RAINFALL ENHANCEMENT PROGRAMS AND NEED FOR TRAINING. SAHEL Conference April 2007 CILSS Ouagadougou, Burkina Faso

Transcription

1 RAINFALL ENHANCEMENT PROGRAMS AND NEED FOR TRAINING SAHEL Conference April 2007 CILSS Ouagadougou, Burkina Faso

2 Fundamentals What are clouds made of?

3 Cloud droplets Condensation on hygroscopic particles (a specific kind of water-soluble aerosol) Droplet growth by condensation Condensation occurs through diffusion of water vapor Cloud droplets are typically in the size range of 5-25 μm (i.e., very small) They have a very small terminal velocity Result: No precipitation from only cloud droplets. The water stays in the cloud

4 Warm rain process Collision and coalescence of droplets falling at different terminal velocities leads to raindrop formation Raindrops are millimeters in size Cloud droplets are 100 times smaller in diameter

5 Ice crystals I Ice nuclei are a different kind of aerosol

6 The cold rain process Co-existence of ice crystals and supercooled droplets leads to rapid vapor growth of crystals, followed in suitable convective clouds by accretional growth (riming), fall-out, and melting as rain Graupel

7 The origins of cloud seeding Freeze supercooled droplets by introducing artificial seeds Langmuir and Schaeffer, Schenectady, NY, Nov 13, 1946: Dropped 3 pounds of dry ice pellets into a stratus deck at 14,000 ft, -20C, flying in a race-track pattern 20 min after seeding

8 Deposition/Evaporation during melting Evaporation/Condensation Autoconversion C L O U D W A T E R Collection R A I N VWater A P Vapor O R So where are we? C o n d e n sa tio n M F reezing C o lle c tio n S h e d d i n g e ltin g Collection A E RAerosols O S O L S W A T+ E R R i m M R i m i n g e ltin g C ollection of ice by rain i n g G R A U P E L H A IL Collection E v a p o ra tio n d u rin g m In itia tio n D e p o sitio n C o l l e c t i o n M e ltin g Splintering C L O U D IC E Conversion Collection S N O W e ltin g Deposition P R E C IP IT A T IO N F A L L O U T Much has been learned about natural rainfall and seeding Warm vs cold rain The fundamental weather mod result: AgI seeding produces lots of ice crystals The effects of seeding on precipitation are essentially never as clear-cut as in the stratus cloud photo. Clouds are complex. Controversial results regarding AgI seeding for rain increase from convective clouds Re-analysis of experiments; effects less clear-cut Better progress with continued efforts studying winter orographic snowfall enhancement Greatly decreased funding since the 80s

9 Re-thinking an old method Using salt to stimulate collision/coalescence Use of pyrotechnic flares Background: Maritime air: per cm -3 (larger drops) Continental air: 500-5,000 per cm -3 (smaller) It is well known that maritime clouds rain more easily (the larger droplets collide with each other more readily)

10 Hygroscopic seeding Studies in South Africa Airborne measurement program of long standing Predominant cold rain process Large drops present in one study cloud Formed over a paper mill Hygroscopic particles present in the smoke from the mill

11 Hygroscopic Seeding Uniform distribution of droplets Non-uniform distribution Try to add larger droplets that will initiate coalescence

12 South African program (1990 s) Initiated a randomized experiment using hygroscopic flares Some aircraft measurements New evaluation methods Used the storm as the experimental unit Radar estimated rainfall Many more experimental cases Objective storm-tracking software (TITAN) Allowed for study of time-resolved response to seeding

13 Radar estimate of rainfall within the TITAN framework The storm The TITAN experimental unit Objective radar estimate of rainfall TITAN identifies and tracks individual storms based on a specified reflectivity threshold

14 Results from South Africa (Mather et al.) Radar derived rain mass (kton) Seed Control Q3 Q2 Q Time from decision

15 Results from South Africa (Mather et al.) Some measurements to support early stage of hypothesis Some modeling results to support early stage of cloud response Generated lots of interest NATURAL SEEDED DIAMETER (μm)

16 Interest from Coahuila, Mexico * Monclova * Durango

17 Components of the Coahuila program Preliminary measurements showed the clouds were similar to those in South Africa

18 Coahuila kg flares - 60% potassium chloride - 20% sodium chloride - 5% magnesium - 15% hydrocarbon binder

19 Attempt to replicate the experiment in South Africa: - Double-blind randomization - Seeding with hygroscopic flares - Evaluation based on time-resolved radar-rainfall estimates from objective TITAN software

