Author: Art Rangno
Catalina Cool Season, October – May Rainfall, 1978 through 2025
A Review and Enhancement of the Cloud Seeding Chapters in the 2007 book, “Human Impacts on Weather and Climate”
Another in a continuing possibly semi-useless series by this author…. This example probably indicates why I am not asked to review manuscripts in my expertise; cloud seeding and ice formation in clouds. I try to follow in the footsteps of meteorologist and MIT faculty member, Fred Sanders, of whom it was said, “His reviews were sometime longer than the manuscript he was reviewing.” I bet he didn’t get many manuscripts to review, either!
Oh, well, “we” trudge on.
——————-
Comments and “enhancements” on this work by my friends, William R. “Bill” Cotton and Roger Pielke, Sr., are in red. Necessarily, I am only presenting those portions of Cotton and Pielke’s book (those on cloud seeding) that require an “enhancement” or corrections for the reader and I have attempted to do this delicately. Fortunately, because both are major scientists, they do not respond to criticism with emotion and are quite happy to see errors in their work corrected. 🙂
For the actual, extensive text leading up to these comments, you’ll have to buy the book. Dashes are inserted where material, usually extensive, is skipped to avoid too much copyright infringement.
Summary
The first three chapters are an excellent overall introduction to the topic of weather modification. The “Reference” section alone makes it worth the purchase of this book since it covers so much of the climate domain up to 2006. The weather modification references needed beefing up and are done so here.
Oh, some background on what the climate was doing when this book on climate came out…
Cotton and Pielke, Sr.’s book, hereafter, “CP07,” came out during the middle of a hiatus in global warming, discussed a few years later in Science magazine by its reporter, Richard Kerr (2009) in, “What Happened to Global Warming?”
About the time CP07 was published it also marked a time when the phrase, “global warming” (which was no longer occurring for unknown reasons) began to recede in use and was supplanted by the phrase, “climate change,” something that is always happening on this great planet. It made sense to change phrases since “climate change” would always be true whereas “global warming,” as we were learning in the 2000-2010 era, might not be. (An aside: I am a believer that CO2 will warm the earth in the decades ahead, but not catastrophically; I am strongly influenced in this belief by “influencers,” Roger Pielke, Jr., climate and policy expert, formerly of the University of Colorado but pushed out, and Cliff Mass, weather and climate expert, still “intact male” at the University of Washington thanks solely to tenure.
Let us begin the review of CP07 with their acknowledgments which required an insertion by this writer:
Acknowledgments for CPO7
The study of human impacts on weather and climate continues to be a high- interest topic area, not only among scientists but also the public. Our second edition has continued to build on our funded research studies from the National Science Foundation, the National Aeronautics and Space Administration, the Environmental Protection Agency, the Department of Defense, the National Oceanic and Atmospheric Administration, and the United States Geological Survey. Our numerous research collaborators at the Natural Resource Ecology Laboratory and Civil Engineering at Colorado State University have continued to provide valuable in sight on this subject. Over our multidecadal career, the fundamental insights into weather and climate provided by our education at the Pennsylvania State University have become increasingly recognized. We also want to recognize the perspective on these subjects, and science in general, that Robert and Joanne Simpson have provided us in our careers. Their mentorship and philosophy of research, of course, is but one of their many seminal accomplishments.
Roger Pielke would like to thank everyone who contributed to compiling the information in Tables 6.2 and 11.2 especially Roni A vissar, Richard Betts, Gordon Bonan, Lahouari Bounoua, Rafael Bras, Chris Castro, Will Cheng, Martin Claussen, Bob Dickinson, Paul Dirmeyer, Han Dolman, Elfatih Eltahir, Jon Foley, PavelKabat, George Kallos, Axel Kleidon, Curtis Marshall, Pat Michaels, Nicole Molders, Udaysankar Nair, Andy Pitman, Adriana Beltran-Przekurat, Rick Raddatz, Chris Rozoff, J. Marshall Shepherd, Lou Steyaert, and Yongkang Xue. In addition, Roger would like to thank Dr. Adriana Beltran for her assistance with figures in this edition.
As is always the case, Dallas Staley’s editorial leadership and Brenda Thompson’s assistance in completing the book has been invaluable and is very much appreciated.
“However, we are unable to thank Mr. Arthur L. Rangno for his review of this book before it was published because we forgot to ask him. Mr. Rangno is an acknowledged expert with several peer-reviewed publications on two of the cloud seeding experiments reviewed in this book; those carried out by Colorado State University (the home institution of CP07) at Climax and Wolf Creek Pass, Colorado, and those carried out in Israel conducted by the Hebrew University of Jerusalem. Namely, Art didn’t do sh. to improve our book beforehand since we also mostly ignored his scintillating 1997 email concerning the first edition of our book, CP95. Nevertheless, we are happy to have him review our 2nd edition (CP07) belatedly, i.e., provide a few review “cow pies” here and there on some of our otherwise excellent work.” (Fake quote.)
If you would like to see how truly “scintillating” my email was to the lead author, go here
Since Professor Doctor William R. “Bill” Cotton does not thank any reviewers in CP07 while Prof. Doctor Roger Pielke, Sr., thanks many contributors, I am of the belief that the cloud seeding chapters, 1-3 in CP07, were not reviewed by anyone until now. Dr. Cotton is on record as favoring commercial vendors of orographic cloud seeding.
So, off we go!
Chapter 1: The rise of the science of weather modification by cloud seeding
Throughout history and probably prehistory man has sought to modify weather by a variety of means. Many primitive tribes have employed witch doctors or medicinemen, and human sacrifices to bring clouds and rainfall during periods of drought and to drive away rain clouds during flooding episodes. Numerous examples exist where modern man has shot cannons, fired rockets, rung bells, etc. in attempts to modify the weather (Changnon and Ivens, 1981).
—————
|
Seeding of supercooled cumulus clouds produced more controversial results. Dry ice and silver iodide seeding experiments were carried outat a variety of locations with the most comprehensive experiments being over New Mexico. Based on four seeding operations near Albuquerque, New Mexico, Langmuir claimed that seeding produced rainfall over a quarter of the area of the state of New Mexico. He concluded that “The odds in favor of this conclusion as compared to the rain was due to natural causes are millions to one.” Langmuir was evenmore enthusiastic about the consequences of silver iodide seeding over New Mexico. The explosive growth of a cumulonimbus cloud and the heavy rainfall near Albuquerque and Santa Fe were attributed to the direct results of ground-based silver iodide seeding. In fact Langmuir concluded that nearly all the rainfall that occurred over New Mexico on the dry ice seeding day and the silver iodide seeding day were the result of seeding.
The claim by Langmuir was found to be false in an analysis by the Scientific Services Division of the Weather Bureau, represented by Ferguson Hall. Others chimed in with Hall also publishing in Science magazine: Gardener Emmons, NYU, Bernard Haurwitz, NYU, Hurd Willett, MIT, and George Wadsworth, MIT.
———————
Convinced that cloud seeding was a miraculous cure to all of nature’s evils, Langmuir and his colleagues carried out a trial seeding experiment of a hurricane with the hope of altering the course of the storm or reducing its intensity. On October 10, 1947, a hurricane was seeded off the east coast of the United States. About 102 kg of dry ice was dropped in clouds in the storm. Due to logistical reasons, the eye wall region and the dominate spiral band were not seeded. Observers interpreted visual observations of snow showers as evidence that seeding had some effect on cloud structure. Following seeding, the hurricane changed direction from a northeasterly to a westerly course, crossing the coast into Georgia. The change in course may have been a was the result of the storm’sinteraction with the larger-scale flow field. Nonetheless, General Electric Corporation became the target of lawsuits for damage claims associated with the hurricane.
In summary, Project Cirrus launched the United States and much of the world into the age of cloud seeding. The impact of this project on the science of cloud seeding, cloud physics research, and the entire field of atmospheric science was similar to the effects of the launching of Sputnik on the United States aerospace industry.
This discussion lacks mention of perhaps the most influential paper of all those that motivated cloud seeding throughout the world; that of Kraus and Squires (1947) reporting spectacular results from an Australia seeding experiment. The KS47 paper, appearing in the high-end journal, Science, purported that two drops of dry ice totaling 250 lbs. into a Cumulus congestus cloud at 23,000 feet spawned a massive, isolated storm that towered to possibly “40,000 feet,” lasted for hours and dropped heavy rains over “20 square miles.” This was the only cloud that appeared to respond so impressively to dry ice seeding on that flight day. The Kraus and Squires report was seen as evidence that drought might be alleviated with few hundred pounds of dry ice, and as a result, widespread cloud seeding took off as entrepreneurs hastily formed cloud seeding companies such as North American Weather Consultants, Atmospherics, Inc., Irving P. Krick Associates, among many others. The series of photographs of the apparent response of that Cumulus cloud to seeding continued to be published as proof what cloud seeding could do for almost 40 years after KS47 (e.g., Orville 1986).
Caveat Emptor:
HOWEVER, in independent dry ice seeding tests on Cumulus clouds by the U. S. Weather Bureau (Coons et al. 1949, Coons and Gunn 1951) no explosion of a cloud occurred as KS47 had described after seeding. Instead, they reported, the seeded Cumulus clouds generally sank back down after dropping up to 100 lbs. of dry ice into them. Also, they reported that natural precipitation had formed in similar Cumulus clouds in the vicinity. It wasn’t clear to Coons et al. that dry ice seeding had made a difference.
——————-
Chapter 2: The glory years of weather modification
The exploratory cloud seeding experiments performed by Langmuir, Schaefer, Kraus and Squires, and Project Cirrus personnel fueled a new era in weather modification research as well as basic research in the microphysics of precipitation processes, cloud dynamics, and small-scale weather systems, in general. At the same time commercial cloud seeding companies sprung up worldwide practicing the art of cloud seeding to enhance and suppress rainfall, dissipate fog, and decrease hail damage. Armed with only rudimentary knowledge of the physics of clouds and the meteorology of small- scale weather systems, these weather modification practitioners sought to alleviate all the symptoms of undesirable weather by prescribing cloud seeding medication. The prevailingview was “cloud seeding is good!”
We have seen that the pioneering experiments of Schaefer and Langmuir suggested that the introduction of dry ice or silver iodide into supercooled clouds could initiate a precipitation process. The underlying concept behind the static mode of cloud seeding is that natural clouds are deficient in ice nuclei.
2For an excellent, more technical review of static seeding, see Silverman (1986).
———————
The rime-splinter secondary ice crystal production process may not explain all the observations of high ice crystal concentrations relative to ice nuclei estimates but it is consistent with many of them. Still other processes, such as drop fragmentation during freezing (Korolev et al. 2004, Rangno and Hobbs 2005) and fragmentation of delicate ice particles (e.g.,Vardiman 1978) are not well quantified understood at this time may be operating operate in some cases of observed high ice crystal concentrations relative to ice nuclei concentrations.
The implication of these physical studies is that the “window of opportunity” for precipitation enhancement by cloud seeding is much smaller than was originally thought. Clouds that are warm-based and maritime have a high natural potential for producing precipitation. On the other hand, clouds that are cold-based and continental have reduced natural potential for precipitation formation and, hence, the opportunity for precipitation is much greater, although the total water available would be less than in a warm-based cloud.
This is consistent with the results of field experiments testing the static seeding hypothesis. The Israeli I and II experiments were quite successful in producing positive yields of precipitation in seeded clouds (Gagin and Neumann, 1981). The clouds that were seeded over Israel had relatively cold bases(5-8°C) and were generally continental such that there was little evidence of a vigorous warm rain process or the presence of large quantities of heavily rimed graupel particles.
This paragraph was copied verbatim from the first edition of this book, CP95. This paragraph should not have been copied and pasted into CP07 because it was not even valid at the time of CP95. For example, strong evidence of ice multiplication and warm rain processes that undercut the “ripe for seeding” Cumulus descriptions that made the statistical successes of the Israeli experiments so credible (e.g., Silverman 1986) had appeared in 1988 (Rangno).
CP07 (and CP 95) were not aware of, or chose also not to cite, Gabriel and Rosenfeld (1990) which published the “crossover” result of seeding for Israeli II for the first time. Crossovers consist of combining the results of random seeding in two targets. The crossover evaluation had been mandated by the Israeli Rain Committee prior to Israeli II (Silverman 2001). The important result for Israeli II was -2% effect on rainfall, not statistically significant. Thus, Israeli II had not replicated Israeli I, as also concluded by Rangno and Hobbs (1995), and Silverman (2001). For comparison, the Israeli I crossover result had indicated a statistically significant 15% increase in rain (e.g., Wurtele 1971).
The null crossover result in Israeli II was due to apparent increases in rain due to seeding in the north target that were canceled out by indications of a whopping 15% decrease in rainfall on seeded days in the south target. Numerous questions about the “ripe for seeding” clouds of Israel would have been raised had the results for the south target of Israeli II been reported in a “timely manner” and not hidden from view until 1990.
CP07 acknowledge some of the evidence against the Israeli seeding successes later in this chapter, but not all. This is an improvement over CP95 that had not cited any counter evidence regarding those experiments. However, in CP07 the reader will get two interpretations of the same experiments in one book.
———
A number of observational and theoretical studies have also suggested that there is a cold temperature “window of opportunity” as well. Studies of both orographic and convective clouds have suggested that clouds colder than-25°C have sufficiently large concentrations of natural ice crystals that seeding can either have no effect or even reduce precipitation (Grant and Elliot, 1974; Gagin and Neumann, 1981; Gagin et al., 1985; Grant, 1986). It is possible that seeding such cold clouds could reduce precipitation by creating so many ice crystals that they compete for the limited supply of water vapor and result in numerous, slowly settling ice crystals which evaporate before reaching the ground. Such clouds are said to be overseeded.
There are also indications that there is a warm temperature limit to seeding effectiveness (Grant and Elliot, 1974; Gagin and Neumann, 1981; Cooper and Lawson, 1984). This is believed to be due to the low efficiency of ice crystal production by silver iodide at temperatures approaching -4 °C and to the slow rates of ice crystal vapor deposition growth at warm temperatures. Thus there appears to be a “temperature window” of about -10°C to -25°C where clouds respond favorably to seeding (i.e., exhibit seedability).
CP07 were not aware of, or chose not cite, the extensive literature that contradicted the claims of Grant and Elliott (1974) concerning a cloud top temperature range in which cloud seeding is supposedly viable. This led CP07 to “cut and paste from CP95 that there is a viable cloud seeding window for clouds (with tops) having a temperature range from -10°C to -25°C. We note that CP07 do not use the term “cloud top” in this discussion, but that’s what the papers they reference are referring to, not just a temperature range within a cloud deck.
That a viable cloud top temperature seeding window of -10°C to -25°C exists as CP07 purport was dealt a severe blow in the Rockies by several papers in which high (10s to 100 per liter) concentrations of ice particles were reported in wintertime clouds with tops >-25°C including even a contribution from Grant (leader of the Climax, CO, experiments) in Hobbs (1969).
Some of the overlooked papers are: Auer et al. (1969), Cooper and Vali (1981), Marwitz et al. 1976, Cooper and Marwitz 1980, and ground ice crystal concentration reports by Vardiman 1978, and by Vardiman and Hartzell (1976) in support of the Colorado River Basin Pilot Project, and by Grant et al.’s 1982 airborne study that reported no correlation with cloud top temperatures and ice particle concentrations in stable orographic clouds.
So, the “window of opportunity” for cloud seeding was, at the time of CP95 and in CP07, known to be much more limited compared to what they were in their books. I brought some of this counter evidence against this purported seeding window to the attention of the first author of CP07 in an email in 1997 to no avail.
There was also a serious drawback to the Grant and Elliott (1974) study that CP95 and CP07 depended upon; they did not measure cloud top temperatures in the projects they evaluated. Instead, Grant and Elliott used constant pressure surfaces as proxies of cloud tops. The use of a constant pressure temperature was shown to be invalid as an index of cloud top temperatures on several occasions (Rangno 1972, 1986, Bartlett et al. 1975, Elliott et al. 1976, Hobbs and Rangno 1979, Mielke 1979, Hill 1980). These studies also went under the CP95 and CP07 “radar.”
There is a similar drawback to the Gagin and Neumann (1981) study; they measured radar tops, not cloud tops; the latter are higher, and their claim of having measured the top of “every cell” in Israeli II, was not true. It was impossible to see the tops of cells in the north end of the north target of Israeli II, nor even very close by when appreciable rain was falling on the Israel Meteorological Service’s 3-cm wavelength radar that Gagin and Neumann used for this purpose (personal communication, 1987, Mr. K. Rosner, the Israeli experiments’ chief meteorologist).
—————–
Physical studies and inferences drawn from statistical seeding experiments suggest there exists a more limited window of opportunity for precipitation enhancement by the static mode of cloud seeding than originally thought. The window of opportunity for cloud seeding appears to be limited to:
- clouds that are relatively cold-based and continental;
- clouds having top temperatures in the range -10°C to -25°C;
The above sentence about cloud top temperatures is profoundly incorrect because neither in CP95 and in CP07 were the authors aware of, or chose not to cite, the many publications concerning wintertime ice in clouds that contradict the assertion that this temperature range presents a viable cloud seeding window.
