Skip to content

Artificial Precipitation, Hail Suppression and Fog Dissipation

CSIR NET Part C meteorology: glaciogenic and hygroscopic cloud seeding, static and dynamic seeding, evaluation, hail suppression and fog clearing.

  • 19 min read
  • Updated
On this page · 16 sections
  1. Prerequisites
  2. The Hook: A Freezer Full of Supercooled Cloud
  3. The Physical Basis: Precipitation Inefficiency
  4. Glaciogenic Seeding
  5. Hygroscopic Seeding (Warm Clouds)
  6. Evaluating Seeding
  7. Hail Suppression
  8. Fog Dissipation
  9. Governance in One Line
  10. Memory Tricks
  11. Don’t Confuse
  12. Common Mistakes
  13. Practice
  14. Quick Revision Table
  15. Key Takeaways
  16. What’s Next

Prerequisites

The Hook: A Freezer Full of Supercooled Cloud

The Physical Basis: Precipitation Inefficiency

Most clouds never precipitate, and those that do convert only a fraction of their condensed water into precipitation reaching the ground. Seeding is worth trying only where a specific microphysical bottleneck limits that conversion:

BottleneckCloud typeSeeding approach
Supercooled water but too few ice crystalssupercooled stratiform and orographic clouds; supercooled cumulusglaciogenic seeding: add ice nuclei or cool the air to make ice
Too many small droplets, narrow spectrumwarm clouds in polluted or continental airhygroscopic seeding: add giant soluble nuclei
Cloud too short-lived or shallowsupercooled cumulus towersdynamic seeding: release latent heat of fusion to boost growth

The seedability concept follows: a cloud is a candidate only if it has supercooled liquid water (or a coalescence bottleneck), a lifetime long enough for seeded particles to grow and fall, and a path for the precipitation to reach the ground in the target area.

Glaciogenic Seeding

Seeding agents

AgentHow it makes iceNotes
Silver iodide (AgI)heterogeneous nucleation: hexagonal lattice close to that of iceactive at a few degrees below 0 °C and more active as it gets colder; burned from solution in ground or aircraft generators, or in pyrotechnic flares
Dry ice (solid CO\(_2\))sublimes at −78.5 °C; cools the air along its path below −40 °C, causing homogeneous nucleationworks in any supercooled cloud regardless of natural INP; dropped as pellets from aircraft
Liquid propaneboils at about −42 °C; expansion and evaporation cool the air enough for homogeneous nucleationused from ground dispensers, especially for supercooled fog
Other organic and inorganic nucleantsheterogeneous nucleationless common

AgI acts at the particle surface, so its effectiveness depends on particle size and formulation; dry ice and propane act by cooling, so they produce very large numbers of crystals at temperatures just below 0 °C, where AgI is weak.

Static seeding: the Bergeron route

Aim: convert supercooled water to precipitation by adding just enough ice crystals for each to grow large at the expense of the droplets.

  • Typical targets are winter orographic clouds (air forced over mountains, giving a steady stream of supercooled water) and supercooled stratiform clouds.
  • Seeding rate matters. Too few crystals: little effect. Over-seeding: so many crystals that the supercooled water is shared among them, each stays small, and the cloud glaciates without precipitating. Over-seeding can reduce precipitation, and the same idea is the basis of hail suppression.
  • Timing and location matter: the crystals need time to grow and fall within the target catchment, so for orographic clouds seeding material is released upwind.

Dynamic seeding: the latent-heat route

Aim: freeze a large fraction of the supercooled water in a cumulus tower so that the latent heat of fusion adds buoyancy, the cloud grows taller and wider, merges with neighbours and processes more water.

The warming from freezing a supercooled liquid water mixing ratio \( q_l \) is

$$\boxed{\Delta T \approx \frac{L_f\,q_l}{c_p}}$$

Freezing 1 g kg\(^{-1}\) gives \( 3.34\times 10^5\times 10^{-3}/1004 \approx 0.33 \) K. Continued deposition of vapour onto the new ice (at the lower ice saturation vapour pressure) adds a little more heat. A few tenths of a kelvin of extra buoyancy can matter in a tower that is barely buoyant near its top.

