Click for next page ( 24


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 23
2 Current Status of Weather MocliD~cation Operations and Research CURRENT OPERATIONAL EFFORTS In tile annual register of National Weather Modification Projects, compiled and published by the World Meteorological Organization (WMO), 24 countries provided information on more than 100 ongoing weather modification activities in 1999 (PIate 2), with most of the precipitation enhancement programs located in the subtropical semiarid belts on either side of the equator. Tl~ese data, however, pertain only to countries that report such information. and at least 10 other countries were conducting sheather modification programs. A few of these precipitation enhancement and hail suppression programs have been conducted on a continuous basis for more than 40 years. China is the most active country in pursuing weather modification, with act investment estimated at more than $40 million annually, both for hail suppression and precipitation enhancement. In the United States the number of precipitation enhancement and hail suppression programs has varied over the course of the past several decades, while the number of fog dissipation projects has remained nearly constant throughout this time (with the primary example being the program sponsored by Delta Airlines at Salt Lake City International Airport). In the last few years there has been an increase in operational weather modification activities in the United States, with approximately 66 programs (for hail suppression and snow or rain enhancement) being conducted in 2001, according to activities reported to NOAA (Plate 2~. All of these projects are located in the southern and western states of the United States and are sponsored by local, state, or private entities. No federal funding currently supports any project. The increase in operational programs over the past 10 years indicates a growing perceived need for enhancing water resources alla mitigating severe sheather in many parts of the world, including the United States. For users and operators of weather modifications technologies, the decision of whether to implement or continue an operational program is a mattes of cost-benefit risk management, which raises questions about what constitutes "successful" modification. Cloud-seeding experiments have shown mixed results, but many operational cloud-seeding programs continue, based on what is seen as circumstantial or indirect evidence of positive results. For instance, 23

OCR for page 23
24 CRI TICA L l.SSlJES {.\ {VEA TI-lER A JODIFICA TI ON RESE-A RCI l studies of hail-damage insurance claims in North Dakota over a seven-year period show a 43 to 45 percent reduction in claims in counties where hail suppression is carried out (Smith et al., ] 9979. Studies of rain enhancement programs in this state report up to a 15 percent increase in rainfall (Johnson, 1985) and up to a 5.9 percent increase in wheat yields (Smith et al., 19921. Indirect qualitative assessments of the additional water produced frolic the Utah operational programs described by Griffith (1991) indicated costs in the range of a few do]]ars per acre-foot (StauffUr and Wi]]iams, 20009. The Tasmanian program calculated a cost-benefit ratio of 13 to 1 (Ryan and King, 19973. These results are viewed as a beneficial for hydropower energy production (Cotton and Pielke, 19953. There is little or no research associated with any of these operational programs, which highlights the need for intensive studies to further develop a scientific basis tar weather modification technology. Marty current precipitation enhancement projects, particularly in developing countries, use old technology and Jaclc the latest instruments and other operational tools. The use of modern observational tools, models, experimental design techniques, and statistical evaluation techniques are prerequisites for shedding light on cause-and-effect ~ elationships. CURRENT SCIENTIFIC EFFORTS Currently there are very few weather modit~catio~ research programs in the world. As discussed in Chapter 1, research in weather modification was actively pursued after the initial discoveries in the late 1 940s and pealed in the late 1 970s, when funding in the United States alone was around $20 million per year. This amount dwindled after 1980 to less than $50O7000 per year and has continued to decline in recent years. A few research projects on a smaller scale have continued in the United States and several other countries, including South Africa, Thailand, Mexico, Argentina, Israel, Japan, and the United Arab Emirates. In the following sections and in Appendix A, the statics and current scientific understanding of various aspects of weather modification are reviewed. Precipitation Enhancement Weather modification research requires the involvement of a wide range of expertise due to the multifaceted nature of the problem and the large range of scales that are addressed. The chain of events in precipitation development ranges from at least the mesoscale dynamics determining, the characteristics of the cloud systems down to small- scale microphysics determining the nucleation and growth characteristics of water droplets and ice particles (e.g. see Pruppacher and Klett, 1998; graham, 1979, 1986b; Dennis, 1980; Rogers, 1976~. Our knowledge of the individual steps in this chain has increased significantly in the past 20 years, but major gaps still exist in our understanding of certain physical processes. Although most rainfall enhancement experiments focus on modifying the mierophysical aspects of clouds, it is important to emphasize that cloud mierophysieal and dynamical processes are intimately linked, and that the major controls on precipitation occurrence and amounts are the mesoseale and larger-scale atmospheric dynamics (e.g. see Cotton and Aptness 1989; Vali et a]., 19881. At present, however, no

