Catastrophe
Research

Summary maps from the National Interagency Fire Center’s wildfire outlooks for January 2025 and January 2026 look strikingly similar. Released on December 1st, each map forecasts wildfire risk for the upcoming January. Given that Los Angeles experienced the most destructive wildfires on record in January 2025, the designation of ‘normal’ risk in December 2024 may seem unexpected and fuel speculation about this year’s risk.

But the underlying conditions tell a different story. The end of 2024 was exceptionally dry, yet December is often a wet month in Los Angeles. As a result, the January 2025 outlook included this caveat: “However, if there are no significant precipitation events in December, there is a chance for January to continue to lean toward above normal significant fire potential for the South Coast Predictive Service Area [Southern California] in the next outlook.”

This year, in contrast, Los Angeles has already received more than 400% of its typical August–December rainfall. Fortunately, this year’s prediction doesn’t rely on rains we assume are coming, they have already arrived.

by Charles Powell (University of Cambridge) and Ruth Petrie 

The 2025 hurricane season was defined by extended periods of suppressed tropical cyclone activity punctuated by explosive bursts of intensity, marked by three Category 5 hurricanes. This pattern fits neatly into the model-predicted trend of increasing intensity of individual storms amid overall reduced activity.

• In the early season strong high-pressure systems and wind shear kept hurricane activity low despite very warm sea surface temperatures in the main development region.
• In the late season conditions became more favourable allowing for rapid intensification of storms like Hurricane Melissa.
• Short-lived influence from atmospheric waves triggered this season’s most powerful hurricanes when the background conditions became intermittently more favourable for development.

Early in the season, from June through July, the Atlantic basin was effectively locked down by anomalously high mean sea-level pressure (MSLP). This pattern promoted subsidence and drying of the mid-troposphere, while weak lapse rates and widespread warmth across the subtropical Atlantic further reduced large-scale convective support, despite the sea surface temperatures (SSTs) in the Main Development Region (MDR) being at near record highs. The suppressing environment was reinforced by a strong Tropical Upper Tropospheric Trough (TUTT), which amplified vertical wind shear and compounded the effects of high pressure and dry air, creating a hostile environment for tropical cyclogenesis.

By late summer, atmospheric conditions began to shift. Subtropical sea surface temperature anomalies weakened, mid-level humidity partially recovered, and the Caribbean developed into a region of enhanced convective potential. These improved thermodynamic parameters created an environment conducive to rapid intensification, as demonstrated by hurricanes such as Melissa, which leveraged these transient favourable conditions to achieve major hurricane status.

One of the most notable aspects of the 2025 season was the influence of equatorial wave forcing. Charles Powell (University of Cambridge), in collaboration with Inigo, has submitted a 2025 season retrospective to Weather, examining how clusters of intense tropical cyclone activity were associated with favourable phases of the Madden–Julian Oscillation (MJO) and eastward-propagating Kelvin waves. These wave events generated short-lived periods of enhanced upper-level divergence, creating localized conditions conducive to deep convection and storm development despite an otherwise unfavourable large-scale environment. This intermittent, wave-driven activity accounts for the season’s above-average Accumulated Cyclone Energy (ACE), concentrated within a small number of major hurricanes rather than distributed across numerous weaker systems.

The 2025 season was largely unfavourable for hurricanes but was punctuated by brief, wave-driven convective surges that produced some of the most powerful hurricanes on record. Even with ocean temperatures at historically high levels, this year highlights that hurricane formation still hinges on a delicate interplay of thermodynamic and atmospheric conditions.

Australia is about to enter its peak risk period for extreme weather, spanning October to April, when bushfires, cyclones, and floods are most common. The likelihood and intensity of these events are shaped by large-scale climate drivers, most notably the El Niño–Southern Oscillation (ENSO) and the Indian Ocean Dipole (IOD). These systems shift between positive, neutral, and negative phases, each defined by sea surface temperature anomalies in the Pacific and Indian Oceans, respectively. 

Currently, both ENSO and the IOD are in negative phases, with forecasts suggesting a gradual return to neutral conditions by February (Southern hemisphere outlooks).  

So, what does that mean for Australia? 

During neutral ENSO and negative IOD conditions, sea surface temperatures northeast and northwest of Australia are anomalously warm, respectively. Air travelling across these regions will heat up, ascend and rain out over Australia. These conditions also favour the formation of tropical cyclones, which can make landfall in Northern Australia. When ENSO enters a negative phase (La Niña), this Pacific circulation pattern strengthens further, amplifying rainfall and cyclone activity. 