- a timescale limited by the availability of significant supercooled water before depletion by entrainment and natural precipitation
We must also recognize that implementing a seeding experiment or operational program that operates only in the above listed windows of opportunity is extremely difficult and costly. It means that in a field setting we must forecast the top temperatures of clouds to assure that they fall within the -10 – 5°C to perhaps to -15°C – 25°C temperature window.
The temperature range above is adjusted to reflect the modern knowledge of ice in clouds in the Rockies and the improved AgI formulations that work more efficiently at higher temperatures. However, at the higher cloud top temperatures in this range (-10°C to -15°C) in the high barriers of the Rockies the wintertime clouds tend to be too thin for appreciable modification potential.
In summary, the static mode of cloud seeding has been shown to cause the expected alterations in cloud microstructure including increased concentrations ofice crystals, reductions of supercooled liquid water content, and more rapid production of precipitation elements in both cumuli (Cooper and Lawson, 1984) and orographic clouds (Reynolds and Dennis, 1986; Reynolds, 1988; Super and Boe, 1988;Super and Heimbach, 1988; Super et al., 1988).
The documentation of increases in precipitation on the ground due to static seeding of cumuli, however, has been far more elusive with the Israeli experiment (Gagin and Neumann, 1981) providing the strongest evidence that static seeding of cold-based, continental cumuli can cause significant increases of precipitation on the ground.
The above statement shouldn’t have been copied and pasted from CP95 into CP07 considering the published contrary evidence available that was available prior to both editions.
The evidence that orographic clouds can cause significant increases in snowpackis far more compelling, particularly in the more continental and cold-based orographic clouds (Mielke et al., 1981; Super and Heimbach, 1988).
By citing Mielke et al. 1981, CP07 indicate a lack of awareness of the published literature regarding the experiments conducted at Climax, CO. Mielke et al. 1981 stratified their results by 500 mb temperatures, which have no meaning for cloud properties as Mielke (1979) himself reported. For those of us who follow the cloud seeding literature, this was a bizarre occurrence in peer-reviewed literature; the official journal reviewers and journal editor who were responsible for this science oxymoron, take note.
A dark cloud (no pun intended) was cast on all of the Colorado State University cloud seeding reports when Rangno and Hobbs (1987) used independent data reduced by NOAA personnel to evaluate the Climax experiments precipitation and upper-level temperature. CSU personnel had reduced the data in Climax II to speed reporting of its results. But this was contrary to the prior statements made by the experimenters about these having been independently made measurements (Mielke 1995).
Rangno and Hobbs (1987) found that the CSU errors created the replication of Climax I by Climax II . Moreover, Climax II’s early tainted success through it’s first two years (reported to the Bureau of Reclamation’s cloud seeding division in Grant et. al. 1969, helped spur the decision by the BOR to fund the massive seven million dollar Colorado River Basin Pilot Project as well as provide CSU and their consultants with about $500,000 to design that experiment in June 1968. The Climax II errors had a profound effect; Climax I had been virtually error-free.
Furthermore, Rhea (1983) uncited by CP 95 or by CP07, showed that the optimistic result by Mielke et al. (1981) for Climax II was due to a mistiming of precipitation gauge readings between the target and the control gauges, and not due to cloud seeding. Thus Climax II did not replicate Climax I as was widely believed. (Rhea 1983 was published AFTER he made revisions to his paper as had been suggested by Grant et al. in a “Comment” and “Reply” journal exchange behind the scenes. However, Grant et al. (1983) did not revise their original “Comment” after Rhea made his revisions, an act that misled readers when reading the published exchange in the journal. Rhea, in a private communication to me in 1986 considered the Grant et al. “Reply” a “smokescreen.”
But even these conclusions have been brought into question. The Climax I and II wintertime orographic cloud seeding experiments (Grant and Mielke; 1967; Chappell et al., 1971; Mielke et al., 1971, 1981) are generally acknowledged by the scientific community (National Academy of Sciences, 1975 1973; Tukey et al., 1978) for providing the strongest evidence that seeding those clouds can significantly increase precipitation.
Nonetheless, Rangno and Hobbs (1987, 1993) question both the randomization techniques and the quality of data collected during those experiments. They and concluded that the Climax II experiment failed to confirm that precipitation can be increased by cloud seeding in the Colorado Rockies when the NOAA-published precipitation and upper level data for the Climax II experiment was used to evaluate it. Even so, Rangno and Hobbs(1987) did show that precipitation may have been increased by about 10%in the combined Climax I and II experiments
CP07 did not read Rangno and Hobbs (1993), The 10% increase in snow CP07 assume occurred was due to Grant and Mielke (1967) building in a huge seeding effect in Climax I in the high 500 mb temperature storm category via the choice of controls mid-way through Climax I. This initial act caused the entire Climax experiments to suggest an ersatz 10% increase in snow due to seeding in the high 500 mb temperature category.Moreover, the 10% was not statistically significant via 1000 re-randomizations of the combined experiments CP07 refer (I. Gorodnoskya, personal communication, University of Washington Academic Computing Center, 1987, unpublished result).
Once the controls were hard-wired, no further indications of a seeding effect occurred as can be seen in the diagrams below from Rangno and Hobbs (1993). One can assume that Grant and Mielke were sincere in their belief that a large cloud seeding induced increase in snowfall was being produced at Climax when they chose their controls at the halfway point, but had they realized that it had ended after their choices, the story of Climax I would have turned out far differently.
should be compared, however, to the original analyses by Grant et al.(1969) and Mielke et al.(1970, 1971)which indicated greater than 100% increase in precipitation on seeded days in the high 500 mb temperature category for Climax I and 24% for Climax II. Subsequently, Mielke (1995) explained a number of the criticisms made by Rangno and Hobbs regarding the statistical design of the experiments, including revealing that CSU personnel, and not independent NOAA ones, as had been claimed on several occasions, were responsible for the errors in the precipitation and upper- level data that created a false replication of Climax I, in particular the randomization procedures, the quality and selection of target and control data, and the use of 500 mb temperature as a partitioning criteria. It is clear that the design, implementation, and analysis of this experiment was a learning process not only for meteorologists but statisticians as well.
PS to the reader: It was not a learning process for the CSU experimenters as claimed above.
Professor Grant was informed in my presence by three different people (I was not one of them) on three occasions in the early 1970s while I was the Assistant Project forecaster for the CRBPP that the stratifications by 500 mb temperatures by he and his group did NOT index cloud top ones as he was claiming. These refutations of his claims were based on the statements of the prior Park Range Project contractor in Rhea et al. (1969), and on the rawinsonde data from the on-going Colorado River Basin Pilot Project (e.g., Rangno 1972, Elliott et al. 1976). 500 mb and cloud top temperatures are, in general, poorly correlated.
So, Grant stopped claiming they were strongly related after he received this information in the early 1970s and learned from it?
No.
Grant continued to claim (as in Grant and Elliott 1974, Grant and Cotton 1979, Grant 1986) that the 500 mb temperature was representative of cloud top temperature (see Table 8 in Grant and Elliott 1974). Strangely believe it, Mielke et al. with Grant as a co-author (1981) again stratified results of seeding in Climax I and II by 500 mb temperatures which Mielke himself knew had no physical meaning re cloud tops or cloud microstructure!
Moreover, the CSU experimenters repeatedly and falsely claimed that a graduate student, Furman (1967), had established a relationship between 500 mb and cloud top temperatures (e.g., as in Grant and Elliott 1974). Furman (1967) says nothing about such a relationship in his master’s thesis that consisted of but three days of vertically pointed 3-cm wavelength radar at Climax (Hobbs and Rangno 1979). Another CSU graduate student, Hjermstad (1970), to refer to Furman’s study as, “scant” in coverage.
To the outside community, the experimenters were presenting quite a different picture of what Furman had done. What do we make of this?
For a naive, idealistic newbie into the weather modification/cloud seeding scene like me in the early 1970s in Durango, CO, this was amazing and troubling stuff to experience! CP07 (CSU folk) try to minimize what happened, I think, by claiming it was a “learning process” when what actually happened was so counter to what we think of as “science”; i. e., that scientists change their minds when new facts come in that contradict their hypotheses.
The results of the many reanalyses of the Climax I and II experiments have clearly “watered down” the overall magnitude of the possible increases in precipitation in wintertime orographic clouds. Furthermore, they have revealed that many of the concepts that were the basis of the experiments are far too simplified compared to what we know today. Furthermore, many of thecloud systems seeded were not simple “blanket-type orographic clouds” but were part of major wintertime cyclonic storms that pass through the region.As such, there was a greater opportunity for ice multiplication processes and riming processes to be operative in those storms, making them less susceptible to cloud seeding.
The above is a good summary.
Two other randomized orographic cloud seeding experiments, the Lake Almanor Experiment (Mooney and Lunn, 1969) and the Bridger Range Experiment (BRE) as reported by Super and Heimbach (1983) and Super (1986) suggested positive results. However,these particular experiments used high-elevation silver iodide generators, which increases the chance that the silver iodide plumes get into the supercooled clouds. Moreover, both experiments provided physical measurements that support the statistical results (Super, 1974; Super and Heimbach, 1983, 1988). Using trace chemistry analysis of snowfall for the Lake Almanor project, Warburton et al.(1995a) found particularly good agreement with earlier statistical suggestions of seeding-induced snowfall enhancement with cold westerly flow. They concluded that failure to produce positive statistical results with southerly flow cases was likely related to seeding mis-targeting of the seeded material.
The reader should be aware that the results of the second randomized Lake Almanor experiment were not fully reported. The effect of seeding in the so-called “cold westerly” cases where a large seeding effect was suggested in the first Lake Almanor experiment, was omitted in the analyses of the second experiment (Bartlett et al. 1975).
Omittted results are always a sign of concern as happened in the Israeli II experiment. Also of concern, no one has reanalyzed the first Lake Almanor experiment with its overly large percentage increases in snow, also of “concern.” They don’t seem realistic to me, an expert in ersatz seeding reports and in cloud microstructure. Someone gimmee that list of random decisions for Lake Amanor I and I’ll check it out!
These two randomized experiments strongly suggest that higher-elevation seeding in mountainous terrain can produce meaningful seasonal snowfall increase.
Independent scrutiny is needed for both of those experiments to strengthen this conclusion.
We noted above, that the strongest evidence of significant precipitation increases by static seeding of cumulus clouds came appeared to come from the Israeli I and II experiments…until Gabriel and Rosenfeld (1990) reported the full results of Israel II.
Rosenfeld and Farbstein(1992) suggested that the differences in seeding effects between the north and south target areas during Israeli II that were reported by Gabriel and Rosenfeld (1990) is was the result of the incursion of desert dust into the cloud systems. They argue that the desert dust contains more active natural ice nuclei and that they can also serve as coalescence embryos enhancing collision and coalescence among droplets. Together, the dust can make the clouds more efficient rain-producers and less amenable to cloud seeding.
Note: This “divergent effects” hypothesis began to gain early traction in the scientific community (e.g., J. Simpson, 1989).
Even these experiments have come under attack by Rangno and Hobbs doubted the Rosenfeld and Farbstein (1992) claims, and he launched a reanalysis of both Israeli experiments in 1992 that was published in 1995 (Rangno and Hobbs). That publication drew numerous critical comments from seeding partisans in 1997 From their reanalysis of both the Israeli I and II experiments, they Rangno and Hobbs argued demonstrated that the appearance of seeding-caused increases in rainfall in the Israeli I experiment was due to “lucky draws” or a Type I statistical error as had first been “red flagged” in Wurtele (1971) for Israeli I due to the highest statistical significance on seeded days in that experiment having been located in a Buffer Zone that was meant to be unseeded between the two targets. That it was largely unseeded was stated by the Israeli I chief meteorologist in Wurtele’s paper, Mr. Karl Rosner. Wurtele’s paper should have been cited in CP95 and CP07.
Gabriel and Rosenfeld (1990) and Furthermore, they Rangno and Hobbs argued showed that during Israeli II, naturally heavy rainfall fell over a wide region that included both targets on north target seeded days, with Rangno and Hobbs expanding the analysis of Gabriel and Rosenfeld (1990) to include Lebanon and Jordan. The widespread heavier rainfall on north target seeded days gave the appearance that seeding caused increases in rainfall over the north target area, but since the seeded days in the north were the control days for the south target and more ordinary storms happened in the south target on its seeded days, created the impression that seeding had decreased rainfall there. Namely, the presence of “dust/haze” as claimed by Rosenfeld and Farbstein (1992) had nothing to do with the outcome of the Israeli II experiment as evaluated by Rangno and Hobbs (1995). The lower natural rainfall in the region encompassing the south target area gave the appearance that seeding decreased rainfall over that target area:
Not cited by CP07 in this segment is Silverman (2001) in his major review of numerous static glaciogenic seeding experiments, that included the Israeli experiments. He concluded, as did Rangno and Hobbs (1995), that the two Israeli experiments no longer were credible in proving that rain had been increased by cloud seeding.
We argued above that the “apparent” success of the Israeli seeding experiments was due to the fact that they are more susceptible to precipitation enhancementby cloud seeding. This is because numerous studies (Gagin, 1971, 1975, 1986; Gagin and Neumann, 1974) have had shown that the clouds over Israel are continental having cloud droplet concentrations of about 1,000 cm-3 and that ice particle concentrations are generally small until cloud top temperatures are colder· than-14°C. Furthermore, there is was little evidence found in those early studies for ice particle multiplication processes operating in those clouds.
See Rangno and Hobbs (1988) for a critique of those early studies by Professor Gagin listed above and why they were unrepresentative of most Israeli clouds. Also see Rangno (1988) for evidence that those early studies were, indeed, highly erroneous as was verified on numerous occasions in later Israeli cloud studies using aircraft (e.g., Levin et al. 1996) and satellite data.
Rangno (1988) and Rangno and Hobbs (1995) also reported on observations of clouds over Israel that strongly suggested containing they contain large supercooled droplets and quite high ice crystal concentrations at relatively warm temperatures. In addition, Levin et al.(1996) corroborated the Rangno (1988) and Rangno and Hobbs (1988, 1995) inferences when they found high ice particle concentrations, 10s to hundreds per liter, in convective clouds with tops no colder than -13°C. presented evidence of active ice multiplication processes in Israeliclouds. This further erodes the perception that the clouds over Israel were as susceptible to seeding as originally thought.
Naturally, the Rangno and Hobbs (1995) paper generated quite a large reaction in the weather modification community. The March issue of the Journal of Applied Meteorology contained a series of comments and replies related to their paper (Ben- Zvi, 1997; Dennis and Orville, 1997; Rangno and Hobbs, 1997a,b,c,d, e, the most important of those “Replies”; Rosenfeld, 1997; Woodley, 1997). These comments and responses clarify many of the issues raised by Rangno and Hobbs (1995). Nonetheless, the image of, what was originally thought of as the best example of the potential for precipitation enhancement of cumulus clouds by static seeding has become considerably tarnished.
What have we learned from this chapter? Caveat emptor concerning reports by those who conducted a “successful” cloud seeding experiments later telling us how ripe with seeding potential those clouds were.
Amen. Thanks, guys, for this concluding remark largely due to the present writer’s work and skepticism. This conclusion should have been placed earlier so the reader is not getting two versions of the of the Israeli experiments in having increased rain.
Chapter 3: The fall of the funding science of weather modificationby cloud seeding
For nearly two decades vigorous research in weather modification was carried out in the United States and elsewhere. As shown in Fig. 3.1, federalfunding in the United States for weather modification research peaked in the middle 1970s at nearly $19 million per year. Even at its peak, funding for weather modification research was only 6% of the total federal spending in atmospheric research (Changnon and Lambright, 1987) and this amount includedconsiderable support for basic research on the physics of clouds and oftropical cyclones. Nonetheless, research funding in cloud physics, cloud dynamics, and mesoscale meteorology was largely justified based on its application to development of the technology of weather modification.Research on the basic microphysics of clouds particularly benefited fromthe political and social support for weather modification. •
By 1980, the funding levels in weather modification research began to fall appreciably and by 1985 they had fallen to the level of $12 million. After 1985, funding in weather modification research became so small and fragmented that no federal agency kept track of it. Currently the Bureau of Reclamation has onlyabout
$0.25 million per year that can be identified as weather modification. They have operated a program in Thailand that was supported by the Agency for International Development. Basic research in the National Science Foundation that can be linked to weather modification is on the order of $1 million. Likewise, the Department of Commerce has no budgeted weather modification program but has supported a cooperative state/federal program at about the $3.5 million level. This on again- off again “pork barrel” program is supported by congressional write-insrather than a line item in the National Oceanic and Atmospheric Administration (NOAA) budget. In FY-2003, the Bureau of Reclamation administered this program,but no such funds were earmarked for either the Bureau of Reclamation or NOAA inFY-2004 or FY-2005. In this program, states having strong political lobbying support for weather modification are earmarked for support in this program.Overall the total federal program for weather modification in the United Statesis on the order of 10% of its peak level in the middle 1970s. What caused this virtual crash in weather modification research?
