Heat and Pollution as Weapons in the Class War: Reclaiming the Clouds of Athens

The summer of 2026 saw extreme heat events across Europe. In many respects Greece was spared the worst. Still, according to recently published data from the National Observatory of Athens (NOA), it was still one of the warmest Greek summers on record, averaging 1° Celsius above the 1991-2020 mean, with August around 1.6°C above normal. At the same time, AtmoHub, the specialist atmospheric research platform coordinated by NOA, recorded the effects of a series of wildfire episodes on the atmosphere over Athens and Attica. Significantly, they noted that the air in Athens is affected in different ways at different heights, and is impacted not only by local and bio-regional fires and smoke, but also by continental and beyond air pollution events. 

Athens makes these relations unusually visible. Situated in a basin surrounded on three sides by mountains, the city is an extraordinarily dense field of concrete construction, unshaded thermal mass and relatively little green space. It is consequently especially susceptible to both urban heat island and urban pollution island effects, as well as to the feedbacks between them. These conditions are now being amplified by climate change. Extreme heat increasingly combines with drought, forest fires, dust carried across the Mediterranean from the ongoing desertification of the Sahara and Sahel regions in Africa, as well as more local traffic and shipping pollution, fuel-poverty driven wood and rubbish burning, and the multiple consequences of decades of austerity in public services and local government, combined with uneven urban development driven by private property speculation of various kinds.

Thus, heat and pollution are not simply effects of climate change. In the contemporary city, heat and pollution are produced and unevenly organised. Heat is organised through the materials and density of buildings, through roads and traffic, through land values and patterns of investment, through access to trees, water and shade, through working conditions, energy prices and the ability – or inability – to retreat into air-conditioned interiors. Pollution is organised through many of the same systems. The atmosphere is a profoundly political space.

But these conditions are not experienced equally. The impacts of extremes of heat and pollution always play out through existing socio-economic inequalities, intersecting through class, gender, race and legal status. Wealthier Athenians can afford air-conditioning in summer and heating in winter. They are more likely to work in similarly conditioned environments, move around in air-conditioned cars and live in greener, less polluted neighbourhoods. They can leave the city during the hottest periods. Those without these possibilities live and work in environments over which they have much less control.

In previous texts I have described these relations as forms of environmental slow violence. I increasingly think that we need stronger language: Heat and pollution are becoming weapons in the class war.

This of course does not mean that someone consciously picks up heat and points it at the poor. The violence is systemic. Property relations, transport systems, labour conditions, energy infrastructures and investment decisions organise differentiated exposure and protection environments. They produce an atmospheric geography of class. Air-conditioning captures the contradiction particularly clearly. During an extreme heatwave it can be life-saving. Yet the dominant model of individual air-conditioning privatises climatic adaptation: those who can afford it purchase a temporary interior climate while compressors consume electricity and reject heat into already overheated streets. The deteriorating atmospheric commons is compensated for by increasingly energy-intensive private bubbles of climatic protection.

The wealthy are both disproportionately insulated from environmental degradation and disproportionately responsible for producing it. This pattern operates at every scale, from the air-conditioned apartment and car to global patterns of consumption, extraction and carbon emissions. It is commendable that the current administration in the Municipality of Athens has recognised and are responding to this reality. Supported by some genuinely excellent long-term research led both by the Municipality’s planning department – I’m thinking in particular of the ‘Athens Resilience Strategy for 2030’ – as well as important work and initiatives by scientists at NOA and academics at NTUA, a number of important initiatives are now underway in Athens, with the municipality explicitly treating cooling as a social right, and providing a network of cooled spaces during significant heat events.

Reclaiming which Cloud(s)?

The theme of this year’s Trans-Europa Festival is Reclaiming the Cloud(s). The plural is important. The first cloud we have encountered above: the atmosphere itself and the turbulent mixture of gases, particles, heat, moisture, smoke, dust, pollen and living matter within which Athens and its inhabitants breathe. But there is another cloud spreading across Attica: the apparently immaterial cloud of digital computation and data-centres.

The immaterial metaphor of the digital cloud is, of course, misleading. The cloud is extraordinarily material. It consists of mines and minerals, semiconductor fabrication plants, fibre-optic and undersea cables, electrical grids and enormous buildings full of servers. It requires land, electricity and water, and produces significant amounts of heat.

A new landscape is currently being constructed across eastern Attica shaped by this second cloud. Microsoft’s €976 million Greek cloud investment consists of three data-centres at Spata and Koropi, with a combined specified capacity of 38.4 MW. It was designated a Strategic Investment by the Greek state and has been developed through special spatial plans for strategic investments, partially displacing normal planning procedures. One of the Koropi sites alone occupies more than 100,000 square metres of previously agricultural land. 