OCR for page 23
CURRENT STA TlJS OF THEA TilER MODIFf CA THAN OPERATIONS AND RE.S~-ARCll 25 theoretical framework or experimental methodology exists that could support any intentional modifications of the atmosphere on these larger scales (see Chapter 44. Precipitation enhancement from ~~ixed-pl~ase clouds (i.e., clouds or parts of the clouds containing temperatures below 0C) has been the focus of most weather n~odification research and operations around the world. The microphysics and dynamics of these cloud systems are complex and, especially in the case of convective storms, are characterized by large natural variability. Establishing cause-and-effect relationships through the complete chain of eyelets leading to precipitation formation is extremely challenging. Glaciogenic seeding immaterial (see also Chapter 1) is the most common seeding material used for precipitation enhancement. Hygroscopic seeding material, such as salt powders, also has been used but has generally proved to be less attractive than glacioge~ic seeding material. During the past decades however, tests have been conducted on ~~ixed-p}~ase clouds using small (sub-micron to tens of microns in diameter) hydroscopic particles released by pyrotechnic flares. The r esuJIs of glaciogenic and hydroscopic precipitation enhancement techniques are distilled in the following, section (see Box 2.1 for a summary), and the detailed methodology is presented ifs Appendix A. Glaciogenic Seeding Experiments Based on the quantity of glaciogenic seeding material used to enhance ice content, two seeding concepts have historically been proposed and widely referred to as "static" and ' dynamic" seeding. In the static seeding concept the aim is to capitalize on the less-than-optimal ice crystal concentrations often present in nature, which leads to prolonged periods of supercooled water, especially in orographic clouds. These regions of supercooled water have to exist for a sufficient length of time for ice crystal growth and precipitation to occur. In the dynamic seeding concept the emphasis is on the release of latent heat by rapid freezing, which enhances buoyancy and invigorates cloud growth, thereby increasing precipitations production. It should be noted that these concepts are not mutually exclusive because they both result in increased ice crystal concentrations and affect cloud dynamics. The same seeding material is used in both seeding concepts and only the quantity of seeding material is varied. While the dynamic seeding concept is primarily applicable to convective clouds, the static seeding concept has been widely utilized in orographic and layer-type clouds as well as in convective clouds. In convective clouds, both "static" and "dynamic" responses can occur in a mutually interactive fashion (Rosenfeld and Woodley~ 1993) Static Seeding. Convective C.'loucis The top half of Table 2.1 lists examples of static glaciogenic seeding experiments designed to test whether precipitation can be increased in convective clouds in response to seeding with ice nucleating agents. For static seeding of convective clouds, statistically significant rainfall increases were not obtained or, in the case of the Israeli experiments, continue to be debated (Gabriel and Rose~feld, 1990; Rosenfeld and Farbstein, ]992; Rangno and Hobbs, 1995; Rosenfeld and Nicely 1996; Levi and Rosenfeld, 19961. In each

OCR for page 23
26 CRITICS L [LSSZJES IN [YEA TI-lER AlODIFJCA'T~ON RESEARCIJ TABLE 2.1 Examples of Static Glaciogenic Seeding Experiments in Precipitation Enhancement Type ofeloud Experiment Uc .: Convective clouds Arizona projects Battan and Kassa~der, 1967 Israeli expel iments Gagin and Newmann, 1 974 Projeet Whitetop graham, 1964, 1979 High Plains Ex,nerime~t Smith et al., 1984 (TRIPLEX) I Winter orographic Lake Almanor Mooney and Lunate 1969 clouds experiment Sierra Cooperative Pilot Reynolds and Dennis, 1986; Projeet (SCPP) Deshler et al., 1990; SCPP, 1982 Climax T and 11 Grant and Mielke, ] 967; Mielke etal.' 1981 Bridger Range Super and Heimbaeh, 1983; experiment Super, 1986 Tasmanian experiments Ryan and Kin:, 1 997 . . . . . . . case, however, useful results or guidance was obtained which contributes to the current knowledge base in weather modification. Among these results are: . that physical measurements in clouds are essential to provide an understanding of the underlying processes; . that high concentrations of fee crystals occur naturally in some cumulus clouds at temperatures as warm as-10 C thus allowing rapid production of precipitation particles; ~ that the window of opportunity for enhancing rair~fal] from a given cloud (system) is limited; that treatment can both enhat~ee and reduce rainfall; and that results based on small clouds might not be transferable to dynamically more vigorous and larger cloud complexes. Static Seeding. Stinter Grog' aphic Clods In the case of static seeding of winter orographic clouds (bottom of Table 2.1), important results include: recognition of the complex interactions between terrain and wind flow in determining regions of cloud liquid water and, later, through microwave radiometer measurements, flee existence of a layer of supercooled water; acknowledgment of the need to target and track the dispersion of seeding material and, again later, the demonstration of complex flow including ridge-parallel tlows below the ridge crest exist in pronounced terrain; evidence of marked increases in fee particle concentrations leading to increased precipitation depending upon the availability of supercooled liquid water; re-emphasis of the need for physical data that can be used together with numerical models to identify the spatial and temporal changes in cloud structure;

