Seeding Clouds to Enhance Precipitation: Methods and Effectiveness

Bart Geerts

Dept. of Atmospheric Science

University of Wyoming

cloud seeding artificial snow making

• clouds are seeded with nuclei so that cloud droplets become precipitation-size (rain or snow)

• air needs to be saturated • precip enhancement may occur

many miles downwind • advertent “weather modification”

• water is sprayed under high pressure in cold air

• the water contains ice nuclei (Pseudomonas syringae bacteria) so that tiny droplets freeze immediately upon exiting the nozzle

• air can be dry • snow falls close to source

• clouds consist of tiny droplets, ~100 cm-3, floating with the wind

• most droplets remain liquid below the freezing point (0°C)

• growth by condensation is very slow, usually just more droplets form

• numerous collisions are needed to make a rain drop

• homogeneous nucleation requires cooling to -39°C

• nature has few ice nuclei on which heterogeneous nucleation can occur (~10 m-3 at -10°C, 14°F)

cloud physics 101

sector plates

Rasmussen & Libbrecht, 2003

Rasmussen & Libbrecht, 2003

dendrites

Rasmussen & Libbrecht, 2003

rosettes

supercooled water is ubiquitous in clouds

the Bergeron process

• most precipitation on Earth derives from snow, via the Bergeron process

90% RH

90% RH

90% RH

vapor

100% RH

100% RH

100% RH

two cloud seeding methods

• large, hygroscopic nuclei (e.g. sea salt) act as condensation nuclei that jump-start the growth of drops by collision-coalescence

• commercial operations where clouds are too warm for glaciogenic seeding

• effectiveness not proven

• homogeneous nucleation – dry ice (CO2 at -80°C) – liquid propane

• ice nuclei (usually tiny AgI crystals) – applied pyrotechnically – AgI nucleation temp -6 to -9°C – most commonly used in cold clouds

hygroscopic glaciogenic

Remote-Controlled Ground Generator Demo Unit

Seeding Agent

Propane Burner

Inside the Chamber

Two target clouds for glaciogenic seeding

• thunderstorms: dynamic effect – latent heat release by glaciation

invigorates thunderstorm • also used to mitigate/prevent hail

damage

• orographic clouds: static effect – without ice crystals, clouds would

simply rise over the mountain and evaporate on the lee side

– Bergeron process is jump-started by injection of ice crystals

As air is forced up over a mountain barrier, it cools.

In the lee of the mountains, the air descends & warms.

When the air is moist, condensation occurs, and a cloud of ~20 micron water droplets forms. These droplets are far too small to precipitate.

In the lee, the droplets evaporate.

These droplets remain liquid well below 0°C. Eventually, nature usually introduces ice in these supercooled clouds, by tiny particles called ice nuclei.

Such ice nuclei are rare.

. First Ice

Once ice forms, the particles grow quickly through the Bergeron process, forming flakes that become large enough to precipitate. Snow begins to fall.

. First Snow

The key is to get the snow to fall out upwind of the crest. Snow falling downwind in the descending, drying air will evaporate.

Cloud seeding provides additional ice nuclei that function at warmer temperatures, allowing ice formation to begin sooner. This allows snow to grow sooner, and increase the amount of snow falling upwind of or around the crest.

. First Ice, seeding

. First Ice, not seeded

Most of the precipitation in the

Western USA is orographic.

Water demand is rising as population increases, yet supply may decrease

in a warming climate.

Wintertime precipitation trend over the next 100 years remains highly uncertain

Wintertime precipitation trend over the next 100 years remains highly uncertain

2050 – 2000 wintertime (NDJFMA) precipitation according to the CCSM A2 simulation (Rasmussen et al. 2012)

WY

CO UT

AZ

ID

NM

Reduce consumptive use Increase water production

New Water/Reuse from the System

• Use wasted brackish water • Recycle wastewater • Use desalted ocean or inland

water

Add to Rainfall • Weather modification,

esp. cloud seeding Reduce Evapotranspiration • Vegetation Management

Add to Inflow • Water from coal bed

methane production • Importation alternatives

Reduce Outflow from System

Reduce Loss to Groundwater

Add to Groundwater •Aquifer storage and recovery •Conjunctive use

Concept Examples Water shortage?

