Warming the entire surface of Mars by 35 degrees would require releasing engineered particles into its atmosphere at a sustained rate of 60 litres per second, produced from minerals already present in Martian soil, for as long as the warming is maintained. That figure comes from a 61-page non-peer-reviewed research roadmap submitted to EarthArXiv in May 2026 by scientists from the University of Chicago, Harvard, MIT, JPL, Northwestern, Stanford, and twelve other institutions. The paper does not argue that warming Mars is a good idea. It sets out, in explicit technical and financial detail, what it would take to find out whether it is even possible.
The starting point is a physical fact about Mars that is easy to state and hard to fix. Mars sits at a stable average temperature of 210 Kelvin, roughly minus 63 degrees Celsius. That temperature is maintained by a precise balance between absorbed sunlight and heat radiated back into space: 1.6 times 10 to the power of 16 watts in and out, holding constant. The planet carries essentially no ocean and only a thin atmosphere, which means it responds to changes in energy input within less than a year, faster than any ocean world. Warming it requires either reflecting more sunlight onto the surface or trapping more of the heat that is already there. The roadmap considers three practical ways to do this, each suited to a different scale of warming and a different phase of human activity on Mars.
The most near-term method involves solid-state greenhouse membranes, physical materials stretched over patches of Martian soil to trap heat the way a greenhouse does on Earth. Aerogel, a material already used for thermal control on Mars rovers, achieves warming of more than 60 Kelvin at a thickness of around 3 centimetres. At that density, 1 square kilometre of aerogel coverage would require 3 million kilograms of material shipped from Earth, costing approximately 6 billion dollars in transport costs before manufacturing is even counted. The roadmap’s proposed solution is to manufacture aerogel on Mars from local resources using solar power, at an estimated energy cost of 30 megajoules per kilogram. A 1-megawatt solar installation coupled to an aerogel production plant could extend the warmed surface area at a rate of 0.3 square kilometres per year. The immediate, practical purpose of this is not planetary transformation but water. A warmed patch of Martian ice sublimates water vapour that can then be captured and condensed. Half a hectare of membrane could produce enough water to fuel a single Starship lander.
The second method involves orbiting solar reflectors, space mirrors in sun-synchronous polar orbit approximately 750 kilometres above Mars, redirecting sunlight onto specific surface targets. A single square kilometre of reflector in that orbit can deliver an average of 0.6 megawatts to a 1-kilometre-radius patch of ground at 40 degrees north latitude. Doubling the sunlight reaching a human base in a useful way would require roughly 750 square kilometres of reflector surface, at a combined mass of between 7,500 and 75,000 tonnes depending on how thin the sail material can be made, and a total cost under current assumptions of around 3.5 billion dollars. The paper notes that the reflectors would not need a rocket burn to reach Mars: solar sails can propel themselves from Earth orbit to Mars orbit using radiation pressure alone, a mechanism already demonstrated by JAXA’s IKAROS spacecraft, removing the cost of the Trans-Mars Injection burn and atmospheric entry. The biggest engineering obstacle is not propulsion but mass. Current state-of-the-art solar sails are constructed at around 60 grams per square metre. The roadmap calculates that Mars-scale reflectors would need to come in at 20 grams or below.
The third method is the one capable of warming the entire planet, and it is the most uncertain. Engineered aerosols, purpose-built particles released into Mars’s atmosphere, can trap outgoing infrared radiation by resonating with the specific wavelengths at which Mars radiates heat to space. Unlike natural dust, which cools the dayside by reflecting incoming sunlight, these particles are designed to absorb thermal infrared while leaving visible wavelengths largely unaffected. In climate models, 2 milligrams of conductive particles per square metre of atmospheric column doubles the strength of Mars’s greenhouse effect, producing around 4 Kelvin of warming. Scaling to 35 Kelvin of warming would require 60 litres per second of particle production from an on-Mars factory, for the duration of the warming campaign. The power requirement to warm Mars by 4 Kelvin sits at around 500 megawatts; warming by 35 Kelvin would require approximately 5 gigawatts, generated from infrastructure built or manufactured on the Martian surface.