20 Rain Mass Coahuila & 1998 seasons 0.1 Seed 0.11 Control Minutes After Decision

21 Mexico: distributions for all storms

22 Mexico: only storms with non-zero rain at time shown Log of Rain Mass > Log of Rain Mass > Not Seeded Storms Minutes From Decision Seeded Storms Minutes From Decision

23 Hygroscopic seeding Attempt to duplicate the South African experiment in Mexico showed amazingly similar results The effect seems to arise because the seeded storms had a longer lifetime Not part of the original hypothesis Seems to imply a connection between seeding and storm dynamics This connection is not well understood Hypotheses involve downdraft forcing This situation leads to caution about accepting the results During the same time frame an experiment in Thailand using hygroscopic salts also showed promising results

24 Is the rainfall from these storms hydrologically significant? Flux in cubic meters per sec Let s compare with the flow in some major rivers

25 MEXICO EXPERIMENT EXAMPLE All Storms within Radar Coverage Number of Cases in Each Rain-Out Size Category Boulder Ck 1-10 m 3 /s Coahuila 1997 and Number of Cases Colorado River Sacramento Equivalent number of very large storms (5,000 cms): Columbia 2000 Mississippi 4000 Ganges 6000 Zaire 8000 Amazon Nile th percentile 99th Zambesi Precipitation Flux (m 3 /s)

26 Attempt at replication in the United Arab Emirates Preliminary airborne measurement program Droplet distribution found to be continental UAE

27 Results of the randomized statistical experiment No statistically significant difference between seed and control cases TITAN radar image and seeding track Radar-estimated rain mass (log) The experiment failed to duplicate the previous hygroscopic seeding results

28 Tentative explanation Long period of cumulus growth below the inversion Recycling of droplets in repeated updrafts and broadening of the spectrum Natural drizzle formation even before the rainstorm breaks through the inversion Efficient ice multiplication process and lots of cold rain in thunderstorms Efficient precip process without seeding Particle images: lots of drizzle drops

29 Sulawesi, Indonesia Clean environment, maritime CCN Tropical sounding First echoes at + 5 to15c (warm rain) Efficient natural coalescence process Freezing by -5C and ice multiplication No point in seeding Watershed Area 2477 km 2

30 Classification of Mexico data using aerosol burden (satellite aerosol optical depth) Effect is most apparent on the days with significant Typical non-aerosol day Typical aerosol day aerosol (<0.1 optical burden depth) (>.1 optical depth) On cleaner, non-aerosol days, little effect Non-Aerosol cases Aerosol cases Seed & control quartile results

31 Summary Convective clouds produce large amounts of rain Comparable to flow in major rivers AgI seeding of cumulus clouds for rain enhancement is a well used technology in many countries However, it is not a well proven technology Based on experiments in S. Africa, Mexico, and Thailand, hygroscopic seeding seems promising under specific circumstances Though promising, it is not well understood There are many new tools available to establish confidence in seeding technologies

32 Summary Many factors control the amount of rain from convective clouds Aerosols (both CCN and IN) clearly affect the cloud particle structure Very likely affect the rain also (but positively or negatively?) However, the exact relationship has A E Aerosols R O S O L S W A T+ E R C o n d e n s a t i o n VWater A P Vapor O R proven difficult to establish R i m i n g M e ltin g C L O U D W A T E R Deposition/Evaporation during melting Evaporation/Condensation Autoconversion Collection R A I N F reezing Collection C o lle c tio n o f ic e b y ra in R i m i n g Collection E v a p o ra tio n d u rin g m In itia tio n D e p o sitio n Splintering C L O U D IC E Conversion Collection S N O W e ltin g Deposition C o lle c tio n S h e d d i n g M e ltin g G R A U P E L H A IL C o lle c tio n M e ltin g P R E C I P I T A T I O N F A L L O U T

33 Aerosol change: natural and anthropogenic Can mask seeding effects Satellite images of pollution effects on clouds Older experiments may not be relevant today Transferring results from one region to another is very problematic Mexico 1998 Indonesian smoke from fires in 1997 Industries

34 And so A program of research should be considered mandatory even for an operational project Preliminary measurements Aerosol, cloud structure, rainfall climatology Should include evaluation of seeding effect Need a well designed, randomized experiment to know if you are spending the sponsor s money wisely Need to look at hydrologic aspects: where does the water go? Cost/benefit study History tells us it s rarely cost beneficial to take a short-cut around the R&D

35 Thank you