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
On this page · 16 sections
- Prerequisites
- The Hook: A Freezer Full of Supercooled Cloud
- The Physical Basis: Precipitation Inefficiency
- Glaciogenic Seeding
- Hygroscopic Seeding (Warm Clouds)
- Evaluating Seeding
- Hail Suppression
- Fog Dissipation
- Governance in One Line
- Memory Tricks
- Don’t Confuse
- Common Mistakes
- Practice
- Quick Revision Table
- Key Takeaways
- What’s Next
Prerequisites
- Ice nuclei, supercooling and CCN from Cloud Classification and Condensation Nuclei.
- The Bergeron–Findeisen and coalescence processes, and hail growth, from Droplet and Ice-Crystal Growth. Every seeding method is an attempt to switch one of them on or off.
- Buoyancy and CAPE from Vertical Stability and Tropical Convection, for dynamic seeding.
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:
| Bottleneck | Cloud type | Seeding approach |
|---|---|---|
| Supercooled water but too few ice crystals | supercooled stratiform and orographic clouds; supercooled cumulus | glaciogenic seeding: add ice nuclei or cool the air to make ice |
| Too many small droplets, narrow spectrum | warm clouds in polluted or continental air | hygroscopic seeding: add giant soluble nuclei |
| Cloud too short-lived or shallow | supercooled cumulus towers | dynamic 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
| Agent | How it makes ice | Notes |
|---|---|---|
| Silver iodide (AgI) | heterogeneous nucleation: hexagonal lattice close to that of ice | active 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 nucleation | works in any supercooled cloud regardless of natural INP; dropped as pellets from aircraft |
| Liquid propane | boils at about −42 °C; expansion and evaporation cool the air enough for homogeneous nucleation | used from ground dispensers, especially for supercooled fog |
| Other organic and inorganic nucleants | heterogeneous nucleation | less 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
| Method | Use |
|---|---|
| Aircraft at cloud base | release AgI into updrafts that carry it into the supercooled region |
| Aircraft at cloud top | drop dry ice or ejectable AgI flares directly into the supercooled layer |
| Ground-based generators | orographic clouds, where upslope flow lifts the plume into cloud; uncertain how much reaches the target |
| Rockets and artillery shells | carry 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
- 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.
- Statistical evidence (the outcome): a demonstration that seeded cases produced more precipitation than comparable unseeded ones.
Experimental designs
| Design | Idea | Weakness |
|---|---|---|
| Target–control (historical regression) | predict target rainfall from a nearby unseeded control area, using a pre-seeding period | relationship may change; no randomisation |
| Randomised single-area | each suitable case is seeded or not at random | needs many cases because of natural variance |
| Randomised crossover | two areas; on each occasion one is chosen at random for seeding, the other acts as control | assumes no contamination between areas |
| Randomised with blind decision | the seeding crew or analysts do not know which cases are seeded | expensive, 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
| Concept | Idea |
|---|---|
| Beneficial competition | seed 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 rainout | seed early in the cloud’s life to start precipitation before the cloud builds a large supercooled water reservoir in the hail-growth zone |
| Trajectory lowering | make 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 glaciation | freeze all the supercooled water so no hail can grow; would need impractically large amounts of nucleant |
| Promotion of coalescence | hygroscopic 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.
| Method | How it works | Problems |
|---|---|---|
| Heating | raise the temperature so the air can hold the fog’s liquid water as vapour | large energy demand; the cleared air is continually replaced by advection; buoyant plumes can carry heat away |
| Hygroscopic seeding | salt particles take up vapour, lowering the relative humidity; droplets evaporate; large salt droplets fall out | corrosion, environmental effects, effective only on fog that is not continually reforming |
| Mechanical mixing | helicopter downwash mixes the fog with drier air from above | works only if the air above the fog is dry and the fog is shallow; covers a small area |
| Other experimental methods | electrically charged particles, sound | not 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
| This | vs | That |
|---|---|---|
| Static seeding: microphysical, Bergeron route, light seeding | dynamic seeding: latent heat of fusion, buoyancy, massive seeding | |
| AgI: heterogeneous nucleus (lattice match) | dry ice and propane: cooling to homogeneous nucleation | |
| Glaciogenic: supercooled clouds | hygroscopic: warm clouds, giant CCN | |
| Beneficial competition: more embryos, smaller hail | over-seeding in rain enhancement: too many crystals, no rain | |
| Supercooled fog: cleared by seeding | warm fog: needs heating, salt or mixing | |
| Physical evidence: chain of events in the cloud | statistical 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
- Name the seeding agent whose effectiveness rests on its crystal lattice resembling ice.
- Which type of seeding targets clouds lying entirely above 0 °C?
- Which kind of fog can be cleared reliably by seeding?
Answers
- Silver iodide.
- Hygroscopic seeding (giant salt nuclei to speed collision–coalescence).
- 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 area | Control area | |
|---|---|---|
| Events when target seeded | 520 | 400 |
| Events when target not seeded | 450 | 410 |
Which statements are correct?
- The double ratio is about 1.18.
- The simple ratio of seeded to unseeded target rainfall is about 1.16.
- The double ratio is larger than the simple ratio because the control area was slightly drier on seeded occasions.
- 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?
- The temperature must rise by about 0.33 K.
- The heat required is about 920 J m\(^{-3}\), of which more than half is latent heat.
- Clearing a volume of \( 10^7 \) m\(^3\) once needs about \( 9\times 10^{9} \) J.
- 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:
| Project | Cloud |
|---|---|
| P | winter orographic stratiform cloud, top −15 °C, abundant supercooled water, few ice crystals |
| Q | monsoon-season cumulus, top +6 °C, 800 droplets cm\(^{-3}\), narrow spectrum |
| R | isolated 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
| Idea | Fact |
|---|---|
| Basis | seeding improves precipitation efficiency; it cannot create water |
| Glaciogenic agents | AgI (lattice match); dry ice (−78.5 °C); liquid propane (about −42 °C) |
| Static seeding | few crystals, Bergeron growth; orographic and stratiform clouds |
| Dynamic seeding | massive seeding; \( \Delta T = L_fq_l/c_p \) (0.33 K per g kg\(^{-1}\)) |
| Over-seeding | many small crystals, less precipitation |
| Hygroscopic seeding | giant salt nuclei for warm clouds; flares at cloud base |
| Evaluation | randomised designs; crossover double ratio; physical chain of evidence |
| Strongest evidence | supercooled orographic cloud seeding; supercooled fog clearing |
| Hail suppression | beneficial competition (\( d \propto N^{-1/3} \)), early rainout, trajectory lowering; hail cannons have no basis |
| Fog | supercooled: 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 |
| India | IITM’s CAIPEEX research programme |
Key Takeaways
What’s Next
- Weather Radar and Rainfall Measurement: the tools used to observe seeded clouds and measure the outcome.
- Droplet and Ice-Crystal Growth: revisit the processes that seeding manipulates.
- Climatology: Humidity, Precipitation and Water Balance: where precipitation fits in the regional water budget.
- Back to the chapter overview.