Dynamic seeding needs massive seeding (many more nuclei than static seeding), and it is a hypothesis about a chain of events (more buoyancy → taller cloud → more inflow → more rain) where each link is uncertain. Its evaluation has been correspondingly harder.

Delivery

MethodUse
Aircraft at cloud baserelease AgI into updrafts that carry it into the supercooled region
Aircraft at cloud topdrop dry ice or ejectable AgI flares directly into the supercooled layer
Ground-based generatorsorographic clouds, where upslope flow lifts the plume into cloud; uncertain how much reaches the target
Rockets and artillery shellscarry nucleant into convective clouds, especially for hail suppression

Hygroscopic Seeding (Warm Clouds)

Aim: start collision–coalescence earlier in clouds whose droplets are too small and uniform.

  • Introduce giant hygroscopic particles (salt such as NaCl or CaCl\(_2\), urea, or salt powders from hygroscopic flares) at cloud base.
  • They activate first and grow into large droplets quickly (low critical supersaturation; see the Köhler table in the nucleation page). These act as collectors and broaden the spectrum.
  • Fine hygroscopic particles may also compete for vapour with the natural, smaller CCN, reducing the number of droplets activated and so making the droplets larger on average (“competition effect”).
  • Water-spray seeding (large water drops released from aircraft) works on the same principle but needs impractically large quantities.

Hygroscopic seeding is the only option for clouds that never reach freezing temperatures, which makes it relevant in the tropics and in Indian monsoon-season experiments.

Evaluating Seeding

The central difficulty: natural rainfall varies by tens of percent from cloud to cloud and season to season, while plausible seeding effects are of the order of ten percent. A seeded cloud that rains proves nothing.

Two kinds of evidence

  1. Physical evidence (the chain of events): aircraft and radar observations showing that seeding material reached the supercooled region, that ice crystal concentrations increased, that precipitation particles grew and fell in the expected place and time. Trace chemistry (for example silver in snow samples downwind of AgI generators) can confirm targeting.
  2. Statistical evidence (the outcome): a demonstration that seeded cases produced more precipitation than comparable unseeded ones.

Experimental designs

DesignIdeaWeakness
Target–control (historical regression)predict target rainfall from a nearby unseeded control area, using a pre-seeding periodrelationship may change; no randomisation
Randomised single-areaeach suitable case is seeded or not at randomneeds many cases because of natural variance
Randomised crossovertwo areas; on each occasion one is chosen at random for seeding, the other acts as controlassumes no contamination between areas
Randomised with blind decisionthe seeding crew or analysts do not know which cases are seededexpensive, but removes bias

For a crossover design, the commonly used test statistic is the double ratio:

$$\boxed{DR = \frac{\bar R_{T,\text{seeded}}/\bar R_{C,\text{seeded}}}{\bar R_{T,\text{unseeded}}/\bar R_{C,\text{unseeded}}}}$$

where \( T \) is the target and \( C \) the control area. \( DR > 1 \) suggests a positive effect; its significance is judged by re-randomisation (permutation) tests, because rainfall is far from normally distributed. The control area removes the effect of a wet or dry spell that affects both areas.

Hail Suppression

How hail forms (recap)

Hail grows in cumulonimbus with strong updrafts, a large supercooled liquid water content and a limited number of embryos (graupel or frozen drops). With few embryos, each collects a lot of water and grows large. See Droplet and Ice-Crystal Growth for wet and dry growth.

Suppression concepts

ConceptIdea
Beneficial competitionseed to create many more embryos; the same supercooled water is shared among them, so each stone is smaller and more likely to melt before reaching the ground
Early rainoutseed early in the cloud’s life to start precipitation before the cloud builds a large supercooled water reservoir in the hail-growth zone
Trajectory loweringmake embryos grow faster lower in the cloud so that they fall out along lower, warmer paths with less time in the hail-growth zone
Complete glaciationfreeze all the supercooled water so no hail can grow; would need impractically large amounts of nucleant
Promotion of coalescencehygroscopic seeding at cloud base to make warm rain that removes water before it is lifted above the freezing level

The key arithmetic of beneficial competition: if a fixed supercooled water mass \( W \) is shared among \( N \) stones, each has mass \( W/N \) and diameter \( \propto N^{-1/3} \). Ten times more embryos reduces diameter by \( 10^{1/3} \approx 2.15 \): a 3 cm stone becomes about 1.4 cm. Smaller stones fall more slowly, spend longer in warm air below the freezing level, and lose more mass to melting.