The scale of the infrastructural consequences is striking. Research by Charis Papaevangelou and Eugenia Siapera estimates from the project’s own assessment documents that Microsoft’s largest Spata facility could initially consume electricity equivalent to around 82 per cent of the municipality’s existing annual electricity consumption, rising to 159 per cent at maximum installed capacity. They also document local concerns about an already degraded water infrastructure and a proposal to create a more circular system of wastewater treatment and reuse that was rejected because of the additional cost to the investment. Microsoft, for its part, states that its Greek facilities are designed to minimise both energy and water consumption and will use direct evaporative cooling only during the hottest periods – when temperatures rise above approximately 29.4°C – with water recirculated several times before part evaporates and the remainder is discharged into the wastewater system. 

There is a peculiar ecological contradiction here. It is precisely during those periods when Attica is hottest, driest and most climatically stressed that water becomes useful for cooling both human bodies and computational machines. At this same moment, Athens is confronting a much larger crisis of metropolitan water supply. The Greek state is planning major new infrastructure to transfer additional water towards Attica from other hydrological basins. This project is not officially justified as infrastructure for data-centres, and indeed it would be wrong to claim that it is. However, their coexistence poses a question which any ecological politics of Athens must confront: Can ecological transition simply mean continually enlarging the territory from which the metropolis extracts water and energy in order to satisfy whatever new demands emerge? Or do we also have to ask: which resource demands are necessary? Who creates them? Whose interests do they serve? And what kinds of collective life do they make possible?

As much as our mainstream media and mainstream political discourse wish to ignore such issues, these questions of large-scale socio-ecological strategic planning are unavoidable. Such plans are currently being made behind closed doors, in the extractivist interests of what Greek economist and political theorist Yanis Varoufakis has described as a new techno-feudalist elite. It doesn’t particularly matter if the new data-centres being built across Attica are powered and cooled by apparently renewable sources. As I have previously argued, following Murray Bookchin: a wind farm owned by a multinational power company is not necessarily an alternative technology! Decarbonisation without transformations in ownership, land and air rights, and global material relations will simply reproduce existing extractivist systems in new forms. 

The questions concerning socio-ecological planning and strategy have over the last decade or two been most powerfully developed around the growing dialogue of proposals, experiments and demands articulated through various Green New Deal proposals, and the discourse around a ‘Just Transition’. At its most radical, the Global Green New Deal proposes not simply technological substitution but a just transition grounded in democratic ownership and control, environmental justice and new institutional forms. For example, a Green New Deal cannot simply ask how can we produce more renewable electricity? It must also ask: what is energy for? What is water for? What is land for? And increasingly today, what is computation for?

We need as many Green New Deals as there are communities to organise around – and at the same time a meta-dialogue capable of connecting them. Specifically for the context being discussed here, we need to start discussing what an Athenian, Attic, Greek and European Green New Deal might look like, beyond the extractivist and capitalist lens which has shaped most local and European projects to date. This distinction is close to one recently made by the Greek architect, planner and political ecologist Ilias Gianniris, between a ‘Green Transition’ and ‘Ecological Transformation’. A transition can remain essentially technical, technocratic and economic, he argues. Transformation necessarily involves society itself.

The cloud that models the cloud

We find ourselves today adrift in contradictions. The critique of digital infrastructure cannot mean a simplistic rejection of computation, modelling or artificial intelligence. These tools, in some form, will surely be essential to any progressive socio-ecological project. Properly socialised, they will be key contributors to any radical Green New Deal Dialogue. As James Bridle has suggested in Ways of Being and elsewhere, the emergence of artificial intelligences might also provide an opportunity to both question the forms of human intelligence, while recognising the multiple other forms of intelligence already operating within ecological systems. The political question therefore includes asking: what kinds of intelligences and relationships might our computational systems allow us to newly perceive, support and participate within? 

I have already praised many local actors for their work recognising the political ecologies of heat and pollution in Athens. I similarly recognise the important progress being made in developing digital models of the city which can help us to understand its metabolisms. The city was one of the European test sites for the EU-funded DUET project – Digital Urban European Twins – which used cloud computing and open data to explore how digital representations of cities might allow both governments and citizens to test the consequences of different policy decisions. Its stated ambition was explicitly democratic: digital twins as a way of bringing collective intelligence into complex urban decision-making. More recent projects are pushing this further into climate adaptation. EUROASIS, for example, is developing an ‘Urban Heat Digital Twin’ as part of a wider programme combining physical cooling interventions, nature-based systems, climate shelters, heat-health monitoring and new forms of governance. Significantly, the project explicitly commits itself to equity, open science and co-creation.

There is much here that should be supported. But a digital twin is not politically progressive simply because its data concerns the environment. Who owns the data? Who decides what gets measured? Who constructs the model? Which entities become visible within it, and which disappear? Can citizens interrogate its assumptions? Can communities contribute their own sensors and observations? Is it built upon open and interoperable infrastructures, or does the public city become dependent upon proprietary platforms? And, most importantly, who has the capacity to act upon the differences that the model reveals? 

There is a further paradox here. Digital technologies promise to make the environmental metabolism of cities increasingly visible. Citizens are encouraged to contribute data. Sensors measure temperature, humidity and pollution. Satellites observe vegetation and heat. Models combine these observations into increasingly detailed representations of urban systems. Yet the computational cloud through which this information circulates can itself remain remarkably opaque, as can the ownership of the data-centre infrastructure upon which it relies. A genuinely ecological digital twin of Athens would therefore eventually have to model the infrastructure that runs the digital twin of Athens. The observer has to re-enter the system being observed.