National Research Council Report on Weather Modification, 2003

• “Notwithstanding several decades of research, scientific proof of the effectiveness of cloud seeding remains lacking”

• catch-22 paradox: the federal government is not willing to fund research to understand the efficacy of weather modification technologies, yet private interests are willing to spend funds to apply these unproven techniques.

• NRC (2003) advocates a renewed commitment in federal funding to support weather modification research,

• and recommends that the opportunities offered by operational weather modification programs should be capitalized upon.

Weather modification

research

• Meteorology: Taming the sky. Nature, 453, 970-974 (June 2008) • Editorial: Change in the weather. Nature, 453, 957-958.

“Other countries, such as the United States, have simply given up;; the most

promising experiment in America is run not by the federal government but by the state of Wyoming, which is spending $9 million on a seven-year series of cloud-seeding experiments evaluated by experts from the National Center for Atmospheric Research. That's the type of targeted and rigorous study that needs to be done in weather modification, but it took Wyoming to do it.” … ”The stakes are high, as weather modification is one of those areas in which science can have an immediate and obvious benefit for society. It's long past time to invest modest funds in the basic understanding of it.”

Can we trust peer-reviewed publications?

• Science contributor John Bohannon sent a deliberately faked research article 305 times to online journals. More than half the journals that supposedly reviewed the fake paper accepted it. (NPR, 3 Oct 2013)

• Is the analysis objective? If an organization funds a research project that will benefit them financially, then we cannot accept the findings as "evidence" unless different researchers (from unrelated organizations) come to the same conclusions through their own independent research.

• Are the data representative? It is crucial for researchers to ensure that the data they collect is representative of the "true" situation.

• Is the peer review robust? Having your research findings published in a peer-reviewed journal means that other scientists who specialize in that kind of research have verified the quality and validity of the research. It is increasingly difficult for editors to find reviewers willing to make the effort.

Weather modification research

• Hundreds of peer-reviewed papers on weather modification research have been published.

• This all goes back to mid-1940s when scientists at General Electric Research Laboratory in New York demonstrated that dry ice and silver iodide (AgI), which have a similar crystal structure to ice, could initiate ice in a laboratory supercooled liquid cloud. When dry ice pellets or AgI nuclei were released from an aircraft into supercooled stratus clouds, the affected cloud rapidly cleared (Schaefer 1946; Vonnegut 1947).

http://www.gereports.com/thinking-outside-the-cold-box/

Irving Langmuir, Bernard Vonnegut,

and Vincent Schaefer

WWII B-17

The start of glaciogenic seeding

11/24/1948. B-17 flew ~100 m above cloud top. Cloud top temp ~-6°C.

various Intro Meteorology textbooks, e.g. Lutgens and Tarbuck

The start of glaciogenic seeding

http://www.gereports.com/thinking-outside-the-cold-box/

Clearing of this thin cloud layer is obvious, but no in situ or remote sensing probes were available to prove the interpretation that ice crystals formed and consumed the available liquid water.

The start of glaciogenic seeding

Chris Walcek, pers. comm.

The start of glaciogenic seeding

Was glaciogenic seeding responsible for the observed cloud clearing?

Dry ice was crunched into "cm" sized chunks, and dropped from a B-17 porthole, yielding significant "spacing" between individual falling pellets of dry ice Did the resulting ice crystals really grow into snowflakes consuming the SLW? Chris Walcek (Univ. of Albany) believes that the aircraft’‛s downwash could have triggered cloud-top entrainment instability (CTEI)

Was glaciogenic seeding responsible for the observed cloud clearing?

Randall 1980, JAS

cloud-top entrainment instability

Maybe Schaeffer and Vonnegut’‛s cloud clearings were the same as “cloud holes” that are often punched through

supercooled stratus clouds by aircraft?

• Propellers produce significant adiabatic cooling below -39°C, yielding a trail of ice crystals in a supercooled cloud.

Heymsfield et al. (2010 in BAMS; 2011 in Science; 2013)

Can we trust peer-reviewed papers on glaciogenic cloud seeding?

• example: “Radar detection of cloud seeding effects” by Hobbs et al. in Science, 1981.

• The reported seeding rate is just ~0.05 kg/km, implying a spacing between dry ice pellets of ~50 m!

• We have since learned that AgI generators do not produce plumes of enhanced radar reflectivity, at least not in orographic clouds.

• Questions must be asked about persistence, repeatability, and uniqueness.

