Creating the best possible climate models is existential

As spring is in full swing, many may have already forgotten that Tokyo saw its warmest summer and fall last year, smashing the short-lived previous record of 2023. In fact, 2024 was the hottest year on record in Japan, in keeping with a global trend.

The heat was truly unprecedented, with deviations in Tokyo so large and persistent that daily historical maximums of 27 degrees Celsius often became daily minimums in the fall.

With a significantly more intense and extended zansho — that is, lingering summer heat — vendors selling kakigōri, the shaved ice dessert, experienced a sales boom. Meanwhile, at some shrines, prayers were offered for an end to the relentless warm weather, particularly given that a string of record-hot years has in part contributed to recent rice shortages across the country.

Japan didn’t endure a miserable summer alone, with reports showing that in the Northern Hemisphere, the season was potentially the hottest of the last 125,000 years.

A persistent run of global heat that started in May 2023 has rattled some in the climate science community, spurring a hive of research around climate extreme event attribution — the extent to which global warming impacts extreme weather events — and a frantic search for the cause(s) of the ongoing heat spree.

A critical question is whether recent observations show that warming is accelerating beyond the predictions of existing climate models: Even if there is a small chance that we find ourselves in an accelerated heating phase, what does all this mean for policy to tackle the crisis?

To date, evidence of the factors behind accelerated warming is mixed. Possible explanations range from the 2022 Hunga Tonga volcanic eruption and El Nino weather phenomenon to the reduced cover of aerosols, which act as a blanket reflecting radiation, due to a change in international shipping regulations. However, these are thought to make up only a portion of observed heating.

Ongoing research into the dramatic decline in Antarctic Sea ice may yield yet more clues. Especially concerning is that the lost sea ice continues to show no sign of refreezing, and if this persists, it would mean that climate models have underrepresented at least one key factor — known as a “feedback mechanism” — therefore resulting in higher temperature projections overall.

Key developments in global climate policy shaped at events such as the annual COPs, the United Nations climate change conferences, often follow the release of reports by the Intergovernmental Panel on Climate Change (IPCC), which provides a global consensus view on the climate.

Therefore, a potential upward revision of the IPCC’s findings in its next release, the 7th, due in 2027, would have far-reaching implications.

Reviews of underlying methodologies and resulting temperature projections by the IPCC have occurred before but have largely been of little consequence. However, this happened at times when there was less awareness of climate risk and urgency around the need to take action.

Last year, the world was on the brink of or may have already passed the ominous threshold of average global temperatures exceeding 1.5 degrees above preindustrial levels. The next time could be different because we are more aware of the threats involved in passing this mark. That said, an upward revision by the IPCC is not a given.

This body is considered the bastion of scientific integrity on climate change, but critics have noted its delayed incorporation of the latest climate science and use of watered-down language, ostensibly due to political pressure, at certain times.

With extreme event attribution science increasingly linking natural disasters to recent changes in the climate, the integrity of IPCC projections may also be further scrutinized — leading to questions around the provision of unbiased, accurate and transparent scientific information that clearly communicates climate risks.

This is especially important because IPCC outputs are often used by commercial climate service providers who translate or “downscale” the raw macro data into individual firm- or site-level climate risk assessments.

In fact, a more nuanced application of complex geophysical models to help inform real-world decisions is already happening in the insurance industry, where the robustness of commercial catastrophe models has improved over time.

Climate policy often involves weighing the nearer-term costs of decarbonization against future losses in the absence of substantial reductions in greenhouse gas emissions, known as the “social cost of carbon”.

This cost is a highly uncertain estimate — ranging from $32 to $1,000 per metric ton of carbon dioxide — and one that is highly sensitive to warming scenarios.

If climate projections are indeed revised upward even by a relatively small margin (note that small shifts in mean temperatures result in disproportionate increases at the “hotter” end of the spectrum), this could lead to a double whammy: Additional heating would have to be factored into the IPCC’s current middle-of-the-road warming scenario (termed “SSP2-4.5”) and this scenario would trend toward an even more pessimistic one (SSP3-7.0) if progress on net zero is stalled by a lack of political will.

This could send reverberations across the financial industry — where regulators and international consortiums have been championing the early adoption of climate scenario analysis to assess risks — as well as firms who have relied on scenario analysis for disclosure and other purposes.

At a top policy level, this could signal the need to overhaul existing climate risk assessment practices for disclosure and stress testing purposes as well as those used to inform decisions. This could translate into more radical net-zero pathways and higher investments in adaptation infrastructure to respond to greater losses.

Given that estimates of the social cost of carbon are likely to be wildly optimistic, a different physical risk assessment paradigm may emerge, one relying on forward-looking rather than historical data — something previously considered far-fetched.

This includes incorporating the social and economic impacts of tipping points being surpassed — including the collapse of the Atlantic Meridional Overturning Circulation, a system of currents within the Atlantic Ocean — or stress testing ecological and societal resilience under extreme heat above hazardous wet bulb globe temperatures.

It is difficult to craft a positive narrative from the current persistence of record heat while scientific agencies in this field are being systematically dismantled in the United States, the world’s second-highest emitter, at a time when we ought to be studying the climate more closely than ever before.

Other countries must therefore fill the void by increasing funding for advanced climate institutions, including the Japan Meteorological Agency. In turn, these should double down on efforts to improve the climate projections that form the basis for better policy responses.