Taken together, negative-to-neutral ENSO and IOD conditions significantly raise the likelihood of extreme rainfall across Eastern Australia this summer (Rainfall – The chance of above median for October to December – Climate Outlooks). Historical records reinforce this risk. Since 1967, insured losses from floods and windstorms between October and February align closely with ENSO and IOD phases (Insurance Council of Australia database benchmarked to 2022). While only 49% of years showed both ENSO and IOD values below zero, this subset accounts for 8 of the 10 largest loss years on record. This outlook suggests that wind and rain, rather than bushfires, are the most likely drivers of insured losses this summer.  

The Mediterranean Sea has warmed rapidly in recent years (Copernicus Marine Service). In new research using climate model simulations, I work with Stephen Cusack of Stormwise to show that greenhouse gas emissions are the dominant driver of this warming. Other human pollution, such as aerosols, temporarily slowed the trend, but as cleaner air policies take effect that offset is disappearing. The conclusion is clear: human emissions are responsible, and the Mediterranean is unlikely to cool down anytime soon. For hail risk across Europe, that means elevated risks are here to stay.

The past few summers illustrate the potential of this warming. The summers of 2022 through 2024 brought record average daily temperatures that neared or exceeded 28 °C, and 2025 is not far behind (Copernicus Marine Service). A hotter Mediterranean can provide more fuel for storms, and evidence suggests it has already contributed to damaging hail events in Italy, France, and Switzerland (see Piper et al., 2019; Kopp et al., 2023).

For the insurance market, knowing what drives extreme events is key to preparing for the future. While further research is needed to refine projections of European hail risk in the coming decades, this work shows that a warm Mediterranean is here to stay.

The 2025 hurricane season has been quiet so far with four extremely short-lived storms. A quiet (or active) start to any season always brings the question, will the rest of the season continue this way? One way to try and answer this is by looking at patterns of ocean temperature around the world. While El Niño-Southern Oscillation (ENSO) conditions in the Pacific often draw significant attention, there is a similar phenomenon in the Atlantic, referred to as Atlantic Niño/Niña, that also impacts hurricane risk. Notably, the Atlantic Niño/Niña region has experienced significant cooling over recent months, now aligning with La Niña conditions. This cooling may reduce hurricane risk by limiting the formation of hurricane seed storms off the African coast.

The possibility of this pattern was indicated as early as January, when forecasts suggested that Atlantic La Niña conditions could emerge during the hurricane season, albeit there was considerable uncertainty. It remains uncertain whether these Atlantic La Niña conditions will persist throughout the season or significantly influence it given the multitude of variables that impact hurricanes including the Madden-Julian Oscillation, steering flows, and Main Development Region temperatures. Nonetheless, it will be an interesting substory as the hurricane season progresses toward its peak.

Last night, a 30,000 m2 wildfire broke out in the Walthamstow Wetlands – close enough to my home that I could hear sirens as fire crews worked to contain it. There have been at least 14 wildfires in London since the start of June, including three yesterday.

So far in 2025, wildfires have burned more than 400 km2 across the UK – surpassing the previous record set in 2019 (287 km2). Only Romania has seen more land burned in Europe this year. The UK wildfire season typically peaks in spring with large fires this year in Galloway and Ystrad Fflur (65 and 50 km2, respectively). However, worsening drought conditions are making summer fires increasingly common.

In London, 48,533 people live within 100m of the rural-urban interface and are exposed to wildfire risk (Ordnance Survey; The I Paper). As climate change heightens this threat, we must act responsibly. Abandoning disposable BBQs or flicking cigarette butts isn’t just careless – it’s dangerous.

Predicted conditions for the 2025 hurricane season include above average Atlantic sea surface temperatures and neutral to weak La Niña conditions in the Pacific. Historically, this combination has been associated with more active and damaging hurricane seasons.

Sea surface temperatures (SST) in the Main Development Region (MDR) and the El Niño–Southern Oscillation (ENSO) are two of the most important large-scale drivers of Atlantic hurricane activity. Warmer MDR SSTs provide more energy for storm formation and intensification, while ENSO influences atmospheric wind shear. During La Niña conditions (ENSO < -0.5) wind shear is reduced making the environment more favourable for hurricane development and intensification. Conversely during El Niño conditions (ENSO >0.5) wind shear is enhanced making the environment more hostile for hurricane development and intensification.