Changnon and Lambright (1987) listed the following reasons for this reduction in funding:
- poor experimental designs;
Changnon and Lambright, as did CP95 and CP07, did not discuss the Bureau of Reclamation’s Colorado River Basin Pilot Project, the nation’s largest, best-planned, and costliest ever randomized orographic cloud seeding experiment based on the work of CSU cloud seeders. The surviving author, Lambright, did not remember why in his reply to a recent email query by this writer.
- widespread use of uncertain modification techniques;
- inadequate management of projects;
- unsubstantiated Faulty claims of success, published in the peer-reviewed literature, ones that should have been caught in the peer-review of manuscripts.
The “fall” described by CP95 and again in CP07 didn’t have to happen because with solid reviews; there would have been no rise! If you would like to read about how to stop faulty cloud seeding claims from appearing in the peer-reviewed literature, go here:
https://cloud-maven.com/cloud-seeding-and-the-journal-barriers-to-faulty-claims-closing-the-gaps/
- inadequate project funding; and
- wasteful expenditures
Changnon and Lambright concluded that the primary cause of the rapid decline in weather modification funding was the lack of a coordinated federal research program in weather modification. However, there are other factors that also must
- weather modification was oversold to the scientific community, public and legislatures due to peer-reviewed literature that appeared to show that cloud seeding had worked;
- demands for water resource enhancement declined due to an abnormal wet period;
Clearly there is a great need to establish a more credible, stably funded scientific program in weather modification research, one that emphasizes the need to establish the physical basis of cloud seeding rather than just a”black box” assessment of whether or not seeding increased precipitation. We need to establish the complete hypothesized physical chain of responses to seeding by observational experiments and numerical simulations. We also needto assess the total physical, biological, and social impacts of cloud seeding, or what we call taking a holistic approach to examining the impacts of cloud seeding similar to those conducted during the latter stages of the Colorado River Basin Pilot Project (e.g., Marwitz et al. 1976). E.K. Bigg, for example (Bigg, 1988, 1990b; Bigg and Turton,1988) suggested that silver iodide seeding can trigger biogenic production ofsecondary ice nuclei. His research suggests that fields sprayed with silveriodide release secondary ice nuclei particles at 10-day intervals and thatsuch releases could account for inferred increases in precipitation 1-3 weeksfollowing seeding in several seeding projects( e.g., Bigg and Turton, 1988).If Bigg’s hypothesis is verified, an implication of biogenic production ofsecondary ice nuclei is that many seeding experiments have thus been contaminated such that the statistical results of seeding are degraded. This effect would be worst in randomized cross-over designs and in experiments inwhich one target area is used and seed days and non-seed days are selectedover the same area on a randomized basis. Thus, not only is the weathermodification community faced with very difficult physical problems and largenatural variability of the meteorology, but they also are faced with thepossibility of responses to seeding through biological processes. (Bigg’s work is not credible to this reviewer.)
We shall see later that scientists dealing with human impacts on global change are also faced with very difficult physical problems, large natural variability of climate, and the possibility of complicated feedbacks through the biosphere. There is also a great deal of overselling of what models candeliver in terms of “prediction” of human impacts over timescales of decades or centuries.
End of “review and enhancement” of CP07’s weather modification chapters.
———–We interrupt this review for a brief personal presentation———
My self-funded trip to Israel (Rangno 1988) was to expose what I thought were faulty, ripe for cloud seeding cloud reports that were the foundation for the belief that seeding had increased rain. My findings were first confirmed by Levin (1992), reported again in the journals by Levin et al. (1996) and verified in ensuing years several times in satellite imagery and in other airborne measurements. Spiking fubball here, of course. How often do researchers go to another’s lab on their own expense and tell him, “I don’t believe your results? Show them to me.” Cost me a lot of money, but it was worth it. At the time I went, the Israeli cloud reports had gone unquestioned (e.g., Silverman 1986).
Just before I hopped on a plane to Israel, my lab chief, Professor Peter V. Hobbs described me as “arrogant” for thinking I knew “more about the clouds of Israel than those who studied them in their backyard.” But when he submitted my paper to the Quart. J. Roy. Meteor. Soc., 12 January 1987, with a copy to Professor A. Gagin, Professor Hobbs wrote to Gagin that what I was describing in my paper was the same as he thought about the clouds of Israel (that they were not as Professor Gagin was describing them).
Professor Hobbs was not being truthful. I often wonder many times Professor Hobbs might have said that to others, that what I was reporting was what he had thought all along, robbing me of my insight, my historical trip and year long effort all paid for from my savings. When I confronted him about this, he sent me a memo that said not to expect a job in his group in the future.
But, he did hire me back to the job I loved via the magic of reconciliation (mostly) over past wrongs!
———————–
The references cited in this “Review and Enhancement” that do not appear in CP07:
Auer, A. H., D. L. Veal, and J. D. Marwitz, 1969: Observations of ice crystalsand ice nuclei observations in stable cap clouds. J. Atmos. Sci., 26, 1342-1343.
Bartlett, J. P., Mooney, M. L., and Scott, W. L., 1975: Lake Almanor Cloud Seeding Program. Special Regional Weather Modification Conference, San Francisco, CA, 106-111.
Coons, R. D., and R. Gunn, 1951: Relation of artificial cloud-modification to the production of precipitation. Compendium of Meteorology, Amer. Meteor. Soc., Boston, MA. 235-241.
Coons, R. D., E. L. Jones, and R. Gunn, 1949: Artificial production of precipitation. Third Partial Report: Orographic Stratiform Clouds–California, 1949. Fourth Partial Report: Cumuliform Clouds–Gulf States, 1949. U. S. Weather Bureau Res. Paper No. 33, Government Printing Office, Washington, 46 pp.
Gabriel, K. R., and D. Rosenfeld: The second Israeli rainfall stimulation experiment: analysis of rainfall on both target areas. J. Appl. Meteor., 29, 1055–1067, 1990.
Cooper, W. A., and J. D. Marwitz, 1980: Winter storms over the San Juan mountains. Part III. Seeding potential. J. Appl. Meteor., 19, 942-949.
Cooper, W. A., and G. Vali, 1981: The origin of ice in mountain cap clouds.J. Atmos. Sci., 38, 1244-1259.
Furman, R. W., 1967: Radar characteristics of wintertime storms in the Colorado Rockies. M. S. thesis, Colorado State University, 40pp
Grant, L. O., DeMott, P. J., and R. M. Rauber, 1982: An inventory of icecrystal concentrations in a series of stable orographic storms. Preprints, Conf. Cloud Phys., Chicago, Amer. Meteor. Soc. Boston, MA. 584- 587.
Grant, L. O., J. G. Medina, and P. W. Mielke, Jr., 1983: Reply to Rhea 1983. J. Appl. Meteor. and Climate, 22, 1482-1484.
Grant, L. O., Chappell, C. F., Crow, L. W., Mielke, P. W., Jr., Rasmussen, J. L., Shobe, W. E., Stockwell, H., and R. A. Wykstra, 1969: An operational adaptation program of weather modification for the Colorado River basin. Interim report to the Bureau of Reclamation, Department of Atmospheric Sciences, Colorado State University, Fort Collins, 98pp.
Hill, G. E., 1980: Reexamination of cloud-top temperatures used as criteria of cloud seeding effects in experiments on winter orographic clouds. J. Climate Appl. Meteor., 19, 1167-1175.
Hobbs, P. V., 1969: Ice multiplication in clouds. J. Atmos. Sci., 26, 315-318.
Hobbs, P. V., and Rangno, A. L., 1979: Comments on the Climax randomized cloud seeding experiments. J. Appl. Meteor., 18, 1233-1237.
Hobbs, P. V., Lyons, J. H., Locatelli, J. D., Biswas, K. R., Radke, L. F., Weiss, R. W., Sr., and A. L. Rangno, 1981: Radar detection of cloud-seeding effects. Science, 213, 1250-1252.
Korolev, A. V., M. P. Bailey, J. Hallett, and G. A. Isaac, and, 2004:Laboratory and in situ observation of deposition growth of frozen drops. J. Appl. Meteor., 43, 612-622.
Kraus, E. B., and P. Squires, 1947: Experiments on the stimulation of clouds to produce rain. Nature, 159, 489-490.
Levin, Z., 1992: The role of large aerosols in the precipitation of the eastern Mediterranean. Paper presented at theWorkshop on Cloud Microphysics and Applications to Global Change, Toronto. (Available from Dept. Atmos. Sci., University of Tel Aviv). No doi available
Levin, Z., E. Ganor, and V. Gladstein, 1996: The effects of desert particles coated with sulfate on rain formation in the eastern Mediterranean. J. Appl. Meteor., 35, 1511-1523.
Marwitz, J. D., Cooper, W. A., and C. P. R. Saunders, 1976: Structure and seedability of San Juan storms. Final Report to the Bureau of Reclamation,University of Wyoming, Laramie, WY, 324 pp.*
Rangno, A. L., 1972: Case study on some characteristics of the specially monitored storm episodes within the Colorado River Basin Pilot Project. Special Project Report to the Bureau of Reclamation, 105pp.
Rangno, A. L., 1988: Rain from clouds with tops warmer than -10 C in Israel. Quart J. Roy. Meteorol. Soc., 114, 495-513.
Rangno, A. L., and P. V. Hobbs, 1988: Criteria for the development of significant concentrations of ice particles in cumulus clouds. Atmos. Res., 22, 1-13.
Rangno, A. L., and P. V. Hobbs, 1997e: Comprehensive Reply to Rosenfeld, Cloud and Aerosol Research Group, Department of Atmospheric Sciences, University of Washington, 25pp .https://carg.atmos.washington.edu/sys/research/archive/1997_comments_seeding.pdf
Rangno, A. L., and P. V. Hobbs, 2005: Microstructures and precipitation development in cumulus and small cumulonimbus clouds over the warm pool of the tropical Pacific Ocean. Quart. J. Roy. Meteorol. Soc., 131, 639-673.
Rhea, J. O., 1983: “Comments on ‘A statistical reanalysis of the replicated Climax I and II wintertime orographic cloud seeding experiments. J. Climate Appl. Meteor., 22, 1475-1481.
Simpson, J., 1989: Amer. Meteor. Soc. Banquet talk transcript on the occasion of her inauguration as president of that organization, October 4th.
Vardiman, L., 1978: The generation of secondary ice particles in cloudcrystal-crystal collisions. J. Atmos. Sci., 35, 2168–2180.
Vardiman, L., and C. L. Hartzell, 1976: Investigation of precipitating ice crystals from natural and seeded winter orographic clouds. Final Report to the Bureau of Reclamation, Western Scientific Services, Inc., 129 pp.
Wurtele, Z. S., 1971: Analysis of the Israeli cloud seeding experiment by means of concomitant meteorological variables. J. Appl. Meteor., 10, 1185-1192.
Concluding remark:
The entire CP07 reference list, consisting of several hundred references, is a great resource for research and demonstrated how knowledgable these two authors are. However, the list above also indicates how difficult a review of a topic is when the amount of literature that appears in journals today can bury you with important citations being missed.
Review and Enhancement of the Cloud Seeding Portion of the 1995 book, “Human Impacts on Climate and Weather”
I received a copy of this book in 1997 and went into a completely objective tantrum when I read the chapters on weather modification/cloud seeding. Here’s my unedited 1997 letter to one the authors, with, as usual, candid material and why I was upset. And, why wasn’t I asked to review this portion of their book? I coulda improved it. Still, I highly recommend this book overall. It is likely that the first author is responsible for the cloud seeding portion of this book.
1997 2-16-18 Cotton, to, critique of his book with Pielke_emotions in cloud seeding_ocr
The comments embedded in this article in red font is NOT a hindsight view, but comments that were appropriate at the time this book was submitted/published with the relevant and available missing citations listed.
Sincerely,
Art
PS: This has been made more difficult since in the intervening 30 odd years, I have become social friends with the two authors. What to do? Nothing, or go forward?
PPS: It will be interesting to see if CP95 carried out my admonishments in CP2007, the second edition of Human Impacts on Climate and Weather. Standby.
Review and Enhancement of a 1979 Review of Weather Modification
Oh, yeah, baby! 1979! The Sex Pistols with Johnny Rotten and punk bands like Black Flag were on the rise!
My belated review of “Weather Modification” has to be done, IMO: i.e., corrections and comments, to complete a peer-reviewed article that was published in the journal, Reviews of Geophysics and Space Physics, using literature available to the authors at the time of the article was submitted/published that they omitted or didn’t know about.
As an aside, omitting stuff happens a LOT in the domain of “weather modification/cloud seeding.” A recent example, to go into a minor rant, was in the peer-reviewed article by Benjamini et al. 2023 (J. Appl. Meteor. and Climate) who reported a null result of randomized cloud seeding in Israel. Benjamini et al. could not bring themselves to cite my 1995 article with Prof. Peter V. Hobbs (same journal), that concluded cloud seeding increases previously reported by Israeli scientists at the Hebrew University of Jerusalem were illusory. Thus, there was no reason to think cloud seeding would work in another randomized experiment. Sure, it’s painful for them to cite my work, but still….same old same old; omit, omit, omit. Science is not always what you think it should be!
As an expert in some elements that are addressed in the 1979 article, I am happy to be able to improve and clarify it for readers of historic material, should they find it, before the Grim Reaper drops by. Here is the full article, which overall is quite good, except for those areas I am intimately familiar with:
A review and enhancement of the “weather modification” review by Grant and Cotton_1979_one column
I am quite candid about WHY I am doing a series of these “reviews and enhancements” of historic material as you will read above. I tend to get carried away, and so all the minutiae I discuss might be “painful,” too. I do my best, though.
Sincerely,
Art
A Review and Enhancement of “A Critical Assessment of Glaciogenic Seeding of Convective Clouds for Rainfall Enhancement”
I submitted a long “review and enhancement” on this article by Dr. Bernard A. Silverman’s 2001 massive (14,000 word) review article in the Bulletin of the American Meteorological Society (BAMS) in March 2002. My “Comments” were too long, the Bulletin editor said, and so it never even went to peer review. And, he was right, it was too long.
Silverman’s long review article is here. It’s well worth reading and there is much we agreed on in those days, if anyone cares:
2001 Silverman Critical assessment 2001ocr
Most of my “review and enhancement” of Silverman’s excellent, unbiased article despite his pro-cloud seeding stance, concerns the Israeli clouds and cloud seeding experiments. I am an expert in that domain, having spent 11 weeks in Israel in 1986 studying their clouds, rain and sounding data with the full cooperation of the Israel Meteorological Service. The ensuing article was published in the J. Roy. Meteor. Soc. in 1988. The gist of it: the clouds of Israel were not being described correctly by the leaders of cloud the seeding experiments at the Hebrew University of Jerusalem. They described them as plump with cloud seeding potential when they were not.
In the early 1990s, I re-analyzed the Israel-1 and Israel-2 randomized experiments along with my co-author, Peter V. Hobbs, director of the Cloud and Aerosol Research Group at the University of Washington, Seattle. The article, which concluded that there had been no cloud seeding induced increases in rain in these experiments, was published in the J. Appl. Meteor. in 1995. Several cloud seeding-centric scientists commented on that paper in 1997 along with our “Replies.”
When I saw Dr. Silverman’s 2001 article and that he had misdescribed some of my own work, I went into a controlled rage of objectivity (haha), as scientists do from time to time, and decided to write a “Comment.” But, as I added more and more material, my “Comment” became an article in itself, even a “novella” of sorts. Eventually those 2002 “Comments” led to a full blown article: “The Rise and Fall of Cloud Seeding in Israel,” that manuscript submitted in 2017 and rejected by BAMS in 2019 (J. R. Fleming, private communication) after two split reviews. BAMS did not allow me to respond to the comments of the two reviewers or revise my manuscript.
Well, after all these years, I just re-read my critique of Silverman’s 2001 article and thought it had some merit for those of you interested in 1) cloud seeding, 2) Israel. So, here it is, as it was submitted with a couple of minor corrections and an update. Caution: the subject of scientific misconduct is broached. That didn’t help me in 2002, either, but I’ve left it in anyway.
———————————————————————
- Introduction
Silverman (2001, hereafter S01) is to be commended for attempting the prodigious task of making sense of all of the randomized cloud seeding experiments targeting cumuliform clouds during the past 40 years. In large measure he has succeeded and made a significant contribution to the field of weather modification. Still, some comments and clarifications are needed .
