Authors: David Egan Sara Sienkiewicz
An Interview with Gallagher Re's Global Severe Convective Storm Peril Lead, Dr Sara Sienkiewicz.
For decades, Severe Convective Storms (SCS) were relegated to the category of "secondary perils" — risks that were frequent but manageable, rarely threatening the systemic stability of the property reinsurance market. That era is over. With insured losses in the US consistently exceeding USD50 billion annually since 2023, and record-breaking hail events reshaping the risk landscape in Europe, SCS has officially moved to the main stage.
Managing this "frequency peril" requires more than just raw data; it requires the ability to untangle the complex interplay between climate change, expanding urban exposure and evolving building standards.
To lead this charge, Gallagher Re appointed Dr. Sara Sienkiewicz as Global Peril Lead for SCS. With a PhD in atmospheric science and experience in both weather prediction and hands-on catastrophe model development, Sienkiewicz's focus will be on translating complex data into useful insights for (re)insurance underwriters.
Sienkiewicz, who joined in October 2025, is one among a new group of peril leads within Gallagher Re's Global Property Research team, with specialists covering tropical cyclone/hurricane, wildfires and floods, in addition to severe convective storms.
Each peril lead is responsible for broad technical oversight of model development and evaluation, and supporting carriers in making informed decisions in pricing, accumulation management and reinsurance placement.
In this interview with Dave Egan, Gallagher Re's Head of International Property, Sienkiewicz shares how she's framing SCS risk as we move through 2026, what recent events have taught the industry and how analytics — from model evaluation to new regional capabilities — can support better outcomes.
Dave Egan: SCS losses were relatively light in the first quarter of 2026, but that's coming off the back of several challenging years. How would you characterize the recent experience of SCS activity?
Sara Sienkiewicz: We are seeing a "new normal" where the volatility of SCS is becoming a weekly consideration for the industry rather than a seasonal one. In the US, we've seen activity rising significantly after a quiet start to the year, with notable EF3 tornadoes and large hail outbreaks in the Midwest and Southeast.
Over the past few years, the experience has been well above average. Gallagher Re's Q1 Natural Catastrophe Report took a deep dive into the SCS statistics, showing how and why losses have topped USD30 billion in four of the past six years. Several recent large outbreaks have placed significant pressure on the industry in the US, including in March and May of 2025. In June, the first EF5 tornado in the US for 12 years struck North Dakota.
Globally, 2023 was a historic year for Italy, marking a record-breaking period not only for very large hail — including extremely large stones — but also for the persistence of storms. They kept reforming day after day, challenging previous ideas of what severe convective storm activity in parts of Europe can look like. For the industry, it reinforced the need to better understand volatility and how storm behavior may be shifting over time.
Whether it's tornadic storms in Brazil or monsoon-driven activity in Australia, the frequency and intensity of these events are forcing a total re-evaluation of how we price and manage this risk.
As we look towards the rest of the 2026 peak season, and move deeper into Northern Hemisphere Spring, one of the key drivers is increasing solar heating and the contrast between warmer air and retreating continental cold air from winter. That contrast is part of what creates the conditions for intense convective activity — and it's why we monitor this period closely as the season ramps up.
DE: What is your view on the "why" behind these rising losses? Is this climate change or an exposure growth story?
SS: It's a very good question, and the honest answer is that the interaction is complex and it varies by geography.
In the US, exposure plays a very large role. One thing we've seen is that some outbreaks have shifted away from the traditional "Tornado Alley" — the Great Plains — toward the Midwest and Southeast, where there's often more exposure. That creates a "chicken-and-egg" problem: is the risk increasing because storms are shifting, because exposure is expanding into different areas or because both are happening simultaneously?
A useful way to think about it's the "expanding bullseye" effect: as exposure grows — more property, more value, more density — the same level of weather activity can produce higher losses simply because there's more in harm's way.
Conversely, in other regions such as Australia, the densely populated regions haven't expanded much in recent decades — suggesting that the frequency and intensity of natural hazards like SCS is playing a more significant role in losses.
