The Future Arrived in June | Weeknotes #444

---

The Future Arrived in June

I’m going with the ol’ British standby, writing about the weather.

Because the UK and most of Europe is currently trapped under a massive heat dome, causing absolutely unprecedented June temperatures. Here in my part of London, the Met Office predicts we will experience two consecutive days of 39°C on Wednesday and Thursday, with a strong potential for the all time UK temperature records to fall elsewhere in the country.

The main cause for concern is that our night temperatures aren’t dropping below 20°C. The Met Office calls these “tropical nights,” which sounds like a holiday until you remember that the vast majority of British houses are built to trap heat inside. Because of the sustained high night time temperatures, neither buildings nor bodies get a chance to cool down. The heat rolls over from one day into the next, meaning that by Thursday, you’re not dealing with that day’s weather, but the compounding heat left over from Tuesday and Wednesday as well. Also, while temperature records are usually broken by fractions of a degree, the June record is on track to be cleared by three or four degrees this week.

I could put a whole rant in here about adaptation, particularly at the home scale, growing climbers on south-facing walls, the wonders of the nearly forgotten technowledge that is the ‘awning‘; but instead my mind has gone to the mechanics of climate models.

Inside Model Land

Having done a lot of climate futures work, and given my assertion that “worlds are the first new medium of the twenty-first century,” I often take the position that it’s best to explain that a climate model simulates a ‘possible world’ where a specific set of assumptions, such as whether global emissions rise or fall, are allowed to play out. These models do not produce a single definitive future. Rather, they map out a relative population of potential futures.

In her book Escape from Model Land: How Mathematical Models Can Lead Us Astray and What We Can Do about It, statistician Erica Thompson explains how mathematical modelling has become one of the defining technologies of modernity. We live in a society governed in “Model Land,” where economic forecasts, pandemic planning, energy grids, and weather reports all depend on simplified representations of reality. The entire point of a model is that it must leave things out, because if it contained every variable in existence, it would simply be the world itself. Therefore, the big question to ask of a model is what it includes, what it excludes, and what decisions are made based on its assumptions.

Climate models are some of the most complex model systems humans have ever built. They divide the atmosphere, oceans, and land into a massive multi-dimensional grids, using our best understanding of physics to calculate what happens inside each section. While they are incredibly sophisticated, models frequently struggle with small-scale events or rare conditions that line up at the extreme edge of probability. The volatile tail of the graph where records are broken.

When meteorologists or the media state that temperatures are “20°C above average”, they are measuring against a baseline reference period that runs from 1991 to 2020. This baseline is calculated by averaging thirty years of June data together to smooth the yearly weather noise into a readable pattern. However, a baseline spanning 1991–2020 is not really an accurate description of our climate today; it describes a historical period whose midpoint was 2005 a world we have already left well behind.

This shifting baseline is also a common source of misunderstanding when talking about climate scenarios/futures in workshops. When people hear that a particular extreme temperature is associated with a future decade like the 2040s or 2080s, they often mistake it for something they should expect on the calendar, like a timetable. They assume the weather will remain roughly normal until we reach those milestones. But climate scenarios alter the odds of an event happening rather than an event being a fixed temporal appointment.

As most of the work in climate futures has been in narrative strategy, and I’ve often found myself explaining the statistics of climate scenarios by analogy. Something like this:

Imagine a bag filled with numbered balls representing potential weather outcomes. In the old British climate, most balls had typical temperatures written on them, while almost none read 35°C or 40°C. Climate change doesn’t suddenly “come along” and hand you a brand-new bingo bag every decade or so. Instead, as the years pass, it swaps out the cooler balls for hotter ones. In any given year you can still accidentally pull out a cold, rainy summer, or draw a future extreme temperature well ahead of schedule.