Methods and evaluation

  • Delivery is by aircraft (base seeding with AgI into feeder updrafts, or cloud-top seeding) or, in some national programmes, by rockets and artillery shells carrying AgI or similar nucleants into the cloud.
  • Success is measured with hail pads (foam panels whose dents record stone sizes and numbers), crop-damage and insurance statistics, and radar reflectivity. All are noisy, and hail is rare at any one spot, so long randomised experiments are needed. Major randomised experiments have produced inconclusive or negative results.
  • Hail cannons (devices that fire acoustic shock waves into the sky) have no physical basis: a pressure wave cannot change the microphysics of a cloud kilometres overhead, and there is no evidence that they work.

Fog Dissipation

Fog is a cloud at the ground (visibility below 1 km), and fog clearance at airports is the most practical form of weather modification. The method depends entirely on the fog’s temperature.

Supercooled fog (below 0 °C): the success story

Seed with dry ice pellets from aircraft or liquid propane from ground dispensers. Large numbers of ice crystals form, grow at the droplets’ expense by the Bergeron–Findeisen process, and fall out as light snow, leaving clear air. Supercooled fog dispersal has been carried out operationally at airports in cold climates, because it is reliable: the physics is simple, the fog is shallow and the seeding is done close to where it is needed.

Warm fog (above 0 °C): the hard problem

Most fog, including the winter radiation fogs of the Indo-Gangetic plain, is warm. No ice-phase trick is available. The options are to evaporate the droplets or to remove them.

MethodHow it worksProblems
Heatingraise the temperature so the air can hold the fog’s liquid water as vapourlarge energy demand; the cleared air is continually replaced by advection; buoyant plumes can carry heat away
Hygroscopic seedingsalt particles take up vapour, lowering the relative humidity; droplets evaporate; large salt droplets fall outcorrosion, environmental effects, effective only on fog that is not continually reforming
Mechanical mixinghelicopter downwash mixes the fog with drier air from aboveworks only if the air above the fog is dry and the fog is shallow; covers a small area
Other experimental methodselectrically charged particles, soundnot operational

Historical heating schemes burned fuel along runways (the wartime British FIDO system) or used jet engines installed beside runways. They worked at great cost; most busy airports with frequent warm fog now rely instead on instrument landing systems that let aircraft land in low visibility.

How much heat does warm fog clearance need?

To clear fog of liquid water content \( W \) (kg m\(^{-3}\)) at temperature \( T \), the air must be warmed by \( \Delta T \) so that its saturation vapour density rises by \( W \), and the latent heat of evaporating \( W \) must also be supplied:

$$\boxed{q = \rho c_p\,\Delta T + L_v W, \qquad \Delta T \approx \frac{W}{d\rho_{vs}/dT}}$$

At 10 °C, the saturation vapour density is about 9.4 g m\(^{-3}\) and increases by about 0.6 g m\(^{-3}\) per kelvin. For \( W = 0.2 \) g m\(^{-3}\):

  • \( \Delta T \approx 0.2/0.6 \approx 0.33 \) K;
  • sensible heat \( \approx 1.25\times 1004\times 0.33 \approx 420 \) J m\(^{-3}\);
  • latent heat \( = 2.5\times 10^6\times 2\times 10^{-4} = 500 \) J m\(^{-3}\);
  • total about 920 J m\(^{-3}\).

A modest temperature rise suffices; the energy is large only because the volume is large and the fog keeps being replaced. Clearing a corridor 2 km long, 100 m wide and 50 m deep (\( 10^7 \) m\(^3\)) once needs about \( 9\times 10^9 \) J, the heat of burning roughly 200 kg of fuel, before any losses.