This topic was explored in depth in the Q1 Nat Cat report that I mentioned above, which found that exposure growth can explain about 90% of the annual increases in nominal losses we have seen since 2000 in the US.
From a science perspective, you can only hold so many variables constant. If variables aren't constant, the work becomes understanding the trends in both the hazard and the exposure to accurately assess the risk.
DE: From a catastrophe modeling perspective, what are the biggest lessons the industry can take from recent SCS experience?
SS: One of the most important disciplines in using catastrophe models — especially for a peril like SCS — is understanding the assumptions embedded in the model: how it represents weather events, how it translates hazard intensity into damage and whether observed reality fits within the model's bounds of frequency and severity.
Every year brings new data, and those lessons don't come only from hazard — they also come from engineering and claims outcomes: how hail, tornadoes and straight-line winds are damaging structures in practice. Detailed claims data, engineering feedback and event analysis help refine how we understand vulnerability.
Meteorology is also a relatively young science in many respects — modern satellite and radar eras are not that long in historical terms — so real-world events continue to be a major driver of learning and model evolution. Each event can add evidence that improves how we represent risk going forward.
DE: You mentioned the value of claims data in modeling — why is claims data so important? Does it pose any challenges?
SS: Claims data is incredibly valuable, but it's also messy — arguably even more complex than atmospheric data. Different companies store it differently, use different fields and capture different levels of detail. That becomes a real challenge when you're trying to learn from multi-company datasets.
For SCS specifically, roof data is a major example. Roof covering type and roof age are highly influential in how a structure responds to hail — but that level of information isn't always captured consistently at the point a claim is processed.
The goal isn't to turn claims data into something perfect — it rarely will be — but to link it intelligently to exposure and vulnerability assumptions. That's where you can start to see meaningful improvement in the quality of insights and loss estimates.
DE: For readers who aren't model developers, can you clarify how model vendors use this information?
SS: Typically, model vendors don't only provide a hazard footprint — for example, "how likely will a tornado affect a location?" They also include vulnerability functions that translate hazard intensity into expected damage to a structure.
These functions can account for characteristics such as roof type, roof age, building height and overall building age. That information is used to estimate damage as a percentage and then translate that into costs — including replacement costs — which can roll up into insured loss estimates. Gallagher Re's analysis shows that rising replacement costs have been a significant part of the loss growth story of recent years. Over the past decade and a half, price inflation in construction materials, particularly the asphalt materials used for roofing in the US, has been striking. Factoring this variable into modeled loss estimates has become particularly important.
Then, brokers and insurers often use those outputs alongside their own analytics and claims data to test sensitivity and understand how model choices affect loss distributions.
DE: You recently joined Gallagher Re as the Global Peril Lead for SCS. What attracted you to this role, and how does your background in atmospheric science inform your approach?
SS: I joined in October, so it's been about five or six months. I've been fascinated by weather since I was very young, which led me through education all the way to a PhD in atmospheric science. During that time, I studied a broad range of systems — not just severe thunderstorms but also large-scale winter storms and the jet stream — and I had hands-on experiences like using mobile Doppler radar during fieldwork.
After that, I worked at the National Weather Service's Weather Prediction Center, where I developed training tools for forecasters nationwide. I enjoyed the challenge of working with complex atmospheric datasets and building probabilistic tools to support real-time decision-making.
Later, I moved into catastrophe modeling and spent several years at Karen Clark and Company, developing models across perils and regions — including US severe convective storms. I worked primarily on the hazard side, collaborating with engineers on damage estimation and working with financial teams to translate that into loss output. I also spent time helping clients interpret model outputs during live events — for example, to inform operational decisions like deploying claims adjusters.
What attracted me to Gallagher Re was the opportunity to bring that combination — science, models and communication — into a role that is explicitly global and client-facing. I saw the Global Severe Convective Storm Peril Lead position as a way to use my experience not only to understand models, but to help clients and colleagues interpret them appropriately, including their limitations and sensitivities. It's also expanded my perspective internationally very quickly — the teams here have been very open about regional challenges, data realities and partnerships.
