Climate science insight: Why resilience starts with understanding extremes

This research document, authored by Dr Sophie Lewis, Chief Scientist – Engagement at ACCR, addresses the critical need for investors to shift their approach when assessing physical climate risk and portfolio resilience. In recent investor dialogues, focus has often centred on identifying the most likely level of future warming and the specific climate shocks expected in those scenarios. However, from a climate science perspective, meaningful planning is impossible without accounting for high-impact, low-probability occurrences, often referred to as tail risks. In complex systems characterised by uncertain tipping points and irreversible change, it is these tail events that drive actual risk rather than statistical averages.

By focusing solely on the most probable warming scenarios and their associated impacts, stakeholders risk overlooking events that are critical to understanding the resilience of assets, companies, industries, and entire portfolios. The report suggests that instead of prioritising the most likely outcomes, the primary question should be: what happens if these extreme tail risks materialise, and are current systems and institutions robust enough to withstand them?

The physical risks we should be considering in our climate risk assessments

The concept of tail risks encompasses the rarest and most extreme climate events, including unprecedented weather patterns, compounding series of events, and critical tipping points. Although these events have a low likelihood of occurring, they play a disproportionately large role in the climate risk envelope due to their costly and destructive impacts. In recent decades, the nature of these risks has shifted alongside rising global temperatures. As warming increases towards 1.5°C, 2°C, or 3°C, many types of events previously classified as extreme tail risks are becoming common occurrences. For example, a flood event that was considered extreme in 1980 is now moving into the realm of likely today, and further warming is expected to make floods even more frequent and severe.

The report highlights three recent examples of record-breaking tail events, also known as Black Swans. Firstly, the 2025 Los Angeles wildfires were the most destructive in the city’s history. These were driven by a “climate whiplash” event, where two unusually wet winters in 2022-23 and 2023-24 promoted the growth of grass and brush, followed by extreme dryness in 2024 that rendered this vegetation highly flammable. Combined with record low winter rainfall and strong Santa Ana winds, these conditions led to unprecedented destruction. Secondly, a crippling heatwave hit the Pacific Northwest and Southwest Canada in June 2021, shattering temperature records. This included a new all-time Canadian record of 49.6ºC in Lytton, which significantly broke the previous 1937 record of 45ºC. Thirdly, Hurricane Sandy in 2012 was the largest Atlantic hurricane ever recorded; its unusual path resulted in a devastating storm surge that caused extensive flooding across 24 states.

The document also emphasises the impact of compound events, which are back-to-back or simultaneous weather events. Examples include an extreme heatwave immediately followed by bushfires, or a series of damaging storms, such as the nine named storms that impacted Spain, Portugal, and Morocco within a 32-day period during January and February 2026. These compound events can push infrastructure—including buildings, roads, and energy systems—beyond their design limits, resulting in cascading system failures. Increasing global temperatures effectively “loads the climate dice,” making compound and systemic risks more probable with every fraction of a degree of warming.

Tipping points are also identified as major tail risks. When critical thresholds are breached, abrupt and irreversible changes can occur in systems such as ice sheets, the Amazon rainforest, or the Atlantic oceanic circulation. The report notes that coral reef systems have already crossed such a tipping point. Failure to achieve rapid reductions in greenhouse gas emissions could lead to other tipping points being reached, potentially causing massive sea level rise, extreme cold in Europe, more destructive hurricanes, and abrupt carbon release from permafrost. While cities may currently cope with occasional tail events, continued warming will likely overwhelm these systems.

The physical climate risk questions we should be asking

To improve resilience, risk assessments must move beyond focusing on the “most likely” impacts and adopt a full risk spectrum approach. The report recommends that stakeholders ask a broader set of questions, including: What high-end temperature projections are possible for the year 2100 based on emissions pathways and modelled climate responses? What high-impact, low-probability events are modelled for these higher warming levels? Specifically, what events are projected for 2.6°C of warming, which is the level currently modelled based on existing global emission reduction commitments? Finally, can an asset or institution cope with these “how bad can it get” scenarios? Asking whether an asset can withstand specific climate events during the design phase of large projects is crucial for long-term stability. By expanding the consideration of physical risks, investors can better assess whether portfolio companies are truly resilient to high-impact climate change. Future insights will further discuss systemic risks, where impacts cascade across regions and supply chains.