When Hazards Collide

Extreme events are bad enough when they occur in isolation. A single event allows time to recover, time to rebuild, and time to plan for the next one. But when these events occur in tandem, the effects can be much, much worse than merely the sums of the losses of the individual events. They are compounded. This article explains in more detail the different types of compound events, as they are called, why they are expected to worsen, and some of the strategies actuaries may use to better understand and prepare for them.

A textbook example

The January 2018 Montecito, California, debris flow, which occurred just weeks after the 2017 Thomas Fire,¹ is a textbook example of a wildfire-flood compound event in which the resulting losses exceeded the sum of the individual disasters.

The first event was the Thomas Fire. At the time, this was the largest wildfire in California history. It scorched over 280,000 acres, destroyed over 1,000 structures, and resulted in two fatalities. Importantly, it also destroyed vegetation and turned the steep, mountainous soil into a hydrophobic (water-repellent) surface. ²

The second event was the intense rainfall on January 9, 2018. Only one month after the fire, an atmospheric river, which is a long narrow stream of concentrated moisture in the sky, brought heavy rain to the same area. Because the Thomas Fire had destroyed all the ground cover and the soil could not absorb water, the rain triggered massive, fast-moving mudflows that carried boulders, mud and ash down from the mountains, inundating the community of Montecito. The mudflows destroyed or damaged another 400 homes. Because they carried so much debris, the damage to homes and infrastructure was more devastating than that from traditional flooding. The debris covered a 30-mile stretch of U.S. Route 101 with up to 2 feet of mud and closed a major economic artery for 13 days. The mudflows also killed 23 people, a casualty count higher than the fire itself, which had low fatalities. Residents who had already evacuated the month before the fire were less prepared or willing to evacuate again, exacerbating the human tragedy. The burn scars in the landscape created pathways that increased the likelihood and severity of the debris flows, and it is likely that fewer deaths would have occurred, and far fewer homes would have been damaged had the fire not taken place so soon before the rainstorms. ^{3 4}

COMPOUND EVENTS COME IN MANY COLORS

Events like the 2017–2018 fire-debris flow are just one type of compound event; others exist. Compound events can occur in a variety of ways. Here are four different types.⁵

1. Preconditioned compound events occur when background environmental conditions amplify the impacts of a later hazard. The “preconditioning” may develop over weeks, months or even years. The example in the previous section is one such, but there are many more.

In 2022, Italy, particularly the Po River Valley, experienced a devastating compound event that began with a record-breaking, prolonged drought followed by intense, localized flooding. The Po Valley supplies 40% of Italy’s food and saw up to 70% crop loss in some areas, amounting to 6.5 billion euros in losses.^{6 7}

The subsequent flood happened, in part, because the drought created a hydrophobic surface that caused water to run downhill rather than get absorbed. The 300–400 mm of rainfall in May 2023 triggered a devastating flood that caused around $10 billion in economic damage.⁸

2. Multivariate compound events are a second type of compound event where multiple hazards are occurring simultaneously and interacting. Heat with drought is the classic example. As drought progresses, less humidity leads to higher heat. Higher heat reduces humidity. Strong hurricanes with heavy rainfall are another example. Hurricanes, in fact, can provide a quadruple punch, bringing wind, heavy rain, storm surge and tornadoes all at the same time.

Hurricane Katrina (2005) is a prime example of a multivariate compound event, causing $125 billion in damage by delivering a catastrophic combination of wind, heavy rain, massive storm surge and tornadoes. It destroyed or damaged over 300,000 homes, caused nearly 1,400 fatalities and produced a 25- to 28-ft storm surge that destroyed the Mississippi coast.⁹

3. Temporally compounding events are a third type where hazards occur in sequence, with each event weakening resilience or increasing vulnerability to the next. Beginning in December of 2022 and extending into the next year, a series of nine storms across California saturated soils and caused widespread flooding. Notable events included a bomb cyclone that brought damaging winds and record rainfall to the Bay Area and Central Coast. A resurgence of rain in March compounded the already compound event. The storms resulted in at least 22 fatalities statewide, and total economic losses were estimated between $5 billion and $7 billion, with over 700 landslides reported and thousands of individuals ordered to evacuate.^{10 11}

4. Spatially compounding events occur when similar or related hazards strike multiple regions simultaneously, creating aggregated impacts across large systems. A recent example was the March 2026 heat wave across the southwestern U.S. Most of the seven different states reported their warmest March ever, and almost half of the continental U.S. by area experienced average March temperatures highs in the top five, according to Climate Central. Potential agricultural impacts may emerge as assessments become available.
An important insight from this article, in my view, is that many real disasters fit more than one category. The labeling may be less critical than understanding what is possible.

CLIMATE CHANGE AND COMPOUND EVENTS

As this report shows, compound events are occurring more frequently and with greater intensity, and climate change may be a factor.