DE: Why is now the right moment for a global SCS peril lead to join Gallagher Re — and what does your role involve?
SS: SCS has clearly become more central to the industry's loss experience and risk conversations. In a role like this, a big part of the value is coordination and translation.
It's only natural for work to become project-focused and compartmentalized by region or business unit, given differences in data, client needs and market structure. The peril lead role is designed to connect those pieces: to be aware of model evaluations underway, tools that exist or are being developed and the different ways teams are approaching SCS across regions.
I sometimes describe it as being a "central server" for SCS information — so clients and colleagues don't have to go to three different people to get an answer. It also helps us identify when similar work is happening in different places, so we can share knowledge and avoid duplicating effort unnecessarily. The goal is to accelerate learning and make innovation more efficient — and ultimately make it easier for clients to access consistent, high-quality insight.
DE: How is Gallagher Re specifically helping clients navigate these challenges through modeling and analytics?
SS: We provide a mix of external model evaluation and proprietary tools. Our North America Hail Risk Score and SCS Score are critical benchmarks that calculate annual average losses at a granular level. We are currently expanding these into multi-peril tools that link vulnerability data directly to claims.
One of our most exciting initiatives is our partnership with the European Severe Storms Laboratory (ESSL). We are collaborating on a new pan-European SCS model that moves beyond purely statistical methods. By using ERA5 re-analysis data — which reconstructs past weather in a 4D framework — and NASA satellite data, we are creating a more physically grounded model. This will capture multiple sub-perils like straight-line winds and tornadoes, which have historically been underserved in European modeling. We can combine atmospheric data with algorithms designed to better identify where hail and severe storm conditions are most likely to occur over time, with higher temporal granularity. Ultimately, this is about improving how we represent frequency and spatial distribution — which is foundational to better risk assessment.
Gallagher's European hail model has been in use for a long time, and the next step is to move from hail-only to a multi-hazard model — including hail, straight-line winds and tornadoes. This collaboration allows us to provide clients with the latest risk information for a critical European peril, using a probabilistic approach that reflects how severe convective storms drive loss.
By expanding peril coverage and incorporating insights from recent events and claims data into damage functions, the model is designed to be relevant across the full risk lifecycle. Importantly, contractual features such as hours clauses can be captured, supporting more informed risk assessment. Together, these improvements allow the model to be integrated into day‑to‑day decision‑making — from more effective risk selection and profitable growth, through reinsurance solution design, to improved event response and faster claims settlement.
DE: We've heard a lot about AI's potential in this space. Is AI improving SCS modeling today?
SS: AI is an excellent tool for specific, data-rich tasks. For example, our property research team uses AI to remove non-meteorological artifacts from radar data — tasks that are easy for a human to identify but tedious to do at scale.
However, AI struggles with the "sparse data" problems inherent in catastrophe modeling. Atmospheric science is incredibly complex, and relying on AI without thorough planning can be risky. We use AI thoughtfully, but we believe the best outcomes come from "augmented intelligence" — where the machine handles the big data, but the human expert interprets the result.
DE: Finally, if SCS is now a "main stage" peril, what should clients prioritize when thinking about it?
SS: The first step is to gather as much relevant data as possible — especially the exposure attributes that drive vulnerability — and then make sure you have trusted expertise involved in interpreting it.
There're many tools in the market, and not every tool fits every portfolio or objective. The value comes from understanding model assumptions, testing sensitivity and using analytics in a way that supports real decisions: underwriting, portfolio design, accumulation management and reinsurance structuring.
My role, and the broader Gallagher Re capability, is to help translate the science and the model outputs into insight that clients and colleagues can use. I encourage clients to reach out when they have questions. The goal is clarity: not just more information, but better interpretation.
Understanding the volatility of Severe Convective Storms is the first step toward protecting your portfolio. Gallagher Re's Global Property Analytics team combines modern meteorology with rich market expertise to help you stay ahead of the storm.