In his historical overview of cloud seeding, S01 should have mentioned the great effort that an independent agency, the U. S. Weather Bureau (USWB), made in attempting to replicate the early and often spectacular claims of seeding successes that began to appear in the late 1940s (e.g., Kraus and Squires 1947). The USWB (i.e., Coons et al. 1949; Coons and Gunn 1951) seeded dozens of summertime cumuliform clouds in Ohio and along the U. S. Gulf Coast that were similar to clouds in several of the projects examined by S01 with up to “60 lbs of dry ice per mile.”
They found no evidence of any particular seeding effect.
On the contrary, USWB observers found that the turrets usually dissipated or settled back in altitude after they had been seeded.
The USWB scientists also made an unexpected and important discovery: ice was already forming in clouds with tops as warm as –6°C before they had seeded them. The occurrence of ice was at far higher temperatures than was expected from ground ice nucleus concentration measurements on which seeding hypotheses were based. Those ground measurements suggested that ice would not form until clouds reached temperatures of –15° to –20°C. This stunning USWB observation was to be confirmed in more sophisticated aircraft measurements more than ten years later in Missouri during Project Whitetop (Koenig 1963).
The USWB’s independent seeding trials that showed little evidence of seeding having affected rain on the ground, crude as it was, stands tall today relative to the same conclusions about the seeding of cumulus clouds reached by S01 50 years later.
- Why weren’t seeding effects produced in the projects reviewed?
One would think that we know how to seed clouds successfully with indisputable scientific evidence for more rain on the ground after more than 50 years of attempting to do so. However, the production of seeding effects in rainfall on the ground have been preempted by three crucial cloud factors, particularly so in the projects examined by S01: 1) Due to the relatively warm cloud bases in the projects he examined, natural ice forms in, or can be expected to form, in the targeted clouds either within or soon after their summits ascend above –5° to –10°C, the temperature range where seeding might take place; 2) the formation of ice crystals is vastly increased in the temperature range of –2.5° to –8°C in clouds with warm bases once ice has formed due to an explosion of ice splinters and fragments caused by secondary ice-forming mechanisms associated with large drops; and 3) perhaps most surprising to the Bulletin reader: we don’t yet know what the true concentrations of ice particles are in clouds today due to past instrument limitations that have only recently been overcome.
Today’s knowledge concerning ice in clouds has been limited because of our inability to reliably measure ice crystals smaller than about 100 mm in maximum dimension. Thus, published concentrations, even as high as they have been, have had to necessarily omit the contribution of very small ice crystals to the total in the cumulus clouds that have been sampled (e.g., Koenig 1963; Rangno and Hobbs 1991, 1994; Levin et al. 1996). Early measurements with a new probe capable of measuring these small ice crystals suggest that the total ice crystal concentrations in clouds will be several times higher than previously thought (e.g., Lawson and Jenson 1998). Hence we are once again on the verge of learning what the USWB did more than 50 years ago: there is more natural ice in clouds than we imagined.
Therefore, the seeding experiments discussed by S01 were, in a sense, premature since the experimenters had made, and perhaps unavoidably, erroneously low estimates of the amount of natural ice that formed in clouds. In essence, they were pouring water into a river without knowing whether there was a flood already in progress.
This fact has been demonstrated by Stith et al. (2002) who found little liquid water in tropical cumuli above the –12°C level and but “traces” of liquid water by –18°C. The clouds studied by Stith et al. represent the kinds of clouds seeded in several of the projects reassessed by S01, those in tropical regions with warm cloud bases and moderate updrafts. The liquid water in the clouds studied by Stith et al. had been consumed by the explosive natural ice formation taking place in rising turrets at temperatures before they reached the –12°C level and that were glaciated by –18°C.
- Project Whitetop
Project Whitetop (Braham 1979) still remains as one of the most important, enigmatic, and well designed of all the randomized cloud seeding experiments carried out to this day and it was surprising that it was not mentioned by S01. It was worthy of discussion because of its design and results, and because of the seeding method employed, which probably constituted its only major design flaw.
- What made Project Whitetop so special?
Project Whitetop had the three critical attributes that characterize sound experimental design: the experiment was 1) randomized, 2) the target area and the specific rain gauges for evaluation purposes were identified before the experiment began, and 3) the results were evaluated contractually by those removed from the conduct of the experiment, the institution carrying it out, and the evaluators had no vested interests in cloud seeding.[1] The importance of these attributes in the conduct and evaluation of cloud seeding experiments cannot be overemphasized. We do not know from S01 which, if any of his reviewed experiments, had these essential design attributes.
Project Whitetop stirred great interest and controversy when the initial analyses following its conclusion suggested strong decreases in rainfall had been produced by cloud seeding over a wide area (e.g., Lovasich et al. 1969). Later analyses, however, found that rain was also less on seeded days over wide areas upwind of the seeding line as well and subsequently, most scientists now believe that the random draw was uneven and produced a false negative (Type II statistical error or “unlucky draw”) and seeding actually produced a null result overall (Braham 1979).
What went wrong in Project Whitetop? First, just as the USWB scientists had found in similar clouds more than 10 years earlier, natural ice was forming in the Project Whitetop clouds at shockingly high cloud top temperatures, between –5° and –10°C, and clouds glaciated (turned completely to ice) with great rapidity (Koenig 1963; Braham 1964). Without doubt, “ice multiplication” as this phenomenon was later dubbed by Hobbs (1969), seriously compromised the chances of creating more rain on the ground through cloud seeding. This is because the purpose of the seeding was to create more ice in clouds that were (erroneously) believed by the experimenters to have little ice.
The warm-based Missouri clouds, like those projects in tropical settings discussed by S01, and also like those in Israel, produce copious quantities of large drops (>23 mm diameter) and even precipitation-sized drops (>200 mm diameter) as they ascend past the freezing level, making them ultra-ripe for the onset of various natural ice multiplication mechanisms (e.g., Hallet and Mossop 1974; Mossop 1985; Hobbs and Alkezweeny 1968). In themselves, these mechanisms can produce effects similar to those produced by cloud seeding (Rangno and Hobbs 1991, 1994).
- Were the clouds seeded effectively in the experiments reviewed?
Again, Project Whitetop has something to say about the projects assessed by S01. Project Whitetop had, in retrospect, a key design flaw: the seeding method used. Instead of injecting the seeding agent into updraft regions of clouds upwind or over the target as would be done today, the three aircraft used in this experiment dispensed it in lines about 50-km long upwind of the target at a height that was just below the bases of the cumulus clouds that might have been forming in the area. Suitable cumulus clouds (building ones with, or about to have supercooled tops) with updrafts above the aircraft as they dispensed silver iodide was not a criterion for releasing the seeding agent in Project Whitetop.
Whether the silver iodide released by the aircraft in Project Whitetop ever got into suitable clouds, whether it did so at the right locations upwind or in the target, and in what concentrations if it did, was never established. In fact, the “patrol” seeding, as it is sometimes called, appeared to be relatively ineffective from the ground measurements of ice nuclei concentrations that were made downwind in the target (Bouqard 1963).
Thus, a crucial link in the chain of events in the seeding process in Project Whitetop was completely missing. The method of seeding suggests another reason why results of Project Whitetop were doubtful; the seeding hypothesis itself may never have been fully tested.
The questionable seeding method used in Project Whitetop was similar to the one that was adopted in the first experiment in Israel that was begun at about the same time as Whitetop, and this in turn affected the choice of seeding method used in the Puglia experiment (List et al. 1999) that S01 discussed. Patrol seeding was used in the Puglia experiment because it appeared to the Puglia design team that the first experiment in Israel had been a statistical success, and they wished to replicate exactly in their own experiment the seeding methodology that had apparently brought success in Israel. Whether, in fact, the first experiment in Israel was a success is now subject to doubt on several accounts (e.g., Rangno and Hobbs 1995, 1997a,b.)
The assumptions about dispersion made by the Project Whitetop design team also represents one of several recurring themes in cloud seeding experiments: an exaggerated view of dispersion, an accompanying lack of dispersion measurements prior to cloud seeding experiments, and an underestimate of the natural ice concentrations in the clouds to be seeded.
- The two randomized seeding experiments in Israel
The two randomized experiments carried out in Israel deserve special attention beyond that given by S01 because of their importance for a number of years in convincing the scientific community, even the most skeptical scientists, that randomized cloud seeding experiments had finally produced a measurable result (e.g., Mason 1980, 1982; Kerr 1982; Silverman 1986; Dennis 1989; Young 1993; Cotton and Pielke 1995). Indeed, many have believed that they were the onlyexperiments in cloud seeding that had demonstrated a seeding success among all those that have been conducted. In this context, S01 must be admired for his ability to move from one who has previously validated the results of the experiments in Israel to one who now believes as the author does that they did not prove cloud seeding effectiveness after all. However, some descriptions by S01 of the Rangno and Hobbs (1995, hereafter RH95) reexamination of these experiments are incomplete and require further discussion.
First, why did the experiments in Israel have such great credibility as successes to such a wide audience? This was because they seemed to have had, for a time anyway, all of the requisites for unambiguous scientific credibility: a sound design that included randomization, a choice of evenly spaced rain gauges (at least in the first experiment) that was limited to all of the gauges in a pre-existing recording rain gauge network, an apparent confirmation of a statistically significant result in a follow-up confirmatory experiment, and a sound cloud microstructural basis for believing that the statistical successes reported were achievable because the clouds were so deficient in ice.
For example, the experimenters reported over a period of many years that the clouds in Israel achieved rather great depths and low cloud top temperatures (to –21°C) while producing little ice or precipitation (i.e., Gagin and Neumann 1974, 1976, 1981; Gagin 1975, 1980, 1981, 1986, hereafter GN74, GN76, GN81 and G75, G80, G81, and G86, respectively). This left a wide window (–12° to –21° C) for seeding to initiate ice and precipitation in those clouds. The higher temperature mentioned above demarcated the highest temperature at which appreciable concentrations of the silver iodide seeding crystals would have begun to nucleate and the lower temperature where it was reported that the natural ice concentrations were high enough that seeding was not required to boost ice content.
With cloud bases regularly at 800 m or so above sea level at 5-9°C (GN74, G75), this meant, according to these reports, that there was a large population of non-precipitating or barely precipitating clouds moving into Israel from the Mediterranean Sea that were as much as several kilometers deep. Furthermore, in support of this picture, the effects of seeding, according to the experimenters, had been in duration of rainfall, not in intensity (G86), a fact compatible with the kind of seeding done. From these many reports, it all made sense to outside scientists.
It will surprise and trouble many readers who do not follow the cloud seeding literature closely that the reporting of the results of the experiments by those who conducted them was not apropos according to normal scientific expectations. This unfortunate element of these experiments, which necessarily impacts the reliability of the body of literature about them, is discussed in Section e.
- Seeding logistics and heterogeneities in the two experiments in Israel
Perhaps because the patrol method of seeding described in Section 4 had just been adopted by the United States in its major test of seeding in Project Whitetop, the experimenters in Israel, about to embark on their own major seeding trial at about the same time (in the late winter of 1960-1961), also chose this method. In fact, patrol seeding was used almost exclusively in the six years of the first experiment (National Academy of Sciences 1973, GN74).[2]
However, the experimenters in Israel used a seeding track that was 65-75 km long, or about 15-25 km longer than the one used in Project Whitetop (Gabriel 1967). Moreover, they had but a single twin-engine aircraft available to them to seed a longer experimental period over which rainfall was to be evaluated for seeding effects, a 24 h day vs. Project Whitetop’s 14 h experimental period. Most remarkably, seeding by this method was carried out for an average of only about 4 h of the 24 h experimental day and yet still seemed to have produced statistically significant results (Gabriel 1967; Wurtele 1971; GN74). The precipitation climatology of Israel, that makes this an astonishing fact, was discussed by RH95.
Not surprisingly, the experimenters themselves came to realize the inadequacy of coverage of the 4 h per day of patrol seeding by a single aircraft in the first experiment. When their second experiment began in the fall of 1969, they had added no less than 42 ground generators and a second aircraft with aircrews to man them 24 h a day (National Academy of Sciences 1973).
These new and greatly extended sources of seeding in the second experiment constituted an enormous heterogeneity in seeding coverage in the amount of seeding material released, and a shift in the methodology between the two experiments from solely airborne seeding to mainly ground seeding supplemented by airborne seeding. Yet, implausibly, according to the partial statistical reports of the experimenters, the enormous increase in seeding and the shift in methodology produced virtually the same seeding result as in the first experiment, a 13% increase in rainfall in the North target (e.g., GN81) compared with the 15% overall increase in both targets of the first experiment when only seeding 4 h per day took place.
- The cloud microstructure of Israel does not make a case for seeding having produced statistically significant results in the experiments
A second factor that makes it implausible that the 4 h of seeding produced the statistical results reported in the first experiment, or effects in the second is that the clouds of the eastern Mediterranean are, in fact, largely unsuitable for seeding due to high concentrations of ice and the onset of ice at slightly supercooled cloud top temperatures (e.g., Rangno 1988; Rangno and Hobbs 1988, 1995, 1997a,b; Levin et al. 1996).
However, S01 stated that a “fraction” of the clouds of Israel do not correspond to those required for seeding to be effective. S01, by using the phrase “a fraction of” the clouds may have been “unsuitable” for seeding, presumably means that some of the clouds making landfall on the Israeli coastline where they were to be seeded already contained high natural ice crystal concentrations and thus had no seeding potential, or had tops that were too warm for seeding to be effective, or were stratiform in nature with no updrafts below them to draw the seeding material upward as was noted in RH95.
However, the word “fraction”, as used by S01, is ambiguous and may inadvertently mislead Bulletin readers by suggesting that it means “a small amount of.”
In fact, it would be a rare day in which high ice concentrations are not observed in mature and aging cumuliform clouds with tops >–14°C. Levin et al. (1996) gathered ice concentrations in clouds with tops warmer than about –14°C on several rather ordinary shower days in Israel and found tens to hundreds per liter in those clouds. The author directs the reader to papers that discuss the cloud microstructure of Israel (Rangno 1988, 2000; Rangno and Hobbs 1988, 1995, 1997a,b).
2. The stratifications of seeding effects by the cloud top temperatures in the second experiment are unreliable
S01 repeats the cloud top temperature stratifications that were reported on numerous occasions by the experimenters as having strongly partitioned seeding effects in the second experiment. The experimenters reported, for example, that the maximum seeding effect in the second experiment was a 46% increase in rainfall compared with control days when the cloud top temperature was in the range of –15° to –21°C (e.g., GN81).
However, the cloud top temperature stratifications by the experimenters are unreliable for several reasons and should not be quoted (RH95). Also, in view of the tens to hundreds per liter ice particle concentrations found in the clouds of Israel with tops warmer than about –14°C on rather ordinary showery day situations (Levin et al. 1996), combined with the discovery that rain routinely falls from clouds with tops warmer than –10°C (Rangno 1988), it is no longer scientifically credible that the strongest seeding effect was produced in clouds that had top temperatures between –15° and –21°C where excessive natural ice crystal concentrations already exist.
3. Discussion of statistical issues: Israel-1
S01 stated that RH95 concluded that the results of the first experiment were due to a false positive or Type I statistical error based solely on evidence reported by Wurtele (1971). RH95 performed a “cradle to grave” analysis of the first experiment and several significant factors, besides the fact that the greatest apparent seeding effect was in a region that was avoided by the seeding aircraft, led us to our conclusion. We direct the reader to our paper and subsequent discussions of this issue (Rangno and Hobbs 1995, 1997a, 1997b; Rosenfeld 1997).
On the other hand, S01 himself offers no explanation in his evaluation of the first experiment about why the statistical significance was highest in the little-seeded Buffer Zone (BZ) that lay between the two targets. In ignoring this fact S01 does not heed the same large “red flag” that should have raised skepticism about the statistically significant results in the first experiment more than thirty years ago.
No less than the chief meteorologist for the first experiment[3] stated, with notable candor in Wurtele (1971), that the BZ could only have been seeded 5-10% of the time, a conclusion sustained in RH95. Moreover, GN74, themselves puzzling over this same statistical anomaly in the BZ, concluded that relying on a seeding argument for the BZ anomaly was weak (though they did it anyway) since the statistical significance in the BZ also accrued on seeded days in which the lone seeding aircraft did not even fly!
It would be interesting to learn what knowledge S01 has developed to refute these assessments by the experimenters themselves.
b. Israel 2
Two Type I statistical errors in a row (lucky draws) is a slim possibility as some have noted concerning RH95. However, two false positives in a row did not occur in the Israeli experiments when the same crossover scheme was used to evaluate both of them as was called for in the a priori design (Gabriel and Rosenfeld 1990, Silverman 2001).
Second, when using the South target gauges as the control for the North target area, as was also specified in the a priori design (e.g., GN74), a null result for the second experiment was produced again (Gabriel and Rosenfeld 1990; RH95).
However, in a third design component of the second experiment, a statistically significant result was evinced when a few rain gauges in a coastal plain upwind of the North target area were used to assess seeding effects in the target (GN81). The statistical significance so obtained in the third of the three design evaluation components comprises the second statistically significant result in a row that S01 refers to. Are you following me, reader?