Although compound events are not new, they are becoming more frequent and severe, making them a growing concern. Climate change is contributing to the growing risk of compound events for several reasons. One is that it is making many events more frequent, bigger in size and stronger in intensity. These changes increase the likelihood that events will occur close enough in time or space to cause a compounding effect.¹²

The increasing heat from climate change means more drought. More drought means more wildfire activity, according to a report from the National Library of Medicine. More heat also means more rain because warmer air holds more moisture, and all else equal, that translates to higher rainfall rates when it does rain. But all else is not equal. More heat at the ground creates more instability, which can lead to more intense convective storms when other conditions are ripe, which can lead to bigger hail (storms) and stronger tornadoes, in addition to more rain. And more heat in the ocean means more intense hurricanes, if not more of them.

According to the National Academy of Sciences, climate change is disrupting the large-scale atmospheric waves that transport weather systems. The waves are larger in amplitude and slower in (eastward) movement thanks to Arctic amplification, which decreases the pole-to-equator temperature gradient and weakens the jet stream. These wave changes result in more intensely affected areas and more prolonged effects.¹³

Even events that did not reach catastrophic levels before are getting a climate change compounding boost.

Thus, even events that did not reach catastrophic levels before are getting a climate change compounding boost.

ACTUARIAL PREPARATION ASPECTS

For actuaries, compound events matter because they can produce “tail amplification,” generating losses that exceed what traditional single-peril models might suggest. Here’s a look at three implications:

1. Compound events challenge the assumption of independence between hazards. Many pricing frameworks, portfolio aggregation methods and capital models implicitly assume that risks are separable or only weakly correlated. Because climate change is making events more extreme—bigger, longer, stronger—the assumptions of independence as well as stationarity are being altered. Warmer oceans, higher atmospheric moisture content, shifting jet streams and changing soil moisture regimes are increasing the probability that hazards interact in damaging ways.
2. Compound events expose limitations in historical datasets. Traditional actuarial analysis relies heavily on past observations, yet the climate system is evolving. Non-stationarity means that relationships observed historically may no longer hold in the future. An event considered “rare” based on 20th-century data may become substantially more likely in a warmer world. This may create particular challenges for estimating exceedance probabilities, return periods and aggregate loss distributions.
3. Compound events create systemic and correlated losses across sectors and geographies. A severe drought may simultaneously impact crop yields, inland marine transportation, power generation and commercial supply chains. Likewise, extreme heat can drive mortality, worker productivity losses, infrastructure stress and increased wildfire potential simultaneously. These cross-line correlations are especially important for enterprise risk management and reinsurance accumulation analysis.

Modern catastrophe modeling has been at the core of quantifying risk for decades, but, in my observation, it is only just beginning to address these issues. Significant gaps remain. Some models still treat hazards independently or rely on simplified dependence structures. Actuaries may want to ask these questions when interacting with catastrophe vendors and internal modeling teams:

• Is the model physically based or data based?
• Does the model represent hazard interactions dynamically?
• How are antecedent conditions incorporated?
• Are climate-conditioned correlations included?
• How are cascading infrastructure and economic effects represented?
• Is climate change incorporated prospectively or only historically?

Answers to the above questions can better inform strategies for using existing models, even if they are not climate-compound-event ready. Actuaries may still use them effectively by combining outputs across catastrophe models, introducing climate-conditioned correlations, and applying scenario-based stress testing. Separate hurricane, flood, wildfire, or drought models may be linked geographically and temporally to explore how interacting hazards amplify losses beyond isolated-peril estimates.

Actuaries can also condition vulnerabilities on antecedent conditions—such as drought increasing wildfire risk or saturated soils worsening flooding—while incorporating large-scale climate drivers like El Niño–Southern Oscillation to create more realistic dependencies between hazards (see illustration, below).

This illustration by the author shows how La Niña can increase the risk of simultaneous wildfires across the western U.S., severe convective activity across the southern U.S., and hurricane activity in the eastern U.S. “H” and “L” indicate airflow aloft with high pressure (H) in the west and low pressure (L) in the east.

Additional value comes from analyzing portfolio accumulations, event clustering, infrastructure interdependencies and cross-line correlations that may not be captured within individual models. By supplementing traditional catastrophe models with climate science, stress testing and systems-level thinking, actuaries can better evaluate emerging compound-event risks even before fully integrated multi-hazard models become widely available.

In my experience, scenario analysis and stress testing are becoming increasingly important tools for addressing these uncertainties. Rather than relying solely on historical calibration, actuaries may need to evaluate plausible but unprecedented combinations of hazards. For example, what happens if a major hurricane coincides with extreme inland rainfall over already saturated surfaces? Or if a multi-year drought is followed by a severe wildfire season and subsequent debris-flow flooding?

Ultimately, compound events, for me, highlight a broader shift occurring within actuarial science—from viewing weather risk as isolated events, toward understanding climate risk as an interconnected system. Actuaries who develop expertise in compound hazards, climate science, and interactions with systemic risk may be well positioned to support pricing, reserving, capital management and long-term resilience planning in the decades ahead.

Statements of fact and opinions expressed herein are those of the individual authors and are not necessarily those of the Society of Actuaries or the respective authors’ employers.

Copyright © 2026 by the Society of Actuaries, Chicago, Illinois.