Two feedstock pathways exist for these particles, and both are under active laboratory development. The first uses Mars’s atmosphere as raw material. Mars air is 95 percent carbon dioxide, meaning one in three atmospheric atoms is carbon. The MOXIE experiment, carried on NASA’s Perseverance rover, demonstrated that carbon dioxide can be electrolysed into carbon monoxide and oxygen on Mars. A subsequent chemical reaction, carbon monoxide combining with itself to produce carbon dioxide plus solid carbon, yields carbon particles including graphene. Laboratory experiments at Northwestern University have already fabricated test batches of magnesium nanoribbons and confirmed their infrared absorption properties match theoretical predictions using Fourier Transform Infrared spectroscopy. The second feedstock pathway uses metals extracted from Martian soil and rock, particularly magnesium sulphate salts found at concentrations of around 20 percent by weight in salt-rich rock deposits whose total estimated mass is 10 to the power of 18 kilograms.
The largest single uncertainty across all aerosol approaches is how long the particles survive in the Martian atmosphere before settling out or clumping together into masses too large to stay suspended. The required minimum lifetime is one-third of a year, roughly 120 days. On Earth, particles in the lower atmosphere last around one week and those in the upper stratosphere last about one year. Mars’s global dust storms decay over roughly 90 days. No Mars-specific measurement exists for engineered aerosols. The paper identifies this as the single most critical gap in current knowledge, and the research that most directly addresses it requires only a Mars environmental chamber, a wind tunnel, and laboratory clumping tests, at a combined cost the roadmap places below $300,000 for the initial phase.
The cost structure for the full programme scales in stages. Near-term on-Earth laboratory testing across all three warming approaches sits in the range of tens of thousands to several million dollars per experiment, with gating criteria specified in advance: if a test fails its benchmark, the pathway is abandoned before the next stage of investment. Mars-based process experiments, operating as secondary payloads on commercial cargo missions beginning in the 2030 to 2031 launch window, are estimated at tens of millions of dollars each. A pilot aerosol factory on Mars, deployed after human landing, carries a cost the authors describe as unknown but comparable to major interplanetary hardware programmes. The shipping cost to Mars is assumed at $2,000 per kilogram, well below current advertised prices of around $100,000 per kilogram but achievable if in-space propellant transfer and full vehicle reuse come into operation. Every analysis in the roadmap treats the future of launch costs as the dominant external variable: if prices do not fall, the economics of any large-scale warming programme shift substantially toward on-Mars production of every component possible.
The roadmap explicitly addresses the ethics of the proposal before addressing the engineering. It notes that no consensus exists on whether humans should extend life beyond Earth, that the search for existing Martian life should take priority over any warming programme, and that three principles should guide research: establishing whether Mars is already lifeless before making it more habitable, preserving the geological and climate record Mars currently holds, and ensuring that no warming method irreversibly consumes resources that future generations might need. The Outer Space Treaty, ratified by all spacefaring nations, requires that Mars exploration serve the interests of all humanity. Regional or larger-scale warming would almost certainly require international agreement among spacefaring nations before proceeding.
The paper notes that even under optimistic assumptions, warming at kilometre scale is at minimum a decade away from any test, and wider environmental modification would require sustained investment across many decades beyond that. As of May 2026, total power infrastructure at Mars’s surface sits below 1 kilowatt. The roadmap remains a non-peer-reviewed preprint and has not yet undergone formal peer review.
Kite, E.S., Essunfeld, A., Hecht, M.H., Mischna, M.A., Wordsworth, R., Mohseni, H., Boies, A., Averesch, N., Ansari, S., Richardson, M.I., DeBenedictis, E.A., Stork, D., Bamba, A.L., Handmer, C.J., Jourdain, C., Ramirez, R., Mason, C.E., Kling, A., Braude, A.S., Dumitrescu, A., Worden, S.P., Cumbers, J., Lanza, N., Quayum, R., & Cockell, C.S. (2026). A research roadmap for assessing the feasibility of warming Mars. EarthArXiv preprint. https://eartharxiv.org/repository/view/9568/
