding - drawing on lessons from major global events such as the 2011 magnitude 9.2 Tohoku earthquake in Japan and local events, like the 2017 Chiapas earthquake.

The updated ground motion model provides a more realistic representation of how earthquakes affect key exposure areas. “Earthquakes impacting Mexico City are a major driver of tail risk,” said Reinhardt. “Because the city is built on the sediment of an ancient lake - very soft soil that amplifies shaking - the new model better captures this behavior and its impact on loss potential.”

To give users a fuller view of secondary impacts, the model also now includes three additional sub-perils: landslide, liquefaction, and tsunami. “As is often the case with sub-perils, the closer we zoom in, the more important they become,” Reinhardt explained. “With these enhancements, insurers and (re)insurers gain finer granularity and a clearer ability to differentiate risk at the location level.

Flood and typhoon risk in a rapidly changing Southeast Asia

Flooding remains one of the most persistent and costly natural hazards across Southeast Asia, driven in part by heavy rainfall, steep terrain, and rapid urban development. To help insurers, reinsurers, and governments better understand and manage these risks, in June 2025, Verisk introduced brand new inland flood models for Malaysia and Indonesia, alongside a major update to the typhoon model for South Korea. These models integrate the latest science and data to deliver deeper insights into evolving flood and typhoon perils.

The new flood models are built on 10,000-year stochastic event catalogs and a 2024 industry exposure database with one-kilometer resolution. Verisk models support a wide range of asset types - from residential and commercial buildings to marine and infrastructure risks. “We pay a lot of attention when building our models to ensure our exposure databases are as representative as possible of conditions on the ground,” said Emma Lewington, Model Product Manager, Catastrophe & Risk Solutions at Verisk. “For this model, that meant identifying high-rise buildings and valuing large industrial facilities, consistent with our latest generation of flood models.”

In Verisk’s inland flood models, physically based pluvial and fluvial components work together to create a unified view of flood risk. Capturing the contribution of pluvial flooding is particularly critical in fast-growing urban areas like Kuala Lumpur, where reduced permeable surfaces and ageing drainage systems heighten flood risk. “Importantly, these factors are modeled dynamically,” Lewington explained. “For example, as surface accumulation increases, the efficiency of drainage systems decreases - just like we see in reality.”

Jakarta faces similar challenges, which are further compounded by its geography. “Jakarta sits in a low-lying delta and is experiencing some of the world’s fastest rates of subsidence,” Lewington said. “It’s intersected by 13 major rivers, and understandably, flood protection is a major government focus. We wanted to ensure our model accurately represents those fluvial protection measures too as they play a crucial role in reducing potential flood risk and impact resulting modeled losses” Lewington added that traditional tropical cyclone classification often overlooks the associated flood risk. “Cyclones have typically been categorized by wind speed or central pressure,” she said, “but that often misses the severe flooding and economic losses that accompany extreme precipitation.” The 2022 Typhoon Hinnamnor illustrated this well—making landfall in South Korea as a Category 2 storm yet causing major flooding and economic disruption.

Verisk’s updated South Korea typhoon model addresses this gap by explicitly modeling the effects of storm surge and typhoon-induced precipitation flooding. The model applies component-based damage functions to better capture flood impacts across a wide variety of building characteristics. “Wind speeds tend to be lower in South Korea, and intense precipitation is the dominant driver of typhoon losses in this region,” said Lewington. “By understanding a building’s number of stories, we can more accurately represent how value and engineering standards vary—improving risk differentiation and decision-making for insurers and reinsurers.”

Pioneering catastrophe modeling

Each of these new and updated models reflects how the science of cat modeling advances in step with risk. With richer data, more sophisticated physics, and refined local context that builds on global learning, Verisk’s models enable insurers and reinsurers to view risk with unprecedented detail, helping them navigate a landscape where the past is only one part of understanding the future.