OCR for page 23
CURRENT STA TlJS OF T1i'EA Ti-lER MODIFICA Tat: OPERA 710~S A ND RESEARCll 2 7 development of highly efficient silver chloro-iodide ice nuclei and other fast acting, highly efficient ice nucleating pyrotechnic and generator devices (Fig. 2.29; and ~ development of methods to detect traces of seeding agents in snowpack and rain water. D:'nc~mic Seeding Table 2.2 lists four examples in which glaciogenic seeding was used in flee expectation that an increase-in cloud buoyancy would fallow Freezing of supercooled water drops. The intent was to seed supercooled clouds with large enough quantities of ice nuclei (100-1000 cm~3) or coolant to cause rapid glaciation. Increased buoyancy was expected to cause the cloud to grow larger, ingest more water vapors and yield more precipitation. It was postulated that increased precipitation would enhance downdrafts and outflows which, in turn, would initiate new convections and extend the effects of treatment (Woodley et al., 19821. Few ot the hypothesized steps in the chain calf events have been measured in experiments or validated by numerical models (Orville, 19964. However, as in the case of static seeding, dynamic seeding has contributed significantly to our current store of knowledge. Among the findings and results frown dynamic seeding experiments that contribute to the current state of knowledge in weather modification are: ~ the complexities of fee formation in clouds where fee and supercooled water have been found at temperatures as Leigh as -] 0C and as low as -38C, r espectively (Rosenftld and Woodley, 2000~; . the dependence of fee formation upon CCN concentrations and sizes (e.g., heezing of large drops) and the role of primary and secondary ice formation in graupel production which have emerged fob these experiments are areas of uncertainty; ~ the importance of coalescence (and hence aerosols) on cloud structures evolution and lain production (Roser~feld and Woodley, 1993; Johnson, 1987~; . the importance and relationship between cloud dynamics and microphysics and the induced changes resulting from seeding; and . the power and limitations of existing radar systems (Chapter 4) as integral experimental tools and as possible means of verification of seeding results. TABLE 2.2 Examples of Dynamic Glaciogenic Seeding Experiments ilk Precipitation E~l~ance~ne'~t. Experiment Florida Area Cumulus Experiments (FACE) 1 and 2 Texas experiments South African experiments Thailand experiments Reference Woodley et al., 1982; Woodley et al., 1983; Gagin et al., 1986 Rosenfeld and Woodley, 1993 Bruintjes et al., 1987; Krauss et al., 1987 Woodley et al.' 1999 . . . . . . . . . . . . .. . . . .

