Cloud Seeding and Weather Modification

Cloud seeding sits in an odd corner of the UPSC syllabus. It is technology, not pollution, yet Shankar IAS files it under a chapter titled "Environmental Pollution" because the whole point of the technique, in its most publicised Indian use, is pollution control: force a cloud to rain and hope the rain scrubs particulate matter out of the air. This note treats it as what it really is, a genuine atmospheric-science intervention with named Indian field programmes, real efficacy debates, and a broader family of geoengineering cousins, rather than as a vague current-affairs buzzword.

What cloud seeding actually is

Cloud seeding is the deliberate introduction of a nucleating agent into an existing cloud to encourage or accelerate precipitation. The operative word is existing: seeding cannot conjure a cloud out of dry air. It can only nudge a cloud that already has enough moisture and instability to precipitate on its own, but is failing to do so efficiently, into releasing that moisture sooner or in greater volume. The World Meteorological Organization is blunt about this limit in its official statement on weather modification: the energy locked up in weather systems is so large that no seeding technique can manufacture a rain-bearing cloud system where none existed. What seeding can do, under the right conditions, is add artificial nuclei that clouds otherwise lack, making the two natural precipitation-formation pathways run faster.

Those two pathways matter because they set up the entire cold cloud versus warm cloud distinction that UPSC likes to test. In a "warm" cloud, one entirely above freezing, raindrops form purely through collision and coalescence: tiny water droplets condensed on naturally occurring aerosol particles collide, merge, and grow large enough to fall as rain. In a "cold" cloud, one with a supercooled liquid-water layer (water that stays liquid below 0°C because it lacks a surface to freeze onto) alongside ice crystals, the dominant mechanism is the Wegener-Bergeron-Findeisen process. Ice has a lower saturation vapour pressure than liquid water at the same sub-zero temperature, so in a mixed cloud, water vapour preferentially deposits onto ice crystals while the surrounding supercooled droplets evaporate to feed that deposition. The ice crystals grow rapidly at the expense of the liquid droplets, eventually becoming heavy enough to fall, melting into rain (or arriving as snow) on the way down. Cloud seeding exploits each pathway differently, and that difference is the basis for the two seeding families.

Cold cloud (glaciogenic) seeding

Glaciogenic seeding targets cold, mixed-phase clouds and works by supplying artificial ice-nucleating particles so the Bergeron-Findeisen process can start (or intensify) well before enough natural ice nuclei would have appeared on their own. The agent that dominates real-world operations worldwide is silver iodide. Its crystal lattice is hexagonal and geometrically very close to the lattice of ordinary ice, so supercooled droplets can nucleate onto silver iodide particles at temperatures where they would otherwise stay liquid for lack of a suitable surface. This is heterogeneous nucleation: the silver iodide particle acts as a substrate, not as an ingredient that becomes part of the raindrop. Potassium iodide has occasionally been used as an alternative ice-nucleating salt, though it is far less common in current operational programmes than silver iodide.

Solid carbon dioxide, dry ice, works through an entirely different mechanism and is worth distinguishing carefully because UPSC-style questions like to blur the two. Dry ice does not provide a crystal surface for droplets to nucleate on. Instead, pellets of dry ice sublimate at around minus 78°C, and that intense localised cooling can drive the surrounding air below the threshold at which ice crystals form spontaneously, without needing any foreign nucleus at all (homogeneous nucleation). So silver iodide seeds by mimicking ice's shape; dry ice seeds by forcing the air cold enough that the cloud freezes on its own. Both are cold cloud techniques, but the underlying physics is not the same, and conflating them is a realistic exam trap.

Warm cloud (hygroscopic) seeding

Hygroscopic seeding targets clouds that never get cold enough for the ice-crystal pathway, working the collision-coalescence process instead. Fine particles of a water-attracting (hygroscopic) salt, commonly sodium chloride, sometimes calcium chloride, are dispersed into the lower part of a growing cumulus cloud. Because these salts readily absorb water vapour, they act as unusually effective condensation nuclei, growing into comparatively large droplets faster than the cloud's natural aerosol population would. Introducing a spread of droplet sizes, rather than the more uniform droplet population a cloud develops on its own, increases the rate at which droplets collide and merge, effectively fast-forwarding the natural rain-formation process and widening the size distribution needed for efficient coalescence. Hygroscopic seeding also has a secondary, more physical effect: the extra latent heat released as the seeded droplets condense water can strengthen the cloud's internal updraught, drawing in more moisture and invigorating the whole convective cell, not just the seeded droplets themselves.

Both delivery routes rely on getting the agent physically into the cloud, which in practice means one of two methods: burning silver iodide or hygroscopic flares mounted on an aircraft's wings as it flies through or beneath the cloud base, or firing seeding rockets or ground generators that release plumes upward into orographic (mountain-formed) or convective clouds from below. Aircraft-based flare seeding is what India's own field programmes, and the 2025 Delhi trial, have both used.

