What CLOUD actually tests
Cloud droplets usually need aerosol particles on which water can condense. CERN’s CLOUD chamber can vary trace gases, temperature, humidity and ionisation while holding other conditions tightly controlled. A CERN particle beam can reproduce ionisation comparable with cosmic rays, allowing researchers to ask whether changing ionisation changes aerosol formation. CLOUD therefore tests microscopic atmospheric mechanisms; it does not directly recreate the global climate system.
Cosmic-ray ions really can change particle formation
CLOUD’s 2011 results found ion-enhanced nucleation in some chemical regimes while also showing that sulphuric acid, water, ammonia and ions were insufficient to explain observed boundary-layer nucleation on their own. In 2016, CLOUD showed that highly oxygenated organic vapours from vegetation can form particles without sulphuric acid and that galactic-cosmic-ray ions can increase nucleation by one to two orders of magnitude under those clean biogenic conditions. The mechanism is real, but strongly dependent on atmospheric chemistry.
The difficult step is getting from tiny particles to global clouds
Newly nucleated particles are only a few nanometres across and many disappear before reaching roughly 50–100 nanometres, where they can commonly act as cloud-condensation nuclei. Svensmark and colleagues reported an ion-related aerosol-growth mechanism in controlled experiments, but the measured laboratory growth effect was small under the tested conditions and its global atmospheric magnitude remains uncertain. Field studies using solar-cycle changes and sudden Forbush decreases have produced conflicting results, including studies that found no detectable global cloud response.
The newest 2026 CLOUD result makes the chemistry richer
A June 2026 Nature paper from CLOUD showed that methanesulfonic acid produced from marine phytoplankton chemistry can strongly assist particle nucleation and growth in cool marine air and can increase modelled cloud-condensation-nucleus concentrations, especially in polar regions. Most of those measurements were made under normal galactic-cosmic-ray conditions, and the authors state that the data were insufficient to isolate ion-induced effects at low nucleation rates. The result reinforces CLOUD’s central lesson: cloud-forming particles emerge from interacting chemistry, biology and ionisation, not from one universal cosmic-ray switch.
Research record
CERN — CLOUD experimentOfficial · open sourceKirkby et al. (2011) — Role of sulphuric acid, ammonia and galactic cosmic rays in atmospheric aerosol nucleationPeer-reviewed · open sourceKirkby et al. (2016) — Ion-induced nucleation of pure biogenic particlesPeer-reviewed · open sourceKirkby et al. (2023) — Atmospheric new particle formation from the CERN CLOUD experimentPeer-reviewed synthesis · open sourceSvensmark et al. (2017) — Increased ionization supports growth of aerosols into cloud condensation nucleiPeer-reviewed · open sourceCalogovic et al. (2010) — Sudden cosmic ray decreases: No change of global cloud coverPeer-reviewed · open sourceKulmala et al. (2010) — Atmospheric data over a solar cycle: no connection between galactic cosmic rays and new particle formationPeer-reviewed · open sourceBaalbaki et al. (2026) — Role of methanesulfonic acid in atmospheric particle nucleation and growthPeer-reviewed · open sourceDo Stars Control Earth’s Clouds?
Direct answer: Cosmic rays can affect aerosol nucleation and may assist particle survival towards cloud-condensation-nucleus sizes. That mechanism is experimentally real. What has not been demonstrated is that ordinary solar-driven cosmic-ray changes produce a large global cloud response or dominate recent climate change.
Svensmark’s mechanism deserves serious treatment; the headline conclusion does not. CLOUD supports ion-aerosol physics, not a demonstrated cosmic-ray control knob for Earth’s climate.