OCR for page 23
28 Hydroscopic Seeding Experiments CRITICAL 75~5lJE.S IN TT'FATI--IER AlODIFI(-A7'10NREtSEARCll HygIoscopic seeding, as opposed to glaciogenic seedings is directed at promoting bile coatescer~ce of water droplets in flee cloud. The intention is to promote particle growth through coalescence and thereby improve the efficiency of the rainfall formation process. Appropriately sized salt particles, water droplets from sprays of either water or saline solution (Bower, 1952; Biswas and Dennis, 1971; Cotton ]982; Murty et al., 2000; Silverman arid Sukar~janasat? 2000), and l~ygroscopic flares Matter et al., 1997; WMO, 2000) have been used. Statistical results, observations and modeling results for large (>10 Em diameter) have provided some statistical evidence (Murty et al. 2000; Silverman and Sukarnjanasat' 2000) and evidence that under certain conditions Title optimal seed drop size spectrums, precipitation can be enhanced (Farley and Chen, 1975; Rokicki and Young, 1978; Young? 1996~. The hydroscopic flare particle seeding experiments have provided statistical support for rainfall increases due to seeding based Ott single cloud analyses, but the physical processes leading to these increases in precipitation are not well understood. Despite the wide use of hydroscopic seeding, the results have been inconclusive due to a lack of physical understanding adds in some cases, inconclusive statistical evaluations. Table 2.3 lists examples of field experiments or operations in which hydroscopic seeding was employed. Among, the results from these programs that have contributed to the current state of knowledge in weather modification are: that both the South African and Mexican experiments produced remarkably similar statistical results in terms of the differences in radar estimated rainfall for seeded versus non-seeded groups (Plate 3) (Bight, 1997; Silverman, 2000; WMO, 20001; that in the South African and Mexican experiments, reevaluation of the results showed an increase in rain mass 30-60 minutes after seeding, significant at the 96 percent level (u = 0.04) or higher; that marked differences in concentrations of ice particles were found in maritime clouds (high) versus continental clouds (low) signifying the active role of collision and coalescence in maritime clouds compared to continental clouds (Scott and Hobbs, 1977; Cotton, 1972; Koenig and Murray, 19761; that freezing temperatures increased with increasing drop size because larger droplets contain or have a higher probability of colliding with ice nuclei; ~ , ~ ~ that relatively large do oplets (>24 frill) played a role in fee multiplication processes, including mechanical fracturing during melting and evaporation arid fee splinter formation during riming (Hallet and Mossop, 19741; that a delayed response in radar-derived storm properties was a possible function of seeding-induced dynamic processes beyond the classical cloud physics results that links cloud condensation nuclei arid droplet spectra to rain production (WMO, 20001; and that hygroseopie seeding might overcome inhibiting effects on rainfall of air pollution (Rosenfeld et al., 20021.

OCR for page 23
CURRENT STA TlJS OF T~krEA Ti-lER MODIFICA TIOA: OPERA T10VS AND RESEARCI1 29 TABLE o.3 Examples of Hygroscopic Seeding Experiments in Precipitation Enhancement Experiment Reference South Ah lean experiments Indian experiments Thailand experiments Mexico experiments Mathel et al.7 1997 Murty et al.' 2000 Silverman and Sukarnjanasat, 2000 WMO, 2000 Hail Suppression Hall suppression programs are driven by tile severe impacts of hail on many different sectors of the economy. In recent years hail damage to crops in the United States typically leas been around $23 billion annually (Cl~angnon, 19984. Susceptibility to BOX 2.1 Summa'' of Cloud-Seeding Techniques for Precipitation Enhancement Glaciogenic Seeding: Seeding of clouds with appropriate ice nuclei (e.g.' silverer iodide) or cooling agent (e.g., dry ice, liquid propane) to create or enhance the formation of ice crystals, particularly the conversion of supercooled water to ice. The two general approaches ale 1. Static seeding, which focuses on microphysical processes; creation of ice crystals and particles; enhances gravel and snow production bY increasing the slumber ot ice particles and triggering prec1pltat1oll process earlier In the ClOUd S lifetime. Examples. Climax I and II, Israel, Project [Khitetop. 2. Dynamic seeding which increases buoyancy of cloud by converting supercooled liquid drops to ice. The subsequent release of latent heat of fusion increases cloud buoyancy, cloud lifetime, and rain production. Examples. FACE I and II, Texas. Hygroscopic Seeding: Enhance rainfall 1 by seeding clouds with appropriately sized salt particles ot droplets, promoting the coalescence process. . Large l~ygroscopic particle seeding, which seeds clouds with large salt particles (e.g., >10 rim dry diameter) to short-circuit the condensation growth process and provide immediate raindrop embryos to start the coalescence pi ocess. Examples. Project Cloud Catcher, India, Thailand. 2. Hygroscopic flare seeding, which focuses on broadening the initial drop spectrum during the nucleation process by seeding with larger than natural CON (0.5 lam to 3 lam dry diameter) to enhance the coalescence process in warm and mixed-phase clouds. Examples. South Africa, Mexico experiments.