Ice fog

At very low temperatures (well below −30 °C), water vapour from vehicles, power plants and buildings can form ice fog of tiny crystals. Seeding cannot help because it is already ice; the only remedy is reducing the vapour sources.

Governance in One Line

The international ENMOD Convention prohibits the military or any other hostile use of environmental modification techniques (which would include deliberate weather modification); peaceful weather modification is governed nationally and the WMO publishes guidance on it.

Memory Tricks

Don’t Confuse

ThisvsThat
Static seeding: microphysical, Bergeron route, light seedingdynamic seeding: latent heat of fusion, buoyancy, massive seeding
AgI: heterogeneous nucleus (lattice match)dry ice and propane: cooling to homogeneous nucleation
Glaciogenic: supercooled cloudshygroscopic: warm clouds, giant CCN
Beneficial competition: more embryos, smaller hailover-seeding in rain enhancement: too many crystals, no rain
Supercooled fog: cleared by seedingwarm fog: needs heating, salt or mixing
Physical evidence: chain of events in the cloudstatistical evidence: randomised comparison of outcomes

Common Mistakes

Practice

All questions below are original practice questions written for this page. They are not from any CSIR NET paper.

Level 1: Foundation

  1. Name the seeding agent whose effectiveness rests on its crystal lattice resembling ice.
  2. Which type of seeding targets clouds lying entirely above 0 °C?
  3. Which kind of fog can be cleared reliably by seeding?
Answers
  1. Silver iodide.
  2. Hygroscopic seeding (giant salt nuclei to speed collision–coalescence).
  3. Supercooled fog, seeded with dry ice or liquid propane.

Level 2: Part B Style

Practice Q1. In dynamic seeding, 1.5 g kg\(^{-1}\) of supercooled water is frozen. With \( L_f = 3.34\times 10^5 \) J kg\(^{-1}\) and \( c_p = 1004 \) J kg\(^{-1}\) K\(^{-1}\), the warming is closest to:

(a) 0.05 K (b) 0.5 K (c) 1.5 K (d) 5 K

Answer: (b) 0.5 K

\( \Delta T = L_fq_l/c_p = 3.34\times 10^5\times 1.5\times 10^{-3}/1004 \approx 0.50 \) K.

Practice Q2. Dry ice is effective for seeding supercooled clouds mainly because it:

(a) has a lattice structure similar to ice (b) is strongly hygroscopic (c) cools the surrounding air below the homogeneous nucleation threshold (d) releases latent heat as it sublimes

Answer: (c)

Dry ice sublimes at −78.5 °C and chills the air along its path below about −40 °C, where ice forms homogeneously. (a) describes AgI; (d) is backwards (sublimation absorbs heat).

Practice Q3. The concept of “beneficial competition” in hail suppression aims to:

(a) glaciate the whole cloud (b) increase the number of hail embryos so that each stone grows smaller (c) reduce the updraft speed (d) increase the supercooled liquid water content

Answer: (b)

More embryos share the same supercooled water, so each stone is smaller and more likely to melt before reaching the ground.

Practice Q4. If the number of hail embryos sharing a fixed supercooled water supply increases eightfold, the diameter of each stone changes by a factor of:

(a) 1/8 (b) 1/4 (c) 1/2 (d) \( 1/\sqrt{8} \)

Answer: (c) 1/2

Mass per stone \( \propto 1/N \), diameter \( \propto N^{-1/3} = 8^{-1/3} = 1/2 \).

Level 3: Part C Style

Practice Q5. A randomised crossover seeding experiment over two catchments gives mean rainfall per event (mm):

Target areaControl area
Events when target seeded520400
Events when target not seeded450410

Which statements are correct?

  1. The double ratio is about 1.18.
  2. The simple ratio of seeded to unseeded target rainfall is about 1.16.
  3. The double ratio is larger than the simple ratio because the control area was slightly drier on seeded occasions.
  4. A double ratio above 1 proves that seeding increased rainfall.