India's field programmes: CAIPEEX and the state trials

India's cloud seeding research has a genuine institutional home: the Indian Institute of Tropical Meteorology (IITM), under the Ministry of Earth Sciences, which has run the Cloud Aerosol Interaction and Precipitation Enhancement Experiment (CAIPEEX) since 2009. CAIPEEX is not itself an operational rainfall programme; it is the underlying research effort, using instrumented research aircraft flying through Indian monsoon and pre-monsoon clouds to measure how aerosol concentration, cloud droplet number, and black carbon loading affect a cloud's natural precipitation efficiency, and to build the scientific basis on which any operational seeding could be justified. Its early phases (2009, 2010-11) mapped background aerosol and cloud microphysics across the country from bases including Hyderabad; a later phase moved into the Ganga basin and the Western Ghats and included actual seeding trials, releasing hygroscopic and glaciogenic flares into monsoon clouds over Maharashtra to directly test the enhancement techniques CAIPEEX had spent years characterising.

The clearest operational, state-run programme to come out of this research base is Karnataka's Varshadhare, run in 2017 by the state's Rural Water Supply and Sanitation Department at a cost of roughly Rs 35 crore, explicitly following the seeding protocols CAIPEEX had developed. It ran for about eleven weeks (21 August to 7 November 2017), used a network of ground radars plus more than six thousand rain gauges across the state to identify moisture-stressed districts and measure the outcome, and combined both hygroscopic and glaciogenic seeding depending on cloud type. A peer-reviewed analysis of the programme's 618 recorded seeding events, comparing rainfall in the four hours after seeding against an estimated natural baseline, found an average enhancement of roughly 28% over what the clouds would likely have produced unseeded, among the more scientifically rigorous real-world results any Indian cloud seeding effort has produced. Telangana and undivided Andhra Pradesh have also experimented with seeding during earlier drought years, though on a smaller and less continuously documented scale than Karnataka's.

The programme UPSC's 2025 question is really pointing at is Delhi's, which reframed cloud seeding around pollution rather than drought. In 2025 the Delhi government signed an agreement with IIT Kanpur to trial seeding as a way of washing particulate matter out of the capital's winter air, and after securing the required civil-aviation clearance flew a single-engine aircraft over northwest Delhi and parts of the NCR, releasing flares carrying silver iodide and sodium chloride compounds over areas including Burari and Mayur Vihar. The results were modest and contested: IIT Kanpur reported no measurable rainfall, citing atmospheric moisture as low as 15 to 20 per cent against the roughly 50 per cent effective seeding typically needs, but pointed to a measurable dip in PM2.5 readings at some monitoring stations afterward, while a separate IIT Delhi assessment concluded that Delhi's winter atmosphere typically lacks the moisture and vertical lift a cloud needs for seeding to work at all, and argued the technique should be treated as an occasional emergency measure rather than a routine pollution-control tool.

The efficacy and cost-effectiveness debate

That Delhi disagreement is not an isolated spat; it reflects a genuine, long-running scientific argument that UPSC questions on this topic are built to probe. The core measurement problem is the counterfactual: no one can rerun the same cloud twice, once seeded and once not, so every claimed enhancement percentage rests on a statistical estimate of what that cloud would probably have produced unseeded, not a direct before-and-after comparison. This is exactly why credible studies, the Karnataka one included, lean on large sample sizes, radar tracking, and dense rain-gauge networks rather than a handful of anecdotal seeded storms.

The WMO's own statement on weather modification is a useful, honest scorecard here rather than a blanket endorsement or dismissal. It singles out glaciogenic seeding of orographic (mountain-forced) supercooled clouds as the technique with the strongest combined statistical and physical evidence behind it, describes hygroscopic seeding of convective clouds as producing inconsistent results because of how naturally variable those clouds already are, and calls evidence for hail suppression by seeding still insufficient. It also records that operational fog dispersion, rainfall, snowfall and hail programmes are now running in more than fifty countries, so the technology is mainstream even where the science of any single programme's outcome remains genuinely disputed.

Cost-effectiveness compounds the uncertainty. A seeding campaign involves aircraft time, trained crew, flares or seeding agents, and a monitoring network sufficient to even attempt measuring the outcome, all spent against a probabilistic, not guaranteed, gain in rainfall or a temporary, easily-reversed dip in pollution. Down To Earth's coverage of the Delhi trials makes the comparison directly: a similar pollution-focused seeding effort in Lahore briefly pulled the city's air quality index down from above 300 to around 189, but the improvement lasted only a couple of days before pollution levels climbed back, which is the standard critique levelled at seeding as a pollution fix: it treats a symptom for a few hours, not the emission sources that caused it, and diverts attention and budget from the harder structural fixes.