OCR for page 23
30 CRITIC,4 L [~SSIJES IN THEA TI-IER AIODIFICA THIN RESEARCTI damage depends on the crop type, its stage of development, the size of the hail and the magnitude of any wind accompanying the hail. Any theory of hail growth that is complete enough to serve as the basis for suppression must include at least the following elements: (1) hail embryo formation process, including flee microphysics of particle growth arid the region or regions in the storm where such growth occurs; (2) transport of embryos to regions of abundant supercooled liquid water where flee further growth to hail is possible; (3) growth trajectory of the hailstone itself as it passes through the strong updraft of a storm; and (4) the time evolution of the storms updraft arid cellular development. Such processes and variables as ice nucleation, dominant rain formation, cloud-base temperature, environmental wind shear, and updraft strength and widely are also essential elements of hail formation. Sulakvelidze et al. (1974), attempted to combine these elements in a united theory of hail formation. Subsequent work showed the complexity of hail producing convective storms ranging from the "ordinary" through severe multi-cell storms to supe~cell storms (Browning and Foote, 1976; Browning et al., 1976; Foote and Knight, 1977~. Radar measurements, including muJti-Doppler, and aircraft studies have produced hail growth trajectories within the measured storm velocity fields (Foote, 19854. None of these or other studies have provided an adequate description of the essential elements of hail formation. Advocates of hail suppression programs claim positive results based upon reported reductions in crop-hail insurance losses (e.g., 45 percent in the study of Smith et al, 1997 and 27 percent in the study of Eklu~d et al., 1999~. However, natural variability in crop-hail insurance losses from season-to-season and an apparent long-term decline beginning around ]950 in hail losses (Figure 2.1) make these data difficult to interpret unambiguously. Numerical models of storms can be a useful vehicle for testing hail theories. They provide a self-consistent environment for computing hail growth and liquid water depletion. Indeed, much has been learned about the dynamics of storms using cloud models (e.g., Weisman and Klemp, 1982, ]984~. However, as discussed in Chapter 4, models powerful enough to include the details of the dynamics and microphysics in three dimensions still do not exist. Such sophisticated models (e.g., bin-mixed-phase, microphysics with full aerosol interactions) are feasible with computer resources commensurate with those currently supporting climate simulations. Other Severe Weather Phenomena Lightning Cloud-to-ground (CG) lightning has been a major cause of fires in man-made structures and in forests, and it has been the cause of many human deaths. While lightning protection has been a topic of study for several centuries and numerous technologies have been developed (AMS, 1998), studies on lightning suppression or the modification of lightning characteristics by inadvertent or advertent intervention has only

OCR for page 23
C (JRRENT STA TlJS OF T~krEA TI.IER MODIFICA TIM OPERA TIO.VS A ND R. K-SEA F?Cl-I 3 1 20 1 18 t 16 t 14 1 lo o 10 ._ I o ._ IN 4 8~ 6t O - ~ 9T 8t 7t 6t 5t o - /) Crop Hail Insurance Loss Ratios I-- Montana Control Area ~ North Dakota Target .. I i1 - t 1920 1930 1940 1950 1 1 ~ 1 -Y: 1920 1930 1940 Year Year 1 -- I. 1960 1 970 Ratio of Target to Control (North Dakota / l~ontana) 1980 1 990 1'~ ~ .? ,~ * . 1960 1970 1980 1990 FIGURE 2.1 Results from an operational hail suppression program in North Dakota with hail losses the unitless ratio of insurance damage claims paid from hail events to the total insured liability reported in two adjacent areas. From Smith et al., 1997, the upwind Montana area was treated as a control for the North Dakota area in which hailstorms had been seeded. During the heavily seeded years of 1976-198S, the seeded area shows proportionally less hail than the control area. However, the ratio of the two hail loss curves (shown in the bottom figured indicates this trend started as early as 1950. SOURCE: B. Foote, adapted frown Smith et al.' 1997.

