Like It or Not: Geo-Engineering in Space May Hold the Keys to Climate Change
11 09 2024 | Jasper Wigley with significant contributions from George Darrah
The next frontier of SpaceTech solutions is fast approaching. In recent months, SpaceX achieved its greatest feat yet, with Starship completing its first successful test launch and soft simulated landing of both stages. The implications of Starship’s nearing commercial readiness should not be understated – not only does it represent a ~7x increase in launch payload from the Falcon 9 (from 22t to 150t), but, crucially, it promises a dramatic reduction in launch costs by up to 10x. Once commercially operational (likely in the next year or two), this will enable a new wave of space-driven technological solutions, ushering in a new frontier of opportunity for our Earth.
Only decades ago, the notion of commercial activity in space remained in the realm of science fiction. With the core purpose of espionage, the first-of-their-kind U.S. CORONA and Russia’s Zenit satellites reached orbit in the early '60s. Capturing photographs from space, these unmanned reconnaissance satellites ejected the film canisters back to Earth, where they descended with the help of heat shields and parachutes. Upon re-entry, the canisters were snagged by intercepting planes, and if all else failed, built-in flotation systems left them open to ocean retrieval. To most at the time, this sounded like pure fantasy, but it was all too real. Fast forward to today’s orbit, and we are monitoring live maps covering the globe with <30 cm resolution via satellites – and it’s only getting wackier.
US Air Force plane looks to catch space-ejected film cannister mid-air
The relationship between AI and space began due to the capability of machine learning to ingest vast geospatial datasets. Even from the early '80s, scientists worked to develop intelligent computer programs to mimic humans in the geospatial data value chain. These two technological threads have been developing in tandem for decades to produce value for tech across climate and defence – particularly in the ever-improving satellite data that provides an understanding of Earth’s surface. Extra-terrestrial film canister ejection and room-sized computers feel a far cry from nanochips and ultra-high-resolution live data feeds.
The progression from AI/geospatial data solutions to complex autonomous hardware in space holds significant implications for climate. Today, the AI Capex boom continues to stoke the race for international data centre dominance, contributing to the fastest yearly growth in electricity demand since 2007, at 4%. In the context of the climate crisis, this trend is concerning, notwithstanding the benefits it will bring. However, as AI applications begin to take shape, all hope is not lost; we may be experiencing a crucial "hummingbird moment." That is, where two elements co-evolve in tandem (like flowers and bees), giving rise to an interlinked yet unexpected third element (such as the biomechanical miracle of the hummingbird). As theorized by author Steven Johnson in How We Got to Now, one example of this phenomenon is the technological breakthroughs following the invention of Gutenberg’s printing press; and how it led to modern understandings of the universe. Although a select number of religious scholars had long been reading with the first iterations of spectacles, the printing press made access to reading near-ubiquitous. The subsequent market for spectacles, as people realized they were long-sighted, facilitated widespread experimentation with lenses. Johnson tracks how the discovery of the telescope and microscope followed shortly after, leading to our "hummingbird moment" with an unexpected yet interlinked outcome.
Geo-engineering from space could be our hummingbird moment. Decades of innovation in satellite launch and AI/ML did not have this use case in mind. However, the cost of launching commercial payloads to low-Earth orbit has decreased by more than 95% since it first became available, and the modern AI revolution is no secret. As climate data collection, modelling, and predictive capabilities regarding climate trends improve, and as launching in-space geo-engineering solutions becomes cost-viable, it may be unwise not to consider such solutions. That reality is imminent. Advanced environmental monitoring is accelerating with each SpaceX and GPT breakthrough, and early experimentation with geo-engineering has already begun.
These solutions will involve thrusting hardware into space with complex automated processes and data feeds at their core. This currently looks to take shape via three fundamental strategies:
Space-Based Solar Shields (Solar Radiation Management - SRM) - The idea behind space-based solar shields is to create a barrier that can partially block or reflect sunlight before it reaches the Earth. These shields would be positioned at the Lagrange Point L1, a location where the gravitational forces between the Earth, Moon, and Sun create a stable position for such structures. This positioning allows the shields to consistently reduce the solar energy impacting the Earth, thereby cooling the planet.
Constructing these solar shields would require breakthroughs in lightweight, highly reflective materials. One possibility is the use of nanomaterials that are both strong and light, such as graphene or carbon nanotubes coated with reflective substances. These materials would need to be engineered to withstand extreme temperature variations, radiation, and micrometeoroid impacts. The shields could be assembled in orbit using automated robotic systems or even 3D printing technology, reducing the need to launch large, pre-assembled structures from Earth. Furthermore, the materials might include adaptive properties, where reflectivity can be dynamically adjusted based on AI’s predictive capacity regarding environmental conditions or energy needs on Earth.
Orbital Aerosol Injection - Orbital aerosol injection would involve releasing particles like sulphur dioxide (SO₂) from satellites into the stratosphere. Satellites equipped with highly precise dispersal mechanisms could be placed in low Earth orbit (LEO) or even ultra-LEO. These particles would act similarly to volcanic ash clouds, which reflect sunlight and cool the Earth's surface. The goal would be to create a fine, even distribution of aerosols that could persist in the stratosphere for years, reducing the Earth's temperature by reflecting incoming solar radiation. Early tests have been conducted via balloons, and it is likely that a combination of Earth- and space-based infrastructure will materialize to manage the required outputs.
Advanced artificial intelligence (AI) and machine learning algorithms would play a crucial role in controlling the dispersal of aerosols. These systems could analyze real-time atmospheric data, adjusting the release of particles to optimize coverage and effectiveness while minimizing potential side effects, such as excessive cooling or disruption of weather patterns. The use of autonomous robotic systems aboard these satellites could facilitate in-orbit maintenance, refueling, or even the deployment of different types of aerosols depending on the season or geographic needs.
Albedo Enhancement of Clouds - Albedo enhancement of clouds involves increasing the reflectivity of low-lying marine clouds by introducing microscopic particles, such as salt, into them. These particles would act as condensation nuclei, making clouds whiter and denser, thus reflecting more sunlight back into space. Satellites could be equipped with spray nozzles designed to release these particles into the atmosphere. Alternatively, drones launched from LEO satellites could descend to lower altitudes to deliver the particles directly into targeted cloud formations. The satellites would monitor cloud coverage and adjust the release of particles to ensure maximum effectiveness.
Many questions remain about who will pay for this, how it could take shape, and whether we even want an "artificial environment." Indeed, many would argue it should be a last resort. However, it may just be a necessary one. If you are building technology in this space, or otherwise have an opinion, please do not hesitate to reach out.