Climate futures arrive day by day, first as freak weather, then as a run of bad luck, and then eventually as normal. For instance, the technical work underpinning the UK’s Third Climate Change Risk Assessment (CCRA3), published in 2022, examined climate projections for periods including 2021–2040. It concluded that exceeding 40°C before 2040 was possible but very unlikely in its model ensemble. Yet on 19 July 2022, the UK recorded 40.3°C at Coningsby. A rare event with a low annual chance can still come up on the first draw; that doesn’t make the estimate wrong. But it does show why low-probability extremes cannot safely be treated as events belonging only to the distant future. In fact a project I was working on just after the end of the pandemic was all about early thinking to explain why the prediction wasn’t wrong to the public.

To address this limitation, the Met Office and The Royal Meteorological Society recently published a paper using a different approach. Instead of relying solely on the relatively small number of summers contained in the observational record, the researchers used a larger climate-model ensemble to generate many physically plausible summers from past data. The method, known as UNSEEN (UNprecedented Simulated Extremes using ENsembles), allowed atmospheric and surface conditions to combine in ways that have not necessarily occurred but remain possible extrapolating from the present climate and recent past.

Using this much larger collection of simulated outcomes, they estimated that, in the climate of 2023, the UK had roughly a 1-in-24 chance of exceeding 40°C in any given year. Our current scenarios are therefore no longer merely warning that 40°C is possible. They suggest that it is becoming more and more repeatable: not a country touching an extreme temperature once as an isolated incident, but enduring it as a persistent feature of an altered climate.

The trouble with Model Land is that nobody actually lives there. While the UNSEEN ensemble runs thousands of parallel, simulated summers on a supercomputer in Exeter, we are left to sweat through one specific, unyielding iteration of reality. By Wednesday afternoon all the statistics I’ve read over the last 15 years will have faded into the background, and it won’t matter whether a 40°C day is a 1-in-24 fluke or a 1-in-300 anomaly when I’m inside a on bed Victorian terrace conversion whose back wall has spent 48 hours acting like a storage heater.

I’ve spent years narrating scenarios. The job with them (i think) is mostly standing slightly to one side of a possible world and describing it for people. A summer could get this hot. By the 2040s the odds might do this or that. You’re always describing a world you haven’t arrived in. This week though I don’t get to stand there, the world I’m inside isn’t a scenario, it’s just Thursday in June.

---

Photo 365

Terminal Access

In the LRB, Jay Owens this week wrote about the introduction of the Nature’s Rights Bill, which seeks to grant nature inherent legal rights as a rights-bearing entity rather than treating it as “natural capital.”

In the UK, though, rights of nature are so far more gesture than law. Crowhurst Parish Council in East Sussex has bestowed ‘personhood status’ on a 1300-year-old yew tree, which encouraged Rother District Council to announce its support for the rights of trees more widely. Also in Sussex, Lewes Council recognised the rights of the River Ouse, including its ‘right to an active and influential voice’ in planning and development, though ‘the council cannot and has not granted the River Ouse legal personality’.

Reading

I finished Breakneck: China’s Quest to Engineer the Future by Dan Wang. Its really good. A fascinating look at China and America as the “Engineering State vs. Lawyerly Society”. Learnt so much.

I’m also reading Postal Intelligence: The Tassis Family and Communications Revolution in Early Modern Europe. I’ve already recommended it to Benedict despite only being 20% of the way though.

Postal Intelligence connects and situates histories of the post and government intelligence alongside print technology and state power in the wider context of the early modern communications revolution. In the sixteenth century, postal services became central to domestic governance and foreign policy enterprises, extended government reach and surveillance, and offered new control over the public sphere.

Stephen Wilson Jr – Preachers Kid (Single)

I told my friend Jess to check out Stephen Wilson Jr the other week, søn of dad was one of my favourite albums of 2024. Though I realised today that I didn’t post about it on the blog at the time! Listen to SWJ. One of the best songwriters today.

Anyways he has a new single out. Its amazing.

Remember Kids:

At each instant of choice, the protagonist created “diverse futures … which themselves also proliferate and fork,” producing a convoluted “network of times” that embraced “all possibilities.

My Tiny Life – Julian Dibbell