OCR for page 23
32 CRITICAL [LSSlJESINTLEATI-IER AlODIFI(-~ATIONREtSF-ARCt~J recently come to the foreground. Lyons et al. (1998) reported that lightning-producing storms that ingested smoke from biomass burning displayed altered electrical characteristics. Smoke-affected storms had an anomalously large fraction of positive cloud-to-g~ound ligl~tni~g strikes, probably due to changes in microstructure oftl~e clouds. In North America most wildfires are initiated by lightning, but most negative cloud-to-round strokes are of short duration. Providing insufficient time for i~niti~ ~ _ - 7 ~ O t . ~ ~ A ~ ,~ ... . 1 ~ t biomass tires (Uruguay et al., I/. About YE percent of positive strolls are of long enough duration to ignite fires (Pyne et a1.~ 1996) and could possibly also have adverse impacts Ott engineered structures. In clouds unaltered by smoke CCN only about five percent of CG lightning flashes are positive strikes. Hence, if there is an increase ire frequency and durations of wildfires and their smokes one might also expect to inct ease the number of ligl~tning-initiated wildfires. Steiger et al. (2002) in a study of CG lightning anomalies (enhanced lightning frequency over Houston? Texas, attributed the increased frequency of lightning to the possible heat island effect. It also was found that increases in lightning were most pronounced when urban air pollution was highest. Houston has ~ strong oil refinery and automobile presence, and it is well known that oil- related industries produce large amounts of sulfur dioxide, which transforms to sulfates that are verb efficient CCN. These findings appear to corroborate earlier findings of increased thunderstorm f] equencies in an effluent plume in St. Louis, Missouri (Changnon et al., 1981~. More recently Williams et al. (2002) proposed a conceptual mode! by which added smoke and other air pollution aerosols could increase the lightning activity of convective clouds. Furthermore, aerosol and cloud interactions are of central importance in these studies as in studies of the pollution effects on rainfall cited elsewhere. There has been some interest in the suppression of lightning for the purposes of reducing lightning-induced forest fires and diminishing lightning hazards during, the launch of space vehicles. The concept usually proposed involves reducing the electric fields within thunderstorms so that they do not become strong enough for lightning discharges to occur. Qualitative studies of: CG lightning suppression through injecting metallic chaff into maturing cumulonimbus also have recently been suggested (Orville, 20011. A few years ago thunderstorms developed in Arizona in which one complex stolen produced numerous CG and another almost none. Post analysis found that the CG-free storm complex had formed in an area where the military had been conducting chaff experiments that same day, and it was postulated that the chaff had suppressed electric fields in flue storm, resulting in only in-cloud lightning production. Limited fieldwork has been done on this topic. Holitza and Kasemir (1974) arid 1(asemir et al (1976) reported that using chaff seeding, they found a reduction in lightning by a factor of three for seeded versus non-seeded storms. Helsdon ( 1980) numerically simulated the chaff seeding in a two-dimensional cloud model. The results showed that the chaff produced large numbers of positive and negative ions, leading to a decrease in the vertical electric field in the cloud. Howevet, these few studies are qualitative in nature and are not statistically significant due to limited evaluation capabilities at the time ' ' Improved statistical techniquesnamely Bayesian methods, which are ideal for accounting tot uncertainty and providing spatial-temporal analyses (see Appendix B)could provide more conclusive results if chaff seeding experiments were conducted again today.

OCR for page 23
CURRENT STA THIS OF THEA TI-IER MODIFICA TIOA: OPERA TlO.VS AND RESEARCI-1 33 The University of Florida's Lightning Research Center? the international Center for Lightning Research and Testing at Camp Blandish, Florida, and other research centers have carried out studies wherein lightning is triggered by launching a small rocket trailing a grounded wire. It leas been found that lightning flashes can be triggered from clouds to ground roughly 50 percent of the time (Uman et al., 19979. Improvements in out understanding of the physics of lightning have led directly to the design and installation of lightning protection devices for a variety of electrical and electronic systems. Hurricanes Tropical cyclones contribute significantly to the annual rainfall of many areas, but they also are responsible for considerable damage to property and for a large loss of life. Due to increases In population density in the coastal zone of the lower 48 United States over the past 30 years, both casualties and costs due to damage and disruption are expected to continue to rise (see Table 1. 1~. Damage estimates due to Hurricane Floyd in 1999 exceeded $1 billion, and costs associated with evacuation equalled that number (Pielke and Carbone, 2002~. Therefore, the aims of any modification procedure should be to reduce tile wind, storm surge, and rain damage but not necessarily flee total rainfall. Hurricane modification experiments were conducted in the 1 960s and early 1 970s (Project Stormfury) (Simpson and Malkus, 1964; Simpson et al., 1967; Willoughby et al., 1985), but the results were inconclusive, and there currently is no generally accepted scientific conceptual model suggesting that hurricanes can be modified (see Box 4.11. Tornadoes Although modification of tot-nados and other storms producing damaging winds is desirable for safety and cost reasons, there presently is no scientifically acceptable physical hypothesis to accomplish such a goal. Freezing Drizzle and Rain Speculations can be made about the possibilities of reducing aircraft icing episodes or mitigating icing of highways and roads by seeding nearby supercooled cloud regions, but there is loo physical, conceptual model on how to mitigate these hazards and no work has been done in this field. Flash Floods and Large-Scale Flooding No physical conceptual model exists to mitigate these events and no world has been conducted in this field. If the precipitation processes were fully understood, then perhaps procedures could be designed to decrease rains from flood-producing rain clouds. Accurate numerical modeling of such cor~ditions would be necessary for such studies.