At first glance, perhaps the achievement of any statistical significance in even one of the three design components is an impressive result even if it is outweighed by two null results (both of which went unreported at the time).
However, in a wider analysis than that performed by the experimenters, or by Gabriel and Rosenfeld (1990), RH95 found comparable or even heavier rain that fell on the seeded days in the North target area also fell over a wide region in and outside of northern Israel; namely, in central and southern Lebanon, western Jordan, and in Israel south of the North target area including Jerusalem itself! Ironically, the experimenters had their offices in Jerusalem and were somehow oblivious to the heavy rain they were receiving when they seeded the North target area some 100 km to the north.
The single exception to this regional pattern of markedly heavier rain on North target area seeded days was in the coastal plain upwind of the North target area that had been pre-selected as a control zone.
The widespread regional pattern of bias in rainfall discovered by RH95 on North target seeded days suggests to meteorologists that the random draw was flawed or compromised by a bias in the weather systems that favored heavier rain over a synoptic scale region on North target area seeded days.
The second point that weakens the statistical significance obtained in the North target area is that those conducting the experiment did not specify the rain gauges for the evaluation prior to the experiment contrary to good scientific design (e.g., Court 1960). Not naming gauges, or not using all of the established gauges in post analysis allows for “cherry-picking” of those gauges to find whatever result one wanted to find[4].
c. The delayed reporting of statistical results and other actions by the experimenters constituted scientific misconduct and therefore the body of work by those researchers is inherently compromised
S01 should be applauded for his valiant attempt in his review to ferret out the misleading reports of statistical significance. However, there is a far more important test of the reliability of published results that S01 did not consider: if demonstrable scientific misconduct occurs in the reporting of results, then no publication by the wayward author(s) before or after this time can be considered reliable and they should not be quoted.
But what is “scientific misconduct”?
“Drug Maker Admits That It Concealed Tests Which Showed Flaws.”
“The Warner-Lambert Company, one of the nation’s largest drug companies, pleaded guilty yesterday to criminal charges and agreed to pay a $10 million fine for hiding from the Food and Drug Administration faulty manufacturing processes used for several drugs….”
–New York Times, p1, 29 November 1995
Most readers can recognize egregious scientific misconduct such as concealing data that impact and change the conclusions of an experiment in which only favorable and therefore, deceptive results are published, as described in the newspaper story above.
Misconduct also occurs when researchers conceal for many years the results of new experiments that contradict those of their previously published “successful” experiments on which the scientific community and public have depended upon.
Misconduct can also be understood to have occurred when a researcher denies access to his “lab” to bonafide workers in his field who have come to study and validate his unique long-published results, results that only his lab in all the world have produced with the “equipment” he has used.
We can all recognize these acts as contrary to the values of science and its pursuit of truth. An action such as the latter is particularly odious when the researcher’s much ballyhooed results are later shown to be wholly fictitious.
Very regrettably, all three acts of misconduct occurred in the cloud seeding experiments in Israel. It is a matter of record that the experimenters chose to conceal the negative statistical results that accrued in the randomly seeded South target area of Israel-2 from the time that experiment ended in the spring of 1975 until 1990. These omitted statistical results, when incorporated into the mandated crossover evaluation of that experiment, resulted in a null (-2%) seeding effect (Gabriel and Rosenfeld 1990).[5] Thus, the complete second experiment had not replicated the results of Israel-1 as Silverman (2001) also concludes but had been widely believed based on the experimenters’ partial reports limited to the North target area (e.g., Tukey et al. 1978, Kerr 1982; Silverman 1986).
The experimenters also chose to conceal from their colleagues the ongoing results of a third randomized cloud seeding experiment, Israel-3, that was taking place in central and southern Israel. This experiment began in the fall of 1975 and ended in 1994. The random seeding in this experiment suggested year after year that seeding was having no effect or possibly decreasing the rainfall in the target (e.g., Rosenfeld and Farbstein 1992; Rosenfeld 1998). The first interim results of the third randomized experiment were not mentioned until 1992, 17 years after it had begun. This reporting behavior is in contrast to the positive reports that were issued in journals part way through the first and second experiments by Gabriel 1967 and GN74 when the effects of seeding were indicated to be positive.
In summary, during 24 consecutive years of randomized seeding in the south target in Israel-2 and -3 combined, had less rainfall (about 10%) on seeded days than on the control days. This contrary knowledge was hidden from the scientific community for more than 15 years. The original authors of the partial reports passed away before these concealed results were, or could be, reported in journals for the outside community to evaluate.
No one can doubt that the crucial negative statistical results of seeding described above in Israel-2 and -3, would have raised many questions, and should have been made known om a timely manner, first of all, to the experimenters’ own countrymen, to the outside scientific community as a whole, and especially to the scientific community in Jordan downwind of central and southern Israel that might have been impacted by the remote possibility of having their rainfall decreased.
In a third example of misconduct, this writer, known to the leader of the Israeli experiments as a skeptic[6] of the cloud microstructure reports that he had been publishing in journals, was denied access to the experimenters’ two radars during rainy spells to examine the heights (and temperatures) of precipitating clouds during his 11-week visit to Israel in early 1986. One of the radars, a vertically pointing X-band or 3-cm wavelength radar, was located near the offices of the experimenters at a satellite office of the Hebrew University of Jerusalem. The second, a C-band or 5.5-cm wavelength radar, was located on the grounds of Ben Gurion Airport to which this writer bicycled to from Tel Aviv for a meeting with the leader of these experiments who forbade him to go there during storms due to “airport security.”
From the rawinsonde data analyzed by me in 1988, and from the airborne data of Levin et al. 1996 we now know why the experimenters did this.
It doesn’t seem possible to conjure up the magnitude of incompetence required on the part of the experimenters to misinterpret so many cloud measurements over so many years from the many measurement sources they had at their disposal: their own radars, satellite thermal imagery which they used routinely for forecasting cloud seeding opportunities and in their research reports (e.g., Rosenfeld 1980; G80), their own aircraft that for two rainy seasons skimmed the tops of clouds over their vertically pointing radar at Jerusalem (e.g., G80). And, of course, they also had the Israel Meteorological Service (IMS) rawinsonde profiles launched up to four times a day from Bet Dagan (near Tel Aviv )from which it could be discerned that the clouds were not as they were describing them.
In view of these documentable instances of misconduct by the experimenters, none of the publications regarding cloud seeding, or its potential in Israel or elsewhere by those who participated in these acts, can be considered reliable. Such publications should not be quoted until the full story concerning the actions of the experimenters is revealed. Of particular interest is the original list of random decisions for Israel-2 due to the extreme one-sided nature of that draw on seeded days (Gabriel and Rosenfeld 1990). We need to be sure that the list wasn’t compromised when heavy rain was forecast by the Israel Meteorological Service. From experience in commercial projects, I know that it’s satisfying to say when someone asks that you seeded when heavy rain falls.
- But why discuss misconduct in science? Won’t a discussion of, or a finding of “misconduct” diminish public support of science? And won’t that, in turn, lessen the job opportunities that we workers in science might have?
In fact, from our own narrow purview, it could be (and will be by some) argued that we should never discuss or even mention misconduct in any area of science. Rather, we should promote the thought that as scientists, we are not like other people, but, in fact, are superior to them and never do anything wrong or fraudulent because of our training like other people. Sarcasm here.
Of course, we must not only discuss but eradicate misconduct from our ranks or others will. And, yes, it is likely that there will be some erosion of public support for cloud seeding in the face of reports of misconduct in that field. Someone may indeed lose his job.
But from a larger viewpoint, it is an outrage to not consider the question of whether scientific misconduct occurred that resulted in misspent tens of millions of public dollars.
Most worrisome, there are no guarantees that this will not happen again. We do not know, for example, if more data relative to cloud seeding or cloud microstructure are being concealed or consciously misanalyzed by this same group in ongoing efforts to justify what now appears to be a dubious operational cloud seeding program begun by the Israeli government in 1975 that was based on the experimenters’ partial statistical reports of seeding success and descriptions of fictitious clouds.
Update: Due to the re-analysis of the Israeli experiments in 1995, and the subsequent journal exchanges in 1997, a panel was formed by the Israeli government to independently examine the results operational cloud seeding of the Sea of Galilee (Lake Kinneret) that began with the winter of 1975-76. The panel found no viable evidence that cloud seeding had increased runoff into the Sea of Galilee over a 27 year period (Kessler et al. 2006, Sharon et al. 2008). Israel-4 ended with a null result after seven seasons of randomized seeding (Benjamini et al. 2023).
A reputational dark cloud will hang over the group from which these acts originated until the details are fully known and the data in these papers verified. An independent panel of inquiry into these matters, while time consuming, can only benefit all parties by lifting this dark cloud so that we can move ahead.
On the other hand, the field of cloud seeding is unique in the atmospheric sciences. Its past charlatans, quacks, and even misguided, self-deceived but sincere scientists who made ludicrous claims about seeding effects, have been well documented throughout its history (e.g., Fleming 2010). In the early days of modern cloud seeding we had the USWB with their aircraft and their independent scientists (e.g., Coons and Gunn 1951) to invalidate some of the outrageous claims being made.
Today, it seems, we have only the peer-review process to ensure that the truth is told. And, from the many reversals of published findings of major, even widely accepted experiment results, some once seen as having “proved” cloud seeding as evaluated by our best scientists (e.g., National Academy of Sciences 1973), we can conclude that peer-review is but a thin “firewall” indeed against this type of more subtle quackery.
Workers in weather modification today, like S01, with their own continuing silence on the matter of the omitted, crucial statistical results of the experiments conducted in Israel, are exhibiting an eerie tolerance for a pernicious kind of science reporting in journals on cloud seeding that, from this author’s viewpoint, threatens to destroy this field altogether.
REFERENCES
Benjamini, Y, A., Givati, P., Khain, Y. Levi, D. Rosenfeld, U. Shamir, A. Siegel, A. Zipori, B. Ziv, and D. M. Steinberg, 2023: The Israel 4 Cloud Seeding Experiment: Primary Results. J. Appl. Meteor. Climate, 62, 317-327. https://doi.org/10.1175/JAMC-D-22-0077.1
Bouqard, A. D., 1963: Ice nucleus concentrations at the ground. J. Atmos. Sci., 20, 386-391.
Braham, R. R., Jr., 1964: What is the role of ice in summer rain-showers? J. Atmos. Sci., 21, 640-646.
_______________., 1979: Field experimentation in weather modification. J. Amer. Stat.
Assoc., 74, 57-68.
Coons, R. D., E. L. Jones, and R. Gunn, 1949: Artificial production of precipitation. Third Partial Report: Orographic Stratiform Clouds–California, 1949. Fourth Partial Report: Cumuliform Clouds–Gulf States, 1949. U. S. Weather Bureau Res. Paper No. 33, Government Printing Office, Washington, 46 pp.
_________, and R. Gunn, 1951: Relation of artificial cloud modification to the production of precipitation. Compendium of Meteorology. Amer. Meteor. Soc., 235-241.
Cotton, W. R., and R. A. Pielke, 1995: Human Impacts on Weather and Climate, 2nd edition, Cambridge University Press, 288 pp.
Court, A., 1960: Evaluation of cloud seeding trials. J. Irrig. and Drainage Div., Proc. Am. Soc. Civ. Eng., 86, No. IR 1, 121-126.
Dennis, A. S., 1989: Editorial to the A. Gagin memorial issue. J. Appl. Meteor., 28, 1013.
Fleming, J. R., 2010: Fixing the Sky: The Checkered History of Weather and Climate Control. Columbia University Press, 306pp.
Gabriel, K. R., 1967: The Israeli artificial rainfall stimulation experiment: statistical evaluation for the period 1961–1965. Proceedings, Fifth Berkeley Symposium on Mathematical Statistics and Probability, Vol. 5, L. M. LeCam and J. Neyman, eds., University of California Press, 91–113.
_______, and D. Rosenfeld, 1990: The second Israeli rainfall stimulation experiment: analysis of rainfall on both target areas. J. Appl. Meteor., 29, 1055–1067.
Gagin, A., 1975: The ice phase in winter continental cumulus clouds. J. Atmos. Sci., 32, 1604–1614.
________, 1980: The relationship between the depth of cumuliform clouds and their raindrop characteristics. J. Res. Atmos., 14, 409-422.
_______, 1981: The Israeli rainfall enhancement experiments. A physical overview. J. Wea. Modif., 13, 1–13.
________., 1986: Evaluation of “static” and “dynamic” seeding concepts through analyses of Israeli II and FACE-2 experiments. In Precipitation Enhancement–A Scientific Challenge, Meteor. Monog., 21, No. 43, Amer. Meteor. Soc., 63-70.
_______, and J. Neumann, 1974: Rain stimulation and cloud physics in Israel. Weather and Climate Modification, W. N. Hess, Ed., John Wiley and Sons, 454–494.
________., and J. Neumann, 1976: The second Israeli cloud seeding experiment–the effect of seeding on varying cloud populations. Proc. II WMO Sci. Conf. Weather Modification, Boulder, WMO Geneva, 195-204.
_______, and _________, 1981: The second Israeli randomized cloud seeding experiment: evaluation of results. J. Appl. Meteor., 20, 1301–1311.
Hallett, J., and S. C. Mossop, 1974: Production of secondary ice particles during the riming process. Nature, 249, 26-28.
Hobbs, P. V., 1969: Ice multiplication in clouds. J. Atmos. Sci., 26, 315-318.
__________., and A. J. Alkezweeny, 1968: The fragmentation of freezing water droplets in free fall. J. Atmos. Sci., 25, 881-888.
Kerr, R. A., 1982: Cloud seeding: one success in 35 years. Science, 217, 519–522.
Kessler, A., A. Cohen, D. Sharon, 2006: Analysis of the cloud seeding in northern Israel. A report submitted to the Israel Hydrology Institute and the Israel Water Management of the Ministry of Infrastructure, In Hebrew with an English abstract). 117pp. No doi.
Koenig, L. R., 1963: The glaciating behavior of small cumulonimbus clouds. J. Atmos. Sci., 20, 29-47.
Kraus, E. B., and P. A. Squires, 1947: Experiments on the stimulation of clouds to produce rain. Nature, 159, 489-492.
Lawson, R. P., and T. L. Jensen, 1998: Improved microphysical measurements in mixed phase clouds. Preprints, Conf. Cloud Phys., Everett, WA, Amer. Meteor. Soc. 451-454.
Levin, Z., E. Ganor, and V. Gladstein, 1996: The effects of desert particles coated with sulfate on rain formation in the eastern Mediterranean. J. Appl. Meteor., 35, 1511-1523.
List, R., K. R. Gabriel, B. A. Silverman, Z. Levin, and T. Karacoastas, 1999: The rain enhancement experiment in Puglia, Italy: statistical evaluation. J. Appl. Meteor. 38, 281-289.
Lovasich, J. L., J. Neyman, E. L. Scott, and J. A. Smith, 1969: Wind directions aloft and effects of seeding on precipitation in the Whitetop experiment. Proceedings, National Acad. Sci., 64, 810-817.
Mason, B. J., 1980: A review of three long-term cloud-seeding experiments. Meteor. Mag., 109, 335-344.
___________, 1982: Personal reflections on 35 years of cloud seeding. Contemp. Phys., 23, 311-327.
Mossop, S. C., 1985: Secondary ice particle production during rime growth: the effect of drop size distribution and rimer velocity. Quart. J. Roy. Met. Soc., 111, 1113-1124.
National Academy of Sciences, 1973: Weather Modification: Progress and Problems. T. F. Malone, ed., 258 pp. (Available from the National Research Council, Washington, D. C.)
Rangno, A. L., 1988: Rain from clouds with tops warmer than –10°C. Quart J. Roy. Meteor. Soc., 114, 495-513.
___________, 2000: Comments on “A review of cloud seeding experiments to enhance precipitation and some new prospects.” Bull. Amer. Meteor. Soc., 81, 583-585.
____________, and P. V. Hobbs, 1988: Criteria for the development of significant concentrations of ice particles in cumulus clouds. Atmos. Res., 22, 1-13.
_____________, and __________, 1991: Ice particle concentrations and precipitation development in small polar maritime cumuliform clouds. Quart. J. Roy. Met. Soc., 117, 207-241.
___________, and __________, 1994: Ice particle concentrations and precipitation development in small continental cumuliform clouds. Quart. J. Roy. Meteor. Soc., 120, 573-601.
___________, and P. V. Hobbs, 1995: A new look at the Israeli cloud seeding experiments. J. Appl. Meteor., 34, 1169-1193.
___________, and _________, 1997a: Reply to Rosenfeld. J. Appl. Meteor., 36, 272-276.
___________, and _________, 1997b: Comprehensive Reply to Rosenfeld. Cloud and Aerosol Research Group, Department of Atmospheric Sciences, University of Washington, 25 pp.