A radical eco-cybernetic computational model must include its own electricity, water, heat, minerals, cables, buildings and labour. The informational ecology must reconnect with the material and energetic ecology from which it has never actually been separate. The question is not simply whether the city can sense itself, but whether it can learn – and who constitutes the city that is doing the learning. A digital twin could become an extraordinarily problematic instrument of technocratic management and enclosure. But it could also become an infrastructure of ecological commoning.

Towers of Winds: Listening to Αθηνα

The various questions set out above have informed a design research project that I have been developing with collaborators in Athens and London in recent years: Towers of Winds, which has developed into a wider investigation into the ecological semiotics of heat and pollution, and how we might simultaneously intervene within, sense and learn from the atmospheric systems of the city.

The ecological anthropologist Gregory Bateson suggested that ecologies always have two metabolisms. One concerns thermodynamic bioenergetics: flows of matter and energy. The other is semiotic, and concerns flows of information, signs and pathways of communication. Ecological thinking and action need both. This distinction has become increasingly important to the project. An early component, which I called Listening to Αθηνα, proposed distributing commercial and DIY environmental sensors through schools, balconies, workshops, performances and community collaborations. Networks of temperature, humidity and pollution sensors could contribute to real-time environmental models while simultaneously producing autonomous environmental knowledge for community organisation. Rather than understanding citizens merely as sources of data for a centrally managed smart city, we might understand sensing itself as a form of ecological pedagogy and political organisation: learning collectively to perceive the systems within which we are enfolded.

Alongside this sensing research, we have been actively constructing and testing passive downdraft evaporative cooling structures. The project operates across several possible scales: i. experimental hacking interventions into individual Athenian polykatoikia apartment buildings; ii. approximately five-to-seven-metre community-scale towers; and iii. much larger 25m high municipal emergency infrastructures. We are currently seeking research funding for the largest scale, but have had great success with the smaller towers, often achieving close to 10°C of air cooling, from our first experimental prototypes at Latraac Skate Bar in Keramikos in 2024, to our more recent street tests outside the Rheomode studio in Kypseli this summer. Weather permitting, we will have a workshop to build and test an iteration of this tower at the entrance to the Trans-Europa festival site opening evening.

The principle is simple. Water evaporating inside a tower removes heat from the air. The cooled air becomes denser and falls, producing a self-perpetuating downdraft and creating a cooler public microclimate with very little energy input, just water. Larger versions, according to our models, might potentially draw much cleaner air down from above the worst layers of street pollution levels in Athens, under the right conditions.

These are not proposed as technological solutions to Athens’ environmental problems. They are experiments, and this distinction matters. The environmental crises now confronting us are of such complexity that it is naïve to imagine that we can simply design or plan ‘solutions’. Indeed, thinking in terms of linear problems and solutions is itself part of the meta-problem of human instrumental reason within complex ecological systems. We can steer a boat, but we can’t steer an ocean. 

Equally of course, this is not an argument against planning. Quite the opposite. Learning about and working with and the planetary metabolic rifts driven by capitalism requires urgent yet experimental forms of democratic socio-ecological planning. But the complexity of ecological systems requires us to recognise the limits of prediction and control. We need to test many experimental socio-ecological formations, nurture enclaves of otherness and understand projects as incomplete processes capable of learning and co-evolving. 

A radical Green New Deal therefore requires new public infrastructures, new forms of municipalism, new cooperative institutions, new relationships between researchers and communities, and new practices of maintenance, care, sensing and modelling. It requires not simply a Green New Deal but a Green New Dialogue, and the Towers of Winds project is an attempt to generate learning dialogues with the troubled eco-mental systems of Athena.

Reclaiming the Clouds

In conclusion then, we can say that reclaiming the clouds might mean reclaiming several interconnected commons:

The atmospheric commons: the right to cool, clean and breathable air that cannot be restricted to those able to purchase private climatic bubbles.

The computational commons: recognising data-centres, networks and artificial intelligence as material infrastructures whose energy, water, land and mineral requirements are matters for democratic socio-ecological planning, and are not to be the enclosed reserves and extractive infrastructure of an emergent techno-feudalist elite.

And the informational commons: the sensors, data and models through which cities increasingly perceive themselves, to be held openly enough that inhabitants can participate not merely in being sensed, but can also participate individually and collectively in deciding what should be sensed, how it should be interpreted and what should happen as a consequence.

These aren’t three separate clouds. They are different descriptions of the same ecology. The struggle over them concerns who has the capacity to sense, model and transform the material conditions of collective life. If heat and pollution have become weapons in an atmospheric class war, the task cannot simply be to construct better private defences against them. It is to transform the systems that distribute vulnerability and protection – and to begin demanding and producing the atmospheres, semiotics and infrastructures of the city as a new commons.