The wider geoengineering family

Cloud seeding is the most mature member of a broader category of technological interventions in atmospheric management, and UPSC's placement of this topic under a "pollution control interventions" framing is really an invitation to know that broader category, not just this one technique. The most discussed cousin is marine cloud brightening, a proposed form of solar radiation management, the category of climate geoengineering aimed at reflecting incoming sunlight rather than removing greenhouse gases. The idea is to spray fine seawater droplets from vessels over the ocean, seeding low marine stratocumulus clouds with extra sea-salt particles so the clouds develop more, smaller droplets. A cloud with more numerous, smaller droplets scatters sunlight more effectively and appears brighter, so in principle it reflects a larger share of incoming solar radiation back to space, cooling the ocean and atmosphere beneath it. It remains firmly experimental. Climate models suggest it could, in theory, offset a meaningful share of global warming, but researchers and governance bodies alike flag it as high-risk precisely because it is deployable at scale with comparatively simple equipment, raising the same counterfactual measurement problem as cloud seeding at a planetary scale, plus the added danger that one country or actor could alter regional rainfall patterns for everyone downwind without global agreement on whether, or how much, to intervene.

That downwind concern is the deeper governance thread running under all weather and climate modification, cloud seeding very much included. Rain redirected toward one district's clouds is, definitionally, moisture that might otherwise have reached a neighbouring one, so any seeding programme carries an inherent inter-jurisdictional fairness question even before its environmental effects (silver iodide's buildup in soil and water, though studies so far have found no strong evidence of ecological harm at the concentrations actually used) are weighed. India currently has no dedicated statute governing weather modification. What exists is a narrower aviation-safety gate: any seeding flight needs clearance from the Directorate General of Civil Aviation as an aerial-work operation, which is how the Delhi trial was authorised, alongside sign-off from coordinating bodies such as the India Meteorological Department, but this covers flight safety, not the substantive question of who may seed clouds, under what ecological or inter-state review, and with what accountability for the outcome. The WMO's own guidance for member states is procedural rather than prescriptive: obtain expert technical advice before any new programme, subject operational programmes to periodic (ideally annual) independent review, and keep WMO itself informed of ongoing weather modification activity worldwide, a light-touch international norm that several commentators argue India, and most seeding-active countries, have yet to translate into binding domestic rules.

The exam angle

UPSC's question so far has tested cloud seeding at the most literal level, which chemicals go into the cloud, but the syllabus placement and the live governance debate both point toward richer traps ahead. Expect statement-based questions built around exactly the distinctions this note has drawn out: cold cloud versus warm cloud seeding, silver iodide's role as an ice-mimicking nucleation surface versus dry ice's role as a pure coolant that needs no foreign surface at all, and glaciogenic versus hygroscopic agents matched to the wrong cloud type. A second likely trap is treating "cloud seeding" and "geoengineering" as synonyms; examiners could easily pair cloud seeding (a weather-modification technique, local and short-lived in its effect) against marine cloud brightening or stratospheric aerosol injection (climate-geoengineering techniques aimed at global temperature, not a single storm) and ask which is which. A third is the programme-name trap: expect Karnataka's Varshadhare, the national CAIPEEX research effort, and Delhi's IIT Kanpur pollution trial to be scrambled together in a matching question, since they differ in agency, objective (drought relief versus pollution control versus pure research), and scale. Finally, watch for a statement testing whether cloud seeding can "create" rain from a clear sky. It cannot; that a suitable cloud must already exist is the single fact the WMO itself is most emphatic about, and it is the easiest correct-looking wrong statement an examiner can write.

Quick revision points

• Cloud seeding introduces a nucleating agent into an existing cloud to accelerate precipitation; it cannot generate rain-bearing clouds where none exist.
• Cold cloud (glaciogenic) seeding uses silver iodide, whose hexagonal lattice mimics ice, to trigger heterogeneous ice nucleation and drive the Wegener-Bergeron-Findeisen process; potassium iodide is a rarer alternative agent.
• Dry ice (solid CO2) seeds cold clouds differently: it cools the air enough for spontaneous (homogeneous) ice formation, without acting as a nucleation surface itself.
• Warm cloud (hygroscopic) seeding uses water-attracting salts, chiefly sodium chloride, to widen droplet size distribution and speed up collision-coalescence.
• India's research base is IITM's CAIPEEX programme (from 2009); the clearest operational state programme is Karnataka's Varshadhare (2017, roughly 28% measured rainfall enhancement); Delhi's 2025 IIT Kanpur trial targeted air pollution, not drought, with contested results.
• Efficacy is genuinely disputed: WMO rates glaciogenic seeding of orographic clouds the best-evidenced method, hygroscopic convective seeding as inconsistent, and hail suppression as still unproven.
• Marine cloud brightening is a related but distinct technique, solar radiation management aimed at global cooling by brightening ocean clouds, not local rainfall.
• India has no dedicated weather-modification law; seeding flights need DGCA aerial-work clearance, and WMO's own guidance (expert review, periodic evaluation, reporting to WMO) remains non-binding.

Work through the linked question below to see how UPSC turns the seeding-agent detail into a precise, factual trap, then use these distinctions to anticipate where a statement-based or matching version of the same topic could go next.