Rosenfeld, D., 1980: Characteristics of rain cloud systems in Israel derived from radar and satellite images. M. S. Thesis, The Hebrew University of Jerusalem, Israel, 129 pp. (Available from the Department of Geosciences, Hebrew University of Jerusalem, Israel)
___________, 1997: Comment on “Reanalysis of the Israeli Cloud Seeding Experiments.” J. Appl. Meteor., 36, 260-271.
___________, 1998: The third Israeli randomized cloud seeding experiment in the south: evaluation of the results and review of all three experiments. Preprints, 14th Conf. on Planned and Inadvertent Wea. Modif., Everett, Amer. Meteor. Soc. 565-568.
___________., and H. Farbstein, 1992: Possible influence of desert dust on seedability of clouds in Israel. J. Appl. Meteor., 31, 722-731.
Sharon, D., A. Kessler, A. Cohen, and E. Doveh, 2008: The history and recent revision of Israel’s cloud seeding program. Isr. J. Earth Sci., 57, 65-69. https://DOI.org/10.1560/IJES.57.1.65.
Silverman, B. A., 1986: Static mode seeding of summer cumuli–a review. In Precipitation Enhancement–A Scientific Challenge, Meteor. Monog., 21, No. 43, 7-20.
_____________, 2001: A critical assessment of glaciogenic seeding of convective clouds for rainfall enhancement. Bull. Amer. Meteor. Soc., 82, 903-923.
Stith, J. L., J. E. Dye, A. Bansemer, A. J. Heymsfield, C. A. Grainger, W. A. Petersen, and R. Cifelli, 2002: Microphysical observations of tropical clouds. J. Appl. Meteor., 41, 97-117.
Turkey, J. W., D. R. Brillinger, and L. V. Jones, 1978b: Report of the Statistical Task Force to the Weather Modification Advisory Board, Vol. II. U. S. Government Printing Office, pE-3.
Young, K. C., 1993: Microphysical Processes in Clouds. Oxford University Press, 335-336.
Wurtele, Z. S., 1971: Analysis of the Israeli cloud seeding experiment by means of concomitant meteorological variables. J. Appl. Meteor., 10, 1185-1192.
Footnotes
[1] The University of California Statistical Laboratory under the direction of Jerzey Neyman.
[2] Four ground generators were located in hilly terrain in the extreme northeast portion of the country.
[3] The Chief Meteorologist for the seeding experiments in Israel was misidentified by Wurtele (1971) as a meteorologist with the Israel Meteorological Service.
[4] Rangno and Hobbs (1995) used a subset of rain gauges that whose data were routinely published in monthly or annual summaries by the Israel Meteorological Service and therefore, were “pre-selected” by the Israel Meteorological Service, an agency not affiliated with the seeding experiments.
5 Even then, these “full” statistical results for the second crossover seeding experiment were not published in a vacuum; but only after the leader of the experiments passed away in 1987, and after a letter-writing campaign to Israeli government offices by the former Chief Meteorologist of the experiments urging the experimenters to publish the full results.
[6] My July 1983 submitted article, asserting that rain was falling from clouds with tops much warmer than could be accounted for by the experimenters’ cloud descriptions was rejected by J. Climate Appl. Meteor. in 1983 (B. A. Silverman, co-chief Editor, private communication.) Moreover, the leader of the experiments in Israel had provided a lengthy and perhaps pivotal negative review of that paper (A. Gagin, 1984, private communication.)
The Catalina, AZ, 2023-24 Water Year
Check it out. We’re still having generally greater totals than we had during those many droughty years of the late 1990s to about 2012 that caused so much speculation about longterm permanent drought here due to global warming (later rephrased to “climate change” after a hiatus in global warming that began around 1999 and lasted more than a decade). The red line is a quadratic fit to the data. Let’s hope those greater WY totals of late continue despite overall gradual warming over this whole period.
The Catalina WY record happened to begin in 1977-78 right at the beginning of one of the wettest few years in hundreds of years in the Southwest generally as seen in tree ring records. The overall Arizona statewide average (unfortunately presented by calendar years by NOAA) doesn’t show much going on over the past 100 years, see lower graphic through 2022, the latest year available.
Catalina Cool Season (October-May) Was Slightly Above Average
Here’s a long term graph with a curve fit of the Catalina October through May precipitation beginning in 1977.
The Catalina record begins with an extraordinary wet spell in 1977-78 through the early 1980s, augmented by El Niñoes and possibly the Pinatubo eruption in the early 1990s. The initial wet spell was associated with a shift in the so-called “Pacific Decadal Oscillation.” The red line is a polynomial fit to the data. It seems to suggest a recovery is underway from the overall drier winters of the early 2000s despite the 2020-21 Oct-May drought. Will a recovery continue? Is it real?
A New Decadal Climate Oscillation Detected in Past Data a Long Time Ago? A Fool’s Journey? Namias Reacts
Purpose of this post of ancient, unfinished work with the humorous title: Could it inspire someone to continue it in a more sophisticated way than I have?
Named climate/weather influencing “oscillations” have become so numerous in the scientific literature (e.g., Atlantic Meridional Oscillation, Arctic Oscillation, Pacific Decadal Oscillation, El Niño-Southern Oscillation, Atlantic Oscillation, Quasi-Biennial Oscillation, Madden-Julian Oscillation, etc.) one is tempted to ask humorously, “Doesn’t everyone have one?”
Well, I do, but this research is incomplete. And it may ne bogus, illusory; it seems to lead to a dead end. Still, “we” journey on, hoping these early findings, incomplete as they are and needing to be updated, will nevertheless bring someone a Nobel Prize in meteorology. (Also still waiting for some kind of science prize or medal from Israel for the work I’ve done there in exposing faulty cloud seeding results and descriptions of non-existent clouds supposedly ripe with seeding potential (Rangno and Hobbs 1995, Rangno 1988). But, as the Prunes sang, I had too much to dream last night.
https://youtu.be/-xRRT92Fpgs
Let us begin this story from the beginning. Let us explore how a gigantic amount of work, consuming work, really, thousands of hours of personal effort, can lead to a dead end. Maybe.
The oscillation story begins in an undergraduate climatology class I took at San Jose State College in the late 1960s. Each of us was assigned to do a climate research project. I chose to do something on Los Angeles Civic Center rainfall, something I had been charting since childhood. I graphed the days with measurable rain over the period of record for the Civic Center going back to the 1877-78 “rain season.” In California, the period of July 1 through June 30 is deemed the “rain season.” That was the way rainfall data were presented in the newspapers. The California “rain season” is similar to the water year precipitation totals widely used in the western US for the period of October 1 through September 30. If you are a meteorologist in the West, the calendar year is generally eschewed in place of rain season or water year since the latter capture the character of whole winters and better account for snowpacks in mountains.
I saw an interesting phenomenon in the plot of days with measurable rain; there appeared to be an “instability”; a jump to much wetter conditions after a decades long trend of declining days with measurable rain. After reaching what appeared to a minimum of days with measurable rain, there was a sudden jump in the next season to one having considerably more than the average. But, it wasn’t just one season that had many more days with rain! It was most of the next ten rain seasons that had above average days with rain. This had happened in my plot on three occasions; the season following 1903-04, the season following 1933-34, and lastly, the season following 1976-77. Namely, it wasn’t just a one-shot wonder, a season long singularity.
I was EXCITED! So excited I eventually sent my Los Angeles Civic Center plot to a famous professor of climate and weather at Scripps Institution of Oceanography, Prof. Jerome Namias. I had read his papers in the Monthly Weather Review. (I had begun subscribing to this journal when I was 13 years old. By this time I had seen that these same three jumps had occurred at San Luis Obispo, Santa Barbara, and San Diego. There was no evidence of this phenomenon at San Francisco; it faded to the north.
If you’re a Los Angeles or Southern California “precipophile” like me, you might well guess immediately why I wasn’t so much interested in rain totals as with days with measurable rain. A couple of huge storms can hide the character of a whole rain season, but the character of a rain season would be called out by days with rain. An extreme example: on New Year’s Eve, December 1933 into New Year’s Day, 1934, Los Angeles received it’s greatest 24 h rainfall: 7.36 inches! Approximately half fell in December 1933 and a little more than half in January 1934. The season’s rainfall total that year was 14.64 inches, or just about average. HOWEVER, the DAYS with rain was two standard deviations below average! That’s “what done it” for me, that extraordinary rain day and why rain totals might hide the real character of a winter’s rain season.
Naturally, there is pretty high correlation between the days with measurable rain and the season’s total. But from day one, I was convinced that the days with measurable rain was a better indicator of circulation changes over the years while rain amounts added “noise.” Here is the correlation between rain amounts and days with rain for Los Angeles Civic Center:
Professor Jerome Namias seemed to be excited, too. Here is his reply to my graph of LA rain frequency over the period of record.
Here’s the plot that started it all and that Namias and Stidd found interesting. It didn’t look like “white noise.”
By the time I corresponded with Professor Namias, I had been hired as the Assistant Project Forecaster for the nation’s largest randomized orographic cloud seeding experiment, the Colorado River Basin Pilot Project. I was living in Durango, CO, where the project’s headquarters were as I continued my research on a possible oscillation.
I was hoping that my work would eventually qualify as a Master’s Thesis from the meteorology department at San Jose State College (despite poor grades in grad school). As an aside to the reader: I had no business whatsoever in being in grad school taking classes like “numerical methods,” “advanced hydrodynamics.” etc. But I loved my campus life in those days of campus trash cans set on fire to protest the war, album protest rock, the marches, the demonstrations, the be-ins, the draft card burnings, to pinch a quote from National Lampoon’s “Bob Dylan’s Golden Protest” parody:
Over the next couple of years I added to my Los Angeles dataset with ones from San Francisco, San Luis Obispo, Santa Barbara, and San Diego. These stations all had records that went back even earlier than the one at Los Angeles. All but San Francisco, whose record exhibited what one might call, “white noise,” exhibited the trend I had seen in the Los Angeles record! It was amazing to see.
Below is a running, 3-season average for SLO, SBA, LAC, and SAN through the early 1980s. Sadly, I was not paying attention when scanning this original diagram and parts are not shown here, such as the abscissa comprised of the rain seasons. The ordinate is the number of days with measurable rain at these four stations with a running mean of 3 seasons of those totals. It ends with the 1981-82 season, far right. For those in the know, the following season, 1982-83 featured a giant El Niño and that, combined with whatever was going on in what I was charting, produced numbers of days off the chart! I was so happy! The Great Salt Lake was about to overflow, too, from this incredible wet spell that accompanied the shift to a more frequently rainy regime.
Since the rain at these locations is associated with cold troughs in the wintertime westerlies, I imagined that the circumpolar westerlies gradually retracted over the years, then hit some kind of tipping point and sprung back to more a more southerly latitude before beginning the same slow retraction over decades.
In the 1950s and 1960s, the Los Angeles forecast office used the 564 decameter geopotential height contour as a divider of rain; heights at or below that contour was where the rain was and no rain was the rule for heights greater than the 564 decameter contour. This key contour was used as an aid in forecasting rain as troughs approached and entered California.
The last shift to wetter conditions I found happened after the 1976-77 rain season. With the 1977-78 rain season, it became much more frequently rainy along the central and Southern California coast for most of the next seven years.
My California rainfall study ended a few years into this transition to wetter conditions due to two elements: 1) NOAA had stopped publishing the “Daily Series, Synoptic Weather Maps, Part 1, Northern Hemisphere Sea-Level and 500 mb Charts and 2) I got very upset over the misleading cloud seeding literature that was being published in journals and jumped ship into reanalyzing previously published cloud seeding literature for most of the rest of my career almost solely on my own time (e.g., Rangno 1979, Hobbs and Rangno 1979, Rangno and Hobbs 1995).
The NOAA surface and 500 mb charts were important because that’s what I had used to track cyclones across the Pacific for five winter seasons before and after a “shift.” I wasn’t able to do a set of tracks before and after the 1977-78 shift. It was interesting that about 20 years later, the 1977-78 shift I was studying was discovered as the, “Pacific Decadal Oscillation” (Wallace et al. 1994).
By the 1973-74 rain season I was so sure a prolonged shift to wetter conditions in the SW was on the doorstep that I was writing to the LA Times science writer, George Getz, the BuRec’s PR person, Hunter Holloway (the BuRec was the sponsor of the cloud seeding experiment I worked on), and to the Durango Herald about this coming shift to wetter conditions.
I was a little too early; the downward trend continued through the 1976-77 winter. Here’s an example of those writings, one an audacious, self-written “news release” that follows the letter to Mr. Holloway shown here:
The reason I posted the letter above is to PROVE that I really was anticipating “The Shift” BEFORE it happened:
The additional research I carried out went far beyond the rain day graphs:
Pacific cyclone tracks before and after a “shift,”
NH average sea level pressures before and after a shift.
Here are the sea level maps for the 1930s, before and after a shift that occurred with the 1934-35 rain season. For those not acquainted with the synoptic charts of that 1930s era it was something of a golden age of ship reports before they disappeared on these maps during World War II. Thus northern hemisphere sea-level pressure and cyclone tracks changes could be reliably charted.
A graphically obtained “delta” map of sea-level pressure changes before and after a shift comprises the third graphic. Not much can be seen to have happened when looking at these average maps for whole December through March seasons due to having semi-permanent pressure systems like the low in the Gulf of Alaska, and the “Pacific High.” I deemed looking at cyclone tracks (the following graphics) as far more useful. Nevertheless, these are never-before-seen-or-done maps by anyone but me. Enjoy:
Strangely believe it, the greatest sea level changes were in the domain of the Arctic Oscillation; in the North Atlantic, Greenland and England. Whodda thunk it? Not much change in the Pac and West, as expected due to persistent pressure fields even in these extremely frequent days with rain and those with many fewer ones. I honestly did not know what to make of this change. Do you? If not, let’s keep moving….
Before doing the sea level pressure maps, I had charted cyclone centers across the central Pacific to the Rockies for the same winter (Dec-Mar) periods as was done above. These charts were much more illuminating concerning the shift that happened in days with rain in central and Southern California:
First, the cyclone track “densities” of the five low frequency days with measurable rain winters preceding the shift. A strong channeling of cyclones was seen during these winters from the lower latitudes of the central Pacific into the Gulf of Alaska and into southern British Columbia to a lesser degree. This suggests the geographically anchored Asia/western Pacific Ocean jet exit had extruded farther east and into the central Pacific. (Upper-level maps are not available for this period.). This was to resemble what happened in the low rain frequency seasons of the early and mid-1970s that preceded another shift.
In the figure below are the tracks following the “shift” to higher frequency rain day occurrences in central and Southern California. The channeling is gone and what appears to be a standard distribution of low centers has replaced it. Cyclones that developed in the western Pacific moved northeast into the Aleutians relatively close to Japan rather than scooting across the lower latitudes of the central Pacific. The difference between these two maps is shown in the final map below this one.
The Delta Cyclone Density Map (pardon the skew):
Many more low centers tracked across the extreme eastern Pacific into California and into the Great Basin low pressure cyclogenesis zone in the lee of the Sierras. This map is more illustrative than the sea-level pressure maps of those changes that happened after the shift to wetter conditions .
Questions: Did channeled cyclones disrupt the sea surface enough to cause temperature anomalies in real time or later? Were the El Niños that followed in the late 1930s and early 1940s in part triggered by the lower latitude cyclones racing across the central Pacific? Was this a global change such that Western Europe and northwest Africa saw a shift to wetter conditions after the ones noted in central and Southern California? Data were not available for this period of study (1930s) except in the Middle East for Jerusalem, Israel. There was no indication of a long period of decreasing days with measurable followed by a sudden and prolonged shift to more frequently rainy days.
My favorite answer to, “What caused this shift when someone asked was based on the work of E. N. Lorenz at MIT. Professor Lorenz is famous for bringing our attention to chaos theory where small initial changes in starting conditions can lead to huge differences later. Today, this concept is used to improve forecasts by introducing minor changes in initial data in the models from that observed to see how much difference results in the model predictions.
There need not be an external cause for climate shifts, he wrote. It may just that systems shift in and out of favorite modal types without an external forcing. Thus, the best answer for what caused the shift I think I detected was, of course, “nothing.” See Lorenz (1968, Climate Determinism) in the Amer. Meteor. Soc. Monograph Vol. 8, No. 30, on, “The Causes of Climate Change.” Strangely believe it, there are no papers on CO2 in this Amer. Meteor. Soc. monograph! Mauna Loa measurements of C02 had just begun and so there was no awareness of how it was steadily increasing over the years.
——-
Work to be done?
Bringing the days with measurable rain plots up to date, storm tracks in the 1970s, and investigate whether the 564 decameter geopotential height contour shows a global expansion after the 1970s shift after hemispheric geopotential heights became available (beginning in the mid-1940s). Or is the shift to a high rain frequency along the central and Southern California coast a local phenomenon where a mean upper-level trough recurrs along the West Coast for years at a time? In a cursory check recently, there seems to be no clear recurrence of the prior three shifts in days with rain. Boo.