(a) 1 and 2 only (b) 1, 2 and 3 only (c) 1, 2, 3 and 4 (d) 3 and 4 only

Answer: (b) 1, 2 and 3 only
  • 1 true: \( (520/400)/(450/410) = 1.300/1.0976 = 1.184 \).
  • 2 true: \( 520/450 = 1.156 \).
  • 3 true: control rainfall was 400 on seeded occasions against 410 on unseeded ones, so after normalising by the control the seeded target looks relatively wetter.
  • 4 false: significance must be tested (for example by re-randomisation), and the physical chain must be consistent. A ratio alone proves nothing.

Practice Q6. Warm fog at 10 °C has liquid water content 0.2 g m\(^{-3}\). The saturation vapour density increases by 0.6 g m\(^{-3}\) per K; take \( \rho c_p = 1255 \) J m\(^{-3}\) K\(^{-1}\) and \( L_v = 2.5\times 10^6 \) J kg\(^{-1}\). Which are correct?

  1. The temperature must rise by about 0.33 K.
  2. The heat required is about 920 J m\(^{-3}\), of which more than half is latent heat.
  3. Clearing a volume of \( 10^7 \) m\(^3\) once needs about \( 9\times 10^{9} \) J.
  4. The same fog at −5 °C could instead be cleared by liquid propane seeding.

(a) 1 and 3 only (b) 1, 2 and 3 only (c) 1, 2, 3 and 4 (d) 2 and 4 only

Answer: (c) 1, 2, 3 and 4
  • 1 true: \( 0.2/0.6 = 0.33 \) K.
  • 2 true: sensible \( 1255\times 0.333 \approx 420 \) J m\(^{-3}\); latent \( 2.5\times 10^6\times 2\times 10^{-4} = 500 \) J m\(^{-3}\); total about 920, latent 54%.
  • 3 true: \( 920\times 10^7 \approx 9.2\times 10^9 \) J.
  • 4 true: at −5 °C the fog is supercooled; propane cooling produces ice crystals, which grow by the Bergeron–Findeisen process and fall out.

Practice Q7. A proposal is made for three seeding projects. Match each with the method that is physically appropriate:

ProjectCloud
Pwinter orographic stratiform cloud, top −15 °C, abundant supercooled water, few ice crystals
Qmonsoon-season cumulus, top +6 °C, 800 droplets cm\(^{-3}\), narrow spectrum
Risolated supercooled cumulus towers, tops −10 °C, barely buoyant near the top

(a) P: hygroscopic; Q: static glaciogenic; R: dynamic glaciogenic (b) P: static glaciogenic; Q: hygroscopic; R: dynamic glaciogenic (c) P: dynamic glaciogenic; Q: hygroscopic; R: static glaciogenic (d) P: static glaciogenic; Q: dynamic glaciogenic; R: hygroscopic

Answer: (b)
  • P has the classic static-seeding signature: supercooled water and a shortage of ice in a long-lived orographic flow.
  • Q is warm throughout with a coalescence bottleneck: only hygroscopic seeding can act.
  • R: freezing the supercooled water releases latent heat of fusion that could tip marginal buoyancy, the dynamic-seeding hypothesis.

Quick Revision Table

IdeaFact
Basisseeding improves precipitation efficiency; it cannot create water
Glaciogenic agentsAgI (lattice match); dry ice (−78.5 °C); liquid propane (about −42 °C)
Static seedingfew crystals, Bergeron growth; orographic and stratiform clouds
Dynamic seedingmassive seeding; \( \Delta T = L_fq_l/c_p \) (0.33 K per g kg\(^{-1}\))
Over-seedingmany small crystals, less precipitation
Hygroscopic seedinggiant salt nuclei for warm clouds; flares at cloud base
Evaluationrandomised designs; crossover double ratio; physical chain of evidence
Strongest evidencesupercooled orographic cloud seeding; supercooled fog clearing
Hail suppressionbeneficial competition (\( d \propto N^{-1/3} \)), early rainout, trajectory lowering; hail cannons have no basis
Fogsupercooled: dry ice or propane; warm: heating, hygroscopic, mixing; ice fog: source control
Warm fog heat\( q = \rho c_p\Delta T + L_vW \); about 920 J m\(^{-3}\) for 0.2 g m\(^{-3}\) at 10 °C
IndiaIITM’s CAIPEEX research programme

Key Takeaways

What’s Next

Advertisement
Advertisement