• In a 2001 opinion letter in BAMS, Hobbs himself is critical about weather mod research and states that the public deserves a more straightforward statement on the uncertainties of cloud seeding to increase precipitation.

• The reality of the 1960s-1980s cloud seeding literature is a monument to subjectivity, lack of peer-review by outsiders, numerous bogus reports, and exaggerated findings.

• This is evidence of short-sightedness, in the long term it became a self-inflicted wound. Federal funding for weather mod research dried up in the late 1980s.

• The fact is the precipitation is an extremely noisy field, and therefore it is difficult to accomplish meaningful control vs. target experiments (Garstang 2005).

Can we trust peer-reviewed papers on glaciogenic cloud seeding?

Is there hope?

• The NRC 2003 reports encourages large randomized experiments with a strong statistical basis, as well as physically-based experiments, to improve our understanding of the changes in cloud microphysical and dynamical processes resulting from cloud seeding.

• Several rigorous statistical studies have come out recently (e.g. Morrison et al. 2009)

• NSF recently funded a study to examine cloud processes in seeded clouds, using new observational tools and high-performance computing.

Divide Peak

Webber Springs Old Battle

Sandstone RS Little Snake

Whiskey Park

Sand Lake

S Brush Cr

N French Cr

Brooklyn Lake

Cinnabar Park

Battle Pass HY47

Hog Park 5502E

Towner Lake GLEES

Douglas Cr

Rob Roy

Upper Cedar Cr

Savery

Saratoga x

0 20 40 km

Target

Target

Control

Control

Wyoming Weather Modification Pilot Project (2006-2012)

Radiometer

Radiometer Sounding Sierra Madre

Medicine Bow

- AgI Generator - SNOTEL

- Project Snowgauge

King Air instruments

FSSP 2DP CIP

PCASP

Icing on the instruments

Shallow orographic cloud over Medicine Bow mountain

Flight pattern over Medicine Bow mountains

- geographically fixed flight legs relative to 3 AgI generators

- 2 full ladders flown before seeding, 2 ladders during seeding

:

strategy to detect impact of AgI seeding: non-simultaneous comparison SEED – NOSEED *

downwind includes natural changes in time

quasi- simultaneous control SEED – NOSEED * upwind

removes these natural changes

* SEED and NOSEED refer to time periods

Why this case?

• Not because it produced much snow … rather, because it was rather simple (“natural laboratory”):

• Temperature was cold (-12 oC at 700 mb) with moderate LWP (~ 0.07 mm)

• No seeder-feeder source aloft

• Not much natural variation in LWP, cloud structure, or precip rate during the flight

Case study – Feb 12, 2013

cloud top mountain top

Pokharel et al. 2013

10 m/s

along-wind leg – Feb 12, 2013

west east 45 km

Medicine Bow mountains rada

r re

flec

tivi

ty

vert

ical

ve

loci

ty

lidar Medicine Bow mountains

liquid water cloud top

Pokharel et al. 2013

Feb 12, 2013

• leg #2 • ~3 km downwind of

AgI generators • 4 passes

• 2 before seeding • 2 during seeding

key strength: vertical perspective gives

reflectivity data very close to ground over complex

terrain Pokharel et al. 2013

Reflectivity FAD Feb 12, 2013

• Cloud is very shallow

• Very light snowfall generated naturally near the surface

• Note shift in near-surface reflectivity to the right during SEED

NOSEED downwind legs

SEED downwind legs

Reflectivity FAD Feb 12, 2013

• Cloud is very shallow

• Very light snowfall generated naturally near the surface

• Note shift in near-surface reflectivity to the right during SEED

NOSEED downwind legs

SEED downwind legs

Reflectivity FAD Feb 12, 2013

• Cloud is very shallow

• Very light snowfall generated naturally near the surface

• Note shift in near-surface reflectivity to the right during SEED

NOSEED downwind legs

SEED downwind legs

SEED - NOSEED

Conclusions

• This is just one case study. Studies such as ASCII should be done more in the context of commercial seeding operations. The observational technology, esp. cloud radar and lidar, are far advanced compared to the weather modification research heydays in the 1960s-1980s.

• High-resolution cloud-resolving, eddy-resolving simulations should be conducted in the context of these observational studies, for model validation of basic dynamics and cloud processes, and to inform about the likely magnitude of the seeding signal on various measurements.

• So there is hope …