And lastly, one must ask, is what I have done a “scientific mirage,” and expression used by Foster and Huber (1997) to denote illusory science?
I would like the answer to be a sudden global expansion of the polar westerlies (shown in the average hemispheric latitude of the 564 500 hPa geopotential height contour that accompanies a shift, something like the “Bond Cycles” observed in historic ice core data: a “reset,” if you will, after a long, long period of withdrawing ever so gradually over the decades. Perhaps our gradually shifting climate to a warmer one has done something to interfere with this “shift” phenomenon?
An obstacle that arose in later years was that climate station at Cal Poly, San Luis Obispo. whose record started in 1868, started having missing monthly data that never showed up in the “Delayed Data” section in NOAA’s Climatological Data in the June and December issues. Eventually, NOAA dropped even having a “Delayed Data” segment in the June and December issues! Boo on that! Santa Barbara, whose record also went back into the 1860s, too, had missing months that never were reported. I remember how discouraged I was when these events happened and thought about giving up. How can stations whose records go back into the 1860s (!) suddenly have incomplete records where observers don’t file reports with NOAA? Where was the California State Climatologist? Asleep at the wheel, I suppose.
Maybe you, one of my two readers, will take this research up to see if it goes anywhere?
Sincerely, Art “I’m dreaming” Rangno. :
Caveat: My grad advisers at San Jose State and a professor at the U of WA were unimpressed with these works.
A Review of the Israeli Cloud Seeding Experience in the Context of the 2023 Israel 4 Null Primary Result
PROLOGUE
I have written an extensive, “comment” and “enhancement” of an article by Benjamini et al. published in the in J. Appl. Meteor. in January 2023. The article was about the results of a fourth randomized Israeli cloud seeding experiment, Israel-4. My “comments” and “enhancement” of Benjamini et al. (2023) posted below would never be published in an Amer. Meteor. Soc. journal. The words are too strong. So, I am going this route, a blog post.
Evidence for such a contentious assertion? Prior experience.
I submitted a paper on the history of Israeli cloud seeding in 2018. The journal, the Bull. Amer. Meteor. Soc. (BAMS) got but two reviews: “Accept, important paper, minor revisions” by one Israeli scientist, and the second review, an outright “reject” in a long review by an Israeli seeding partisan who signed his review. The chief editor of BAMS did not allow me to revise my manuscript where needed (minor corrections), nor rebut the many specious comments by the seeding partisan.
Why is this behavior by the chief editor of BAMS outrageous and in non-compliance with our science ideals?
Replying to the comments of reviewers of manuscripts following peer-review is standard procedure in science after which a final decision on publication is then reached by the editor based on the responses of the author and the revisions made in the submitted manuscript. This is exactly the process that Prof. Dave Schultz, Chief Editor of the Amer. Meteor. Soc. journal, The Monthly Weather Review, and I are going through right now with a cloud seeding manuscript on the Colorado River Basin Pilot Project sub omitted to the AMS’ J. Appl. Meteor. and Climate (as of January 2024).
As an acknowledged expert on Israeli clouds, weather, and cloud seeding (e.g., Rangno 1988, Rangno and Hobbs 1995a), I deemed this refusal by the BAMS Chief Editor to allow me to respond to the comments of the two reviewers the sign of a corrupted journal process within the Amer. Meteor. Soc.: Certain stories about failed science are not to be told, especially if they involve a country people have strong feelings about, as in this case. My 2018 history describes unimaginably inadequate peer-reviews of the original published reports, those describing ripe for seeding clouds and the cloud seeding statistical “successes” that were all scientific mirages crafted by cloud seeding partisans.
The manuscript below has the same elements as the 2018 submission thus guaranteeing its rejection by a partisan AMS leadership. But I feel strongly that certain things need to be said, and questions asked, to stop seeding partisans in Israel from costing their country so much in wasted cloud seeding efforts as they have over so many decades. Sound implausible? Read on…..
Despite what might be considered some “harsh” language at times, I consider myself a friend of Israel and donate to the American-Israeli Cooperative Enterprise, an organization that regularly counters the negative descriptions of Israel in much of the media today in their “Myth vs. Facts” segments.
=========THE MANUSCRIPT==========
ABSTRACT
The result of a fourth long-term randomized cloud seeding experiment in Israel, Israel-4, has been reported by Benjamini et al. 2023. The seven-season randomized cloud seeding experiment ended in 2020 with a non-statistically significant result on rainfall (a suggested increase in rain of 1.8%). This review puts the results of Israel-4 in the context of prior independent reanalyses of Israel-1 and -2, reanalyses that can be said to have anticipated a null result of both the Israel-4 experiment and the lack of evidence that rain had been increased in the 30 plus years of the operational cloud seeding program targeting the Lake Kinneret (Sea of Galilee) watershed discovered in 2006 by an independent panel of Israeli experts. The published literature that overturned the reports of success in the first two experiments, Israel-1 and Israel-2, was omitted by Benjamini et al., and thus, misled readers concerning those first two experiments.
The lack of cloud seeding success in Israel can be attributed to unsuitable clouds for seeding purposes, ones that form prolific concentrations of natural ice at relatively slight to moderate supercoolings which preclude seeding successes using glaciogenic seeding agents.
The phenomenon of “one-sided citing,” practiced by Benjamini et al. via the omission of relevant contrary literature is addressed. Several corrections to and enhancements of the Benjamini et al. article are also included.
===================================================
- Introduction and Background
The results of the first two randomized crossover cloud seeding experiments in Israel, Israel-1 and Israel-2, discussed recently by Benjamini et al. 2023, as well as the descriptions of “ripe for seeding” clouds in Israel by the seeding experimenters, had an important role in the history of cloud seeding. For many years it appeared that the viability of cloud seeding to have produced economically important amounts of rain had been established in those two “crossover” experiments conducted by scientists at the Hebrew University of Jerusalem (HUJ) (e.g., Kerr 1982, Mason 1982, Dennis 1989). In descriptions of the first two benchmark experiments, ones that created the scientific consensus described above, Benjamini et al. (2023, hereafter, “B23,”) do not tell the whole story in their history of cloud seeding in Israel that preceded their evaluation of Israel-4.
This review is meant to fill in the gaps for the reader left by B23 about those first two experiments that had so much practical impact. For example, the Israel National Water Authority (INWA) began a several decades-long operational cloud seeding of the watersheds around Lake Kinneret (aka, Sea of Galilee) based on the seemingly favorable results of Israel-1 and those in the “confirmatory” Israel-2 experiment that followed (Gabriel 1967a; b; Neumann et al. 1967; Wurtele 1971; Gagin and Neumann 1974; 1976). The INWA began seeding Lake Kinneret’s watersheds in November through April, beginning with the 1975/76, the winter season that immediately followed the end of Israel-2.
The statistical results of Israel-1 and -2 were backed by several cloud microstructure reports over the years that underpinned the idea that rain could be increased substantially by seeding Israel’s clouds (e.g., Gagin 1975, 1981, 1986, Gagin and Neumann 1974, 1976, 1981). These reports caused Science magazine’s reporter, Richard Kerr, to proclaim in 1982 that those first two Israeli experiments constituted the “One success in 35 years” of cloud seeding experimentation. Kerr (1982) also wrote:
“The Israeli II1 data must still be reanalyzed by other statisticians, but most researchers are also impressed that the results make so much physical sense. The clouds that Gagin and Neumann hypothesized would be most susceptible to seeding did indeed produce the most additional rain after seeding.”
These statements are compatible with the history that B23 have provided, but it was to be far from the end of the “story.”
Fifteen years after Israel-2 had been completed it was learned that the random seeding of the south target clouds of Israel-2, a crossover experiment as Israel-1 had been, produced the indication that cloud seeding had decreased rainfall by a substantial amount, 15% (Gabriel and Rosenfeld 1990)2. Gagin and Neumann (1981), however, had claimed that the random seeding that took place in the south target was “non-experimental” and so did not report the results of random seeding there. No one challenged this claim.
Until 1981 the result of seeding in the south target seeding had been described as “inconclusive” (Gagin and Neumann 1976), and prior to that, by (Gagin and Neumann 1974) after the first two seasons of Israel-2, that seeding had resulted in a seed/no seed average rainfall fraction in the south target that was “less than 1,” suggesting rain might have been decreased on seeded days.
However, the crossover evaluation of seeding in Israel-2 was not reported until Gabriel and Rosenfeld (1990)2. The design document, approved by the Israeli Rain Committee and completed before Israel-2 began had, however, mandated a crossover evaluation (Silverman 2001) as had been done for Israel-1. Nowhere did Gagin (1981) or Gagin and Neumann (1974, 1976, 1981) explain why they did not perform the mandated crossover evaluation of Israel-2.
Instead of Israel-2 crossover evaluation replicating Israel-1, where seeding appeared to have increased rainfall by about 15% when the data from both targets was combined (e.g., Wurtele 1971), the crossover evaluation of Israel-2 indicated a slight decrease in rainfall of 2% (not statistically significant). Thus, Israel-2 had not replicated Israel-1 in an important way.
But results of Israel-2 were complex, as noted by Gabriel and Rosenfeld (1990) and left questions that they could not resolve. The most revealing statement in Gabriel and Rosenfeld (1990) in reporting the “full” results of Israel-2 was this enigma (my italics and bold font):
“There is a surprising contradiction between this finding and those of the analyses of Tables 4-17. The difference occurs because the historical comparison of Table 18 ignores the unusually high south precipitation on north-seeded days (as well as the north precipitation on south-seeded days). In other words, it is what happened when there was no seeding that causes the differences between the two analyses. The different choice of “control” days for the south, whether all the rainy days of 1949-60 or the north-seeded days of 1969-75, is what crucially affects the comparison. If such large differences-of a magnitude of several standard errors and clearly significant by the usual statistical criteria-occur by chance, then chance operates in unexpected ways on precipitation and historical comparisons become highly suspect (see also Gabriel and Petrondas 1983). Otherwise, one would need to explain why there was so much more rain in the south when the north was being seeded; as of now, no explanation is available, especially as the prevailing wind direction is from the southwest.”
A “Type I statistical error,” the “good draw,” in Israel-2, heavy rains that affected both targets on north target seeded days3, was there for all to see if they wanted to.
Thus, a severe blow to the idea of randomizing cloud seeding experiments occurred in Israel-2 due to the exceptional random draw described by Gagin and Rosenfeld (1990). Randomization could produce wildly unrepresentative results in which slight, but important, rain increases due to seeding could be forever hidden.
The null result of the combined targets in Israel-2 was due to an apparent decrease in rainfall on seeded days in the south target (~15%) that canceled out apparent increases in rainfall (~13%) in the north target. Despite the new result and the many questions it raised, the INWA continued the commercial-like seeding of the Lake Kinneret watersheds during the winter rain seasons for more than 20 years after Gabriel and Rosenfeld’s (1990) disclosure of the “full” results of Israel-2.
The continuation of seeding of Lake Kinneret watersheds in northern Israel by the INWA despite the Israel-2 null result may have been due to the hypothesis put forward by Rosenfeld and Farbstein (1992)4; “dust/haze” had interfered with seeding in Israel -2 by creating high natural ice particle concentrations in supercooled clouds and that the presence of “dust/haze” even resulted in collisions with coalescence-formed rain (“the warm rain” process) that does not require the ice phase. These cloud attributes, they concluded, meant there could be no increases in rainfall due to cloud seeding in the south target nor in the north target when dust/haze was present. Without “dust/haze,” Rosenfeld and Farbstein argued, the clouds of Israel were as ripe as ever for cloud seeding.
2). The Motivation for a Reanalysis of Israel-1 and Israel-2
The publication and the hypothesis of Rosenfeld and Farbstein (1992) formed the motivation for the Rangno and Hobbs 1995, hereafter RH95a) reanalyses of Israel-1 and -2. This writer had spent 11 winter weeks in Israel in 1986 studying the rain-producing characteristics of Israeli clouds and felt Rosenfeld and Farbstein’s hypothesis had little credibility; a full independent review of Israel-1 and -2 was needed as had been suggested in Science magazine (Kerr 1982). And it would be done by someone who knew the clouds and weather of Israel (Rangno 1983, rejected by the J. Appl. Meteor.; Rangno (1988), Rangno and Hobbs (1988, hereafter, RH88).
I am also experienced in exposing suspect cloud seeding claims in the published literature (e.g., Hobbs and Rangno 1978, 1979, Rangno 1979, 1986, Rangno and Hobbs 1980a, b, 1981, 1987, 1993, 1995b). By the time I began reanalyzing the Israeli experiments in 1992 I had also logged more than 400 flights for the University of Washington’s Cloud and Aerosol Group in studies that mostly concerned ice crystaldevelopment in slightly supercooled clouds in polar air masses similar to those that affect Israel (Rangno and Hobbs 1983, 1991, 1994, Hobbs and Rangno 1985, 1990).
3). The results of the Rangno and Hobbs (1995) benchmark reanalyses of Israel-1 and Israel-2 that went uncited by B23
RH95 concluded that neither Israel-1 nor Israel-2 had produced bona fide increases in rain on seeded days, contradicting the HUJ experimenters’ reports and those contained in B23 that cloud seeding had increased rain in each of these experiments. The conclusions of RH95 were given support by Silverman (2001) and later, for Israel-2, by Levin et al. (2010).
Moreover, in R88 it was strongly indicated that the “ripe for seeding” clouds described repeatedly by the experimenters (e.g., Gagin and Neumann 1974, 1976, 1981, Gagin 1975, 1981, 1986) did not exist. The findings in R88 concerning shallow clouds that rained was not news to Israel Meteorological Service forecasters with whom I spoke nor to the Israeli experiments’ “Chief Meteorologist,” Mr. Karl Rosner. Mr. Rosner wrote to me in 1987 that, “sometimes heavy rain fell from clouds with tops at -8°C.” Thus, in contrast to the many HUJ experimenters’ reports cited previously, it was widely known by weather forecasters in Israel that rain fell regularly from clouds with tops >-10°C (~3-4 km thick clouds) as was documented in R88.
The HUJ experimenters had also reported, contrary to the above, that many clouds with radar measured tops between -15°C and -21°C did not precipitate naturally due to a lack of ice in them or that precipitation formed by “warm rain” (collisions with coalescence) process (e.g., Gagin 1981, 1986) did not occur. Those non-precipitating clouds in this low radar top temperature range were responsible for extra-large increases (46%) in rain on seeded days (Gagin and Neumann 1981, Gagin and Gabriel 1987).
Seeding, they also reported, had no effect on naturally precipitating clouds, a finding compatible with the “static” seeding method carried out by the HUJ experimenters where small amounts of the seeding agent, silver iodide are released. Namely, when seeding took place, it rained for more hours on seeded days than on control days, but it did not rain harder.
B23 also refer to the Israel-2 low radar top temperature partition as having been associated with increases in rain.
(Questions)
Is it possible that Israeli weather forecasters and the “chief meteorologist” of the Israeli cloud seeding experiments had a better idea of which clouds rained in Israel than those whose research careers at the HUJ depended on reliable assessments of their own clouds and their cloud seeding potential? Ans. Probably not.
Why?
This writer, while welcomed at the Israel Meteorological Service in January 1986, was denied access to the seeding experimenters’ radar on the grounds of Ben Gurion AP to obtain echo heights by the leader of the Israeli experiments, Prof. A. Gagin. He insisted in our meeting that my monitoring of top heights would only confirm his cloud reports; that it took deep and very cold-topped clouds to rain in Israel.
It was also learned during January 1986 at about this same time that no less than six attempts had been proposed by outside groups to do airborne studies of the seemingly unique clouds of Israel, as shown in RH88, ones that had responded so well to cloud seeding (Personal communication, Prof. Gabor Vali, University of Wyoming, 31 January 1986). Every one of those attempts to study Israeli clouds had been blocked.
Why? And by whom?
- More about Rangno and Hobbs (1995): the most controversial and commented on paper ever published in an Amer. Meteor. Soc. journal and the unusual strategy used by the editor in choosing reviewers
In a moment of brilliance (in retrospect), the editor for our journal manuscript, L. Randall Koenig, chose three reviewers who would be sure to reject the RH95a manuscript and its negative findings concerning cloud seeding. But at the same time, Koenig realized that there would be no easy pass on it; no stone would go unturned by the reviewers, and our findings would be severely tested. In fact, RH95 was significantly better for having cloud seeding partisans, H. Orville, W. Woodley, and D. Rosenfeld, review it (all signed their reviews).
Editor Koenig, himself an expert on weather modification and cloud microphysics (e.g., Koenig 1963, 1977, 1984), was also steeped in the long record of frequent mischief by those in the cloud seeding domain, weighed the arguments of the reviewers and the modifications of RH95a that reflected the reviewers’ criticisms: He made the choice to publish RH95a.