The plot shows 40 years of North Atlantic hurricane seasons plotted in ENSO–SST space where the larger pink bubbles have larger losses and small blue bubbles have low losses. The lower-right quadrant is characterized by cooler SSTs and predominantly El Niño conditions. This quadrant represents the lowest physical risk regime, notably with only one major loss year – 2018 (Michael, Florence).

The upper-left quadrant is characterized by warm SSTs and predominantly La Niña conditions. This thermodynamic setup is highly conducive to both tropical cyclone formation and intensification. The clustering of large loss years in this quadrant is striking, demonstrating a link between the physical parameters and the losses experienced.

For 2025, ENSO forecasts suggest values between -0.5 and 0.0, while SSTs in the Main Development Region are expected to be above average. Extreme SST anomalies are not anticipated, so likely not more than 0.5°C above average. This places the upcoming season in the high-risk quadrant.

Looking at just the losses in the yellow box, which bounds the ENSO and SST predictions, we see years such as 2005 (Katrina, Rita, Wilma), 2017 (Harvey, Irma, Maria), and 2008 (Ike, Gustav) – some of the most devastating hurricane seasons on record.

Although risk is elevated this season, and a high-loss outcome is more likely given the physical setup, being in the high-risk quadrant does not guarantee that 2025 will be a high-loss year.

In wildfires, buildings are only as resilient as their most vulnerable features. It’s often assumed that homes with metal roofs will perform better during fires than those with asphalt shingles, which are more flammable. However, analysis using the Cal Fire Damage Inspection database — which tracks all structures damaged by wildfires in California since 2012 — reveals little difference in performance between single-family homes with metal and asphalt roofs. One reason is that homes with metal roofs are more likely to have wooden walls, while those with asphalt roofs are more often built with more-fire-resistant materials. This simple example highlights the importance of considering a building’s complete set of features when assessing its wildfire resilience. Focusing on just one or two attributes can provide a misleading picture.

Caption: Assessment of the influence of roof type and wall type of wildfire vulnerability. Left plot shows average damage of buildings in wildfires with 50+ buildings with asphalt and metal roofs in the Cal Fire Damage Inspection Database (https://data.ca.gov/dataset/cal-fire-damage-inspection-dins-data). Average calculated assuming categories ‘No Damage’, ‘1-10% Affected Damage’, ’10-25% Minor Damage’, ’25-50% Major Damage’, ’50-100% Destroyed’ which are given values of 0, 5, 17.5, 37.5 and 100% loss, respectively. Right plot shows proportions of different wall siding types for locations used in the lefthand plot.

June 1^st is right around the corner, and with it the start of the Atlantic hurricane season. When we issued our seasonal outlook in March, we forecast slightly above average activity for the upcoming season. Since then, ocean temperatures in the Atlantic have continued to cool from their 2023 and 2024 peak (1st image). This is good news for 2025 hurricane risk as cooler ocean temperatures mean less fuel for hurricanes.

We can translate what the recent cooling in the Atlantic means for hurricane risk using Inigo’s in-house model. Our model updates each month and forecasts ACE west of 60° longitude (ACE 60W). ACE 60W is a measure of total hurricane activity in the western part of the Atlantic Ocean, and our work has found this metric is connected to economic and insured hurricane losses in the US. Our model was forecasting an above average season for much of the early part of 2025, but has backed off those predictions in the latest May update (2nd image).  No prediction is perfect, but hopefully 2025 is quieter than the 2024 hurricane season.

As winter turns to spring, flowers begin to bloom and seasonal hurricane forecasts are about to be released. There will be a lot of seasonal hurricane predictions coming out in the next few months and almost all of them will rely on the six-month sea surface temperature (SST) forecasts released by national modelling centres. This raises the question, how accurate are these six-month SST forecasts?

There’s a lot of different ways that we can evaluate how good or bad a forecast is, but one straightforward way is to look at how well it has performed in the past. The plot shows SST forecasts for the Main Development Region of the Atlantic Ocean over the past 18 months, released by ECMWF, a leading modelling centre. By comparing past forecast anomalies (grey lines), released monthly, to observed SST anomalies (yellow line), we can see that the forecasts have correctly predicted the cooling trend in the Main Development Region since June 2024. However, they have often been too aggressive with the cooling (grey forecasts are mostly below the yellow observations).  Looking ahead to the start of hurricane season in June, the forecasts predict continued cooling, which would significantly lower hurricane risk compared to 2024. Whether the latest forecasts (blue line) are accurate remains one of the major questions that organizations releasing seasonal hurricane forecasts will need to grapple with. Inigo will be releasing our view in the coming weeks, so stay tuned!