OCR for page 23
34 CRITICA L [.SSl.JES IN THEA TIlER AlODlPlCA LION RE.S-~-iARCI-I Inadvertent Weather Modification Human activity is inducing inadvertent effects in the atmosphere on scales ranging from the local (a given point source of pollution, urban heat island, contrails? etc.) to the global (changes in greenhouse gases and aerosols and associated cloud effects). Global effects of cleanses in greenhouse gases, aerosols, arid cloud cover are of fundamental concern, but they go beyond the scope of this report. However, the evidence of local to regional cloud arid precipitation cleanses due to anthropogenically derived aerosols is highly relevant to the issue of deliberate weather modification; and it is discussed below. Aerosol effects on clouds and precipitation are a complex, multi-order problem. In ]957? Gunn and Phillips (1957) documented the detrimental effects of air pollution CCN on clouds and precipitations. Twomey ( 1974) then postulated that increased pollution results in greater CCN concentrations and numbers of cloud droplets, which ill turn increase the reflectance of clouds. Twomey et al. (l 984) argued that enhanced closed albedo has a magnitude comparable to that of greenhouse warming and acts to cool the atmosphere. Evidence of cloud and precipitation changes due to aerosols (changes in "natural" CCN) is becoming widespread. There is ample evidence now that biomass burning and other anthropogenic sources of aerosols affect the radiative properties of clouds and precipitation processes in clouds, leading also to changes in the dynamical processes in clouds (i.e., effects on cloud lifetimes). Increased CCN lead to higher droplet concentrations and a narrower droplet spectrum (which manifests itself as a higher cloud albedo), which leads to suppressed drizzle formation and longer lasting stratiform clouds (e.g., ship-track studies, jJAS, 2000 and Albrecht, 198944. Recent satellite studies of cloud microstructure downwind of biomass burning and industrial pollution sources have also suggested suppressed precipitation formation in the affected clouds, as illustrated in Plates 4 and 5. (Ramanathan et al., 2001; Rosenfeld, 1999, 2000;~. However? Cotton and Pielke (1995) noted that the susceptibility of the drizzle process in marine stratocumulus clouds to anthropogenic emissions of CCN may depend on the presence or absence of large and ultra-giant aerosol particles in the sub- cloud layer. In oilier chords, the drizzle formation process is not solely regulated by the concentrations of CCN and cloud liquid water contents, but possibly also by the details of the spectrum of the hydroscopic aerosol population. In fact, Rosenfeld et al. (2002) showed that sea spray? even under light wind conditions' can restore precipitation from .. . . . . . ~ , . ~ ~ . . ... . . .. polluted convective clouds, doing naturally what deliberate hydroscopic seeding Is attempting to achieve artificially. In addition, tile intriguing evidence of Inca eased positive lightning flashes in storms affected by smoke from the Mexican fires of 1998 is yet another example of the complex effects of aerosols on clouds, precipitation, and the microphysics relevant to cloud electrification (Lyons et al.' 1998~. The effects of desert dust and mixtures with anthropogenic pollutants are important to warm rain and ice processes through their ice nucleating ability' and, possibly through the coating of sulfates? their droplet nucleating ability. The apparent decrease in rainfall in the softly target area in the Israel; II study was linked by Rosenfeld and Farbstein (1992) to the incursion of desert dust. They suggest that desert dust contains more ice nuclei and also provides coalescence embryos (when coated with