It took courage for Editor Koenig to do that and recognizing who he felt had the better arguments. In RH95a were the first two independent re-analyses of Israel-1 and Israel-2, as had been recommended years earlier in Kerr (1982) but ones that were clearly not going to take place. How many other papers in our journals would be the improved and bogus claims eliminated if editors used the strategy of of Koenig and were as informed about the topic of the manuscript?
Perhaps due to the size of the ox being gored, our paper drew comments by the reviewers of our manuscript and others (1997a, b, c, d, e). The number of journal pages involved in “Comments” and “Replies” on a single article is still a record for an Amer. Meteor. Soc. journal. We draw particular attention to our “Replies” to the many, as we showed, specious “Comments” of Dr. Rosenfeld in RH97a and RH97b, and a B23 co-author.
Let the reader decide where truth lies. We urge the reader to carefully review RH95a and our replies for the considerable evidence we present that the Israel-1 and Israel-2 experiments were both mirages of cloud seeding successes, contrary to the assertions in B23.
- Israel-3: enhancing B23’s description
B23 describe the results of the longest randomized cloud seeding experiment ever conducted, Israel-3 (1975-1995), a single target experiment. However, they omit informing the reader that the remarkable “inconclusive” result was a suggested 9% decrease in rainfall on seeded days compared to non-seeded days (Rosenfeld 1998). By omitting the sign of the null result, B23 left the reader to speculate on what the sign of the “null” result was. The suggestion of a decrease in rain on seeded days again points to clouds naturally form precipitation very efficiently in Israel. With the result of Israel -3 in hand, the reader would now learn, with Israel-2 (Gabriel and Rosenfeld 1990), that over a period of 25 plus years (Israel-2 and Israel-3 combined) decreases in rainfall due to seeding were suggested in central and southern Israel by cloud seeding!
- Rectifying B23’s statement concerning operational seeding
B23 state the increase in rainfall during the operational seeding, 1975/76 winter to 1990 reported by Nirel and Rosenfeld (1995) was “6-11%.” In the abstract of the quoted article, the authors state that rainfall due to cloud seeding was increased by 6%, not “6-11%.” This same increase in rain (6%) was also quoted by Sharon et al. (2008).
Moreover, the 6% increase in rain (said to be statistically significant by Nirel and Rosenfeld 1995) was not confirmed by Kessler et al. (2006) in their independent evaluation of operational seeding through the same period. The independent panel reported 4.8% suggested rain enhancement over the same period evaluated by Nirel and Rosenfeld (Figure 1).
Figure 1. The results of operational seeding on the watersheds of Lake Kinneret (aka, Sea of Galilee) as reported by Kessler et al. 2006. (a) is that result of seeding on rainfall reported by Nirel and Rosenfeld (1995), b-d are the results found for various periods, including the very same era evaluated by Nirel and Rosenfeld (1995)5.
- What triggered the formation of an independent panel to evaluate cloud seeding?
The panel was created after RH95a was published and then followed by extensive journal exchanges by RH97a, b, c, d, e, in “Replies” to various “Comments” in 1997. The INWA was then inspired to form an independent panel of experts due to these exchanges to evaluate what they were getting from the operational seeding of Lake Kinneret’s (aka, Sea of Galilee) watersheds rather than relying on the evaluations by the seeding promoters at the HUJ (e.g., Nirel and Rosenfeld 1995). The results found by the panel are shown in Figure 1.
Should the lack of seeding results after 1990 shown in Figure 1 surprise? I don’t think so. This sequence of optimistic claims by seeding experimenters concerning their own experiments followed by reanalyses by external skeptics that find the original claims were “scientific mirages” (Foster and Huber 1997, Judging Science) is a pathology within the cloud seeding realm that has dogged it since its earliest days (e.g., Brier and Enger 1952, versus MacCready 1952).
In view of Figure 1, one must ask, “What if there had been no RH95a”?
We suspect that not citing our independent re-analyses of Israel-1 and Israel-2, Silverman’s (2001) conclusions concerning the first two Israeli experiments, and Wurtele (1971) who first drew the attention to a major red flag in Israel-1, combined with the fact that the HUJ experimenters failed to even understand the precipitating nature of their own clouds for decades with all the tools at their command, all pose monumental science embarrassments for Israel, their scientists, and for the prestigious HUJ from which the faulty reports emanated.
Can there be other reasons for not citing the work of external, foreign workers who overturned benchmark experimental science by the home country’s scientists?
- Did the background airborne microphysical measurements that preceded Israel-4 justify a new experiment?
B23 cite Freud et al. (2015) as having demonstrated cloud seeding potential in the mountainous north region of Israel through a series of airborne flights; but did it support the idea of strong cloud seeding potential as B23 assert?
No.
I was not asked to review Freud et al. 2015, as one might have expected given my background. Nevertheless, I carried out a post publication “comprehensive review” that can be found under 2017 here.
Freud et al. 2015 was a “Jekyll and Hyde” read; some of the best reporting by the HUJ’s cloud seeding unit was contained in it. But it also contained misleading statements. My recommendation after reading what I considered to be a strongly biased study that was going to mislead the INWA concerning cloud seeding potential: “Don’t do a cloud seeding experiment in northern Israel based on the research of Freud et al. (2015)!”
As the INWA could have suspected, Freud et al. (2015) would not be the first time that cloud seeding researchers at the HUJ had misled the INWA about the clouds of Israel being filled with cloud seeding potential. My conclusion regarding the false picture of “abundant” cloud seeding potential in the northern mountains of Israel painted by Freud et al. 2015 was, in essence, affirmed post facto by the “primary” results of Israel 4. The “abundant” cloud seeding potential in northern Israel described by Freud et al. (2015) was not realized or was imaginary to begin with.
A caveat on airborne sampling: One can “lie” with aircraft measurements by sampling only newly risen turrets and avoiding those that are maturing or in aged states with appreciable ice particle concentrations. Gagin and Neumann (1974), for example, stated that they chose only newly risen turrets, narrow ones at that, and flew research flights on mostly dry days, and those choices misled them and the rest of the scientific community regarding the microstructure of Israeli clouds and their cloud seeding potential. Significant rain days in Israel are comprised of large complexes of convective clouds in various stages of development, “tangled masses,” as they were described by Neumann et al. (1967). To their credit, Freud et al. informed the reader that they sampled only newly risen turrets when reporting the low (<2 per liter) modal ice particle concentrations in those turrets.
Freud et al.’s measurements could not have been more incompatible with uncited by B23 measurements of Levin (1992: 1994; Levin et al. 1996). Tens to hundreds per liter of ice particles were found in six flights on four days in clouds having tops >-13°C. Freud et al. 2015 could not bring themselves to inform their readers of similar high ice particle concentrations that they likely encountered during their 27 flights (that is, if they did not deliberately avoid those high ice particle concentration regions). Freud et al. 2015, therefore, may be a first in the evaluation of cloud seeding potential in which measurements of ice particle concentrations in mature and aging clouds were not reported; the absence of such data made their entire report unreliable.
One of the B23 co-authors (DR) has claimed that ice particle measurements measured in their airborne research were “unreasonably high” in Israeli clouds due to probe caused shattering of ice crystals and thus weren’t reliable. D. Axisa, a representative of the manufacturer, Droplet Measurement Systems, of the CAPS probe used by Freud et al. (2015) stated that this statement was false: “They could have reported accurate ice particle concentrations if they had wanted to.” Dr. Axisa is a former president of the Weather Modification Assoc. It seems likely that HUJ researchers are once again withholding vital information on the clouds of Israel6.
- What do we know about cloud seeding in Israel today?
What we know today is that if careful, skeptical and independent analyses of Israel-1 and Israel-2 experiments and equally careful evaluations of the clouds of Israel had been done in the first place by independent Israeli scientists or ones outside Israel that are non-partisan cloud seeding scientists (as was carried out by RH95a, R88, and by Silverman 2001), there would not have been 30 plus years of wasted operational cloud as would be found by independent evaluators in the decades ahead (Kessler et al. 2006, Sharon et al. 2008). Fortunately, we need not guess whether those 10s of millions of dollars were wasted on the seeding of Lake Kinneret watersheds. They were. Inexplicably, the INWA drove through the “stop sign” presented by Kessler et al. (2006) and commercially seeded around Lake Kinneret for another seven years after this report came out according to B23.
- Why hasn’t cloud seeding worked In Israel?
Answer: too much natural ice formation in clouds.
B23 failed to mention that the “ripe-for-seeding” cloud foundation for the statistical results of Israel-1 and Israel-2 no longer exists. The mythical clouds described by HUJ researchers were critical in the acceptance of the Israeli cloud seeding rain increases by the scientific community, as quoted in Kerr (1982) earlier and by Dennis (1989).
A review of the Israeli cloud microstructure shows that they are “ripe,” but not for cloud seeding, but for an explosion of ice as the tops ascend to temperatures below -5°C and age. In most cases, precipitation-sized drops have already formed when the Israeli cloud ascend through this level (Gagin and Neumann 1974, Figure 13.4), and the concentration of cloud droplets exceeding the Hallett-Mossop riming-splintering criterion of >23 µm diameter can be inferred to be copious in that -2.5° to -8°C temperature zone. Furthermore, there is an enhancement of the H-M process when droplets <13 µm are present (Goldsmith et al. 1976, Mossop 1985) and such drops would be present in the semi-polluted air masses; initially, shallow cold layers diluted by the warming of the Mediterranean Sea to depths of 3-9 km on shower/thunderstorm days by the time they reach Israel under cold polar troughs.
Without the “ripe for seeding” clouds, ones with great seeding potential to cloud top temperatures as low as -21°C as described by Gagin and Neumann (1976, 1981 and Gagin 1981), there can be no viable increases in rainfall due to cloud seeding. This does not mean that some small, slightly supercooled clouds can’t be seeded to make small amounts of rain as noted by the HUJ researchers, Gagin and Neumann (1981), and by Sharon et al. (2008). However, those small amounts weren’t deemed viable for a cloud seeding operations.
- The nature of the reporting of the experiments by the HUJ cloud seeding researchers
The omission of the south target result (Gagin and Neumann 1976, 1981) was tantamount to the cancer researcher who only reports on the 50 mice his treatment cured while not reporting on the 50 mice that died from the same treatment. This kind of behavior in virtually every field but weather modification/cloud seeding, would be termed, “scientific misconduct,” specifically of a type called, “falsification” when data are omitted or adjusted (Ben-Yehuda and Oliver-Lumerman 2017, Fraud and Misconduct in Research)6. Inexplicably, Prof. K. Ruben Gabriel, the Israeli cloud seeding statistician, acquiesced in this omission as a reviewer of Gagin and Neumann’s 1981 paper in which this critical omission occurred.
Moreover, reporting the apparent negative effect on rainfall in the south target of Israel-2 would have raised numerous questions about the clouds of Israel: How could seeding Israeli clouds, described as being filled with great seeding potential as had been repeatedly described by the HUJ researchers, have resulted in what appeared to be a large decrease in rainfall in the south target on seeded days? Cloud tops in the south target in Israel average higher temperatures than those in the north (e.g., GN74; RH95a) making findings of decreased rainfall due to cloud seeding (as Rosenfeld 1989, Rosenfeld and Farbstein 1992 suggested) even harder to explain.
Moreover, while interim “positive” reports of cloud seeding increases in rain emanated from the HUJ during Israel-1 and Israel-2, HUJ researchers clearly felt differently about reporting indications of rain decreases in Israel-2 and Israel-3. For example, the scientific community was not informed of the suggestion of decreased rain due to clouds seeding in Israel-3 by the HUJ experimenters until 17 years after randomized seeding had begun (Rosenfeld and Farbstein 1992). Is this what the HUJ stands for? This chronology demonstrates a pattern that HUJ experimenters have had reporting suggestions of decreased rainfall or null results due to cloud seeding and in correcting their flawed cloud microstructure reports to the scientific community and to their countrymen in the years prior to B23.
Moreover when “good draws” or null results are suggested, the HUJ researchers reach for the magic bag to explain why “cloud seeding did it,” not nature. For example, when the Israel-1 chief meteorologist provided a plume analysis that the buffer zone (BZ) of Israel-1 could not have been appreciably contaminated by inadvertent seeding (a conclusion also supported by Neumann et al. 1967), Gagin and Neumann (1974), however, countered with an opposite explanation; the BZ had surely been contaminated on Center seeded days. The reason and data behind these two different explanations for the difference in the two plume analyses was not given except in general unsatisfactory terms.
When a Type I error and massive “good draw” affected the north seeded days of Israel-2 that also brought heavy rain to the south target, the crossover null result was then explained as due to “dust/haze” that produced different cloud microstructures when present in each target, first proposed by Rosenfeld (1989) in an HUJ report.
When RH95a showed that the results of seeding on the coast of Israel in Israel-1 were too close to the cloud base seeding release point to have resulted in rain practically falling on top on the seeding aircraft that flew in a line along the coast, Rosenfeld (1997) wrote a magical explanation filled with conjectures, one requiring nine steps to be fulfilled to explain the troublesome indication of rain increases in the BZ and in the coastal zone on Center seeded days. Please see my extended “Reply,” p11, to the hypothesis of Rosenfeld (1997) at:
http://carg.atmos.washington.edu/sys/research/archive/1997_comments_seeding.pdf.
When the independent panel, Kessler et al. 2002, could find no viable increases in rain in the seeding of the Lake Kinneret watersheds in their interim report, the HUJ seeding team then asserted that “air pollution” was suddenly( after 1990) decreasing rain as much as cloud seeding was increasing it (Givati and Rosenfeld (2005). One might ask, “what happened to ‘dust/haze’”?
Ice crystal concentrations measured in Israeli clouds by our best instruments are “unreasonably high” according to B23 co-author, Rosenfeld (private communication, 2018). Rosenfeld’s statement, however, contrasts with that of Droplet Measurement Technologies (DMT), the manufacturer of the Cloud, Aerosol and Precipitation Spectrometer (CAPS) probe used by the HUJ researchers: “They could have reported accurate ice particle concentrations if they had wanted to” (D. Axisa, DMT scientist, personal communication, 2018).
With the certainty of dust/haze days and incoming Israeli shower clouds affected by “sea spray” as Freud et al. 2015 described on shower days during the time the HUJ experimenters were flying their research aircraft in the early 1970s, monitoring storms with their radars, or examining rawinsondes during rain spells, we can conclude confidently that the lack of reporting on shallow precipitating clouds that occurred regularly in Israel is one of the more inexplicable and troubling aspects in the reporting of the Israeli cloud seeding experiments.
Deepening this enigma is that for two winter seasons in the late 1970s, the experimenters measured the depth of raining clouds with a vertically pointed 3-cm wavelength radar with research aircraft overflights to verify accuracy (Gagin 1980). Dr. Rosenfeld, a B23 co-author who studied clouds and radar imagery at this time, is the sole living person who can tell us what happened (Rosenfeld 1980, master’s thesis). One must necessarily ask if the HUJ experimenters discovered clouds they “didn’t like,” and withheld that information from us as they did the results of seeding in the south target of Israel-2? Without conjuring up a stupefying degree of incompetence, it seems likely.
It is not science that we are dealing with concerning the reporting by the HUJ cloud seeding researchers. There will ALWAYS be another problem that prevented seeding from working and if only corrected, seeding will work, as we are sure to learn when the inevitable “secondary” results of Israel-4 are published.
Will I be given a chance to review an Israeli cloud seeding manuscript as an expert in Israeli clouds, weather, and cloud seeding? It seems unlikely with the journal atmosphere we have today.
- The on-going journal problem of “one-sided citing” as seen in B23; the equivalent of today’s “cancel culture”
The omission of the work by myself and with Prof. Peter V. Hobbs was shocking to see in B23 since all the B23 authors knew of this work. In human terms, external skeptics from a foreign country that expose faulty science in another country are not going to be exactly welcomed (or apparently cited) by that’s country’s scientists when a scientific embarrassment unfolds, as has happened in Israel concerning cloud seeding. While this may seem like an outlandish claim, what happened could be interpreted as tinged with nationalism has previously been shown to obfuscate science (Broad and Wade 1982, p114).
For journal readers who are used to “one-sided citing” in partisan media, our scientific journals are supposed to be immune from these acts due to a peer-review “filter” that is supposed to eliminate this practice before an article reaches the publication stage.
a). Why do authors, like B23, tell only one side of the story?
In the words of Ben-Yehuda and Oliver-Lumerman (2017) of the HUJ, such deceptions are, “…a deliberate attempt to create a false reality, persuade audiences that these realities are valid, and enjoy the benefits that accompany scientific revelations, whether those of prestige, money, reputation, or power….” The effect of one-sided citing on journal readers is well expressed in the U. S. Federal Trade Commission’s (FTC) statement on consumer fraud:
“Certain elements undergird all deception cases. First, there must be a representation, omission or practice that is likely to mislead the consumer [journal reader].”
For the reader, one-sided citing, if it is not obvious, is purposefully