OCR for page 23
ClJRRENT STA THIS OF T7FrEA TI-lER lIODIFICA TION OPERA T10.VS AND REtS~EARCl-I 35 sulfates) that could enhance the collision-coalescence process in clouds thus providing efficient precipitation processes in these clouds. OTHER RESULTS Over the past few decades there have been considerable advances in the basic sciences relevant to weather recodification. For instance, tl~rougl~ cloud modeling there is a better understanding of the -microphysics of clouds and the dynamics of clouds and weather systems. More effective ice nucleants and hydroscopic nucleus flares have been developed. Progress has been made in combining cloud microphysics and cloud dynamics in three-dimensior~al numerical models, which give promise for better definition of where and when seeding intervention may be most effective. New tools and techniques are available for remote sensing of conditions In clouds, delineation of zones identified for seeding, to aching seeded volumes, and monitoring changes in cloud structure following seeding (as discussed further in Chapter 41. Collectively, these areas could be viewed as the scientific infrastructure of weather modification, but many of the relevant advances have yet to be applied to weather modification research. Table 2.4 list results oldish have been obtained *om new observing systems and laboratory and modeling studies that have not necessarily been an integral part of weather modification research over the past three decades. In each case, however, there is a direct or potential application to weather modification that has not yet been fully ~ ealized TABLE 9.4 Otl~er Results Derived from Nests Observing Systems arid Laboratory and Modeling Studies Area of Research Result Aerosols Cloud droplets Cloud ice Sources and sinks Influecee on size distribution and number concentrations of cloud droplets Aquatic-phase chemistry and cloud scavenging Aerosol-induced changes in cloud drop size spectra Role of pol] ution In-cloud recirculation Physics of drop-drop collisions and collision and coalescence efficiencies Drop size freezing More universal occurrence of coalescence in producing (warm) rain Relationship between drop shape and size distribution, radar reflectivity and rainfall rates Particle riming and the secondary production of ice particles Microphysics of hail pi oduction

OCR for page 23
36 CRITI(~4L [.SSI;JFS TV {T'EA Ti-lER A~fODIFI(-A TION RE`S~-ARClI RECOGNITION OF KEY UNCERTAINTIES lN WEATHER MODIFICATION The current state of knowledge in weather modification as summarized in the preceding sections provides sufficient guidance to identify key uncertainties which need to be addressed before substantial progress in weather modification is likely to be made. Box 2~2 provides a list of key uncertainties wl~ich stem from the current state of knowledge in weather modification. These uncertainties transmute into questions which identify roadblocks where research should be focused and which constitute a framework thatwith the concerted application of current technology, modeling' and statistical analysis described in the following chapter and Appendix B can promise substantial progress in determining and demonstrating to what extent we influence, modify, or even control flue weather. Sucl~ a framework clearly identifies critical roadblocks to progress where research resources should be focused. BOX 2.2 Summary of Key Uncertainties The statements in boldface type are considered to have the highest priority. Cloud and precipitation ~nicrophysics issues Background concentration, sizes, and chemical composition of aerosols that participate in cloud processes Nucleation processes as they relate to chemical composition, sizes, and concentrations of hydroscopic aerosol particles Ice nucleation (primary and secondary) . Evolution of the droplet spectra in clouds and processes that contribute to spectra broadening and the onset of coalescence Relative importance of drizzle in precipitation processes Cloud dy'?a~nics issues Cloud-to-cloud and mesoscale interactions as they relate to updraft and downdraft structures and cloud evolution and lifetimes . Cloud and sub-cloud dynamical interactions as they relate to precipitation amounts and the size spectrum of hydrometeors . clouds Microphysical, thermodynamical, and dynamical interactions within Cloud-'nodel~ng i.ss~`e`v . Combination of the best cloud models with advanced observing systems in carefully designed field tests and experiments . Extension of existing and development of new cloud-resolving models explicitly applied to weather modification Application of short-term predictive models including precipitation forecasts and data assimilation and adjoins methodology in treated and untreated situations ~ Evaluation of predictive models for severe weather events and establishment of current predictive capabilities including probabilistic forecasts

OCR for page 23
CURRENT STA TlJS OF TiPrEA TI-lER MODlE! .CA TIOA: OPERA TIO.VS AND RESEARCH 3 7 Advancement of the capabilities in cloud models to simulate dispersion trajectories of seeding material . Use of cloud models to examine effects of cloud seeding outside of seeded areas ~ Combination of cloud models with statistical analysis to establish seeding effects Seeding issues Targeting of seeding agents, diffusion and transport of seeding material, and spread of seeding effects throughout the cloud volume Measurement capabilities and limitations of cell-tracking software, radar, and technologies to observe seeding effects ~ Analysis of recent observations with new instruments of high concentrations of ice crystal . Interactions between different hydrometeors in clouds and how to best model Alum simulation . Modeling arid prediction of treated and untreated conditions for Mechanisms of transferring the storm-scale effect into an area-wide precipitation effect and tracking possible downwind changes at the single cell, cloud cluster, and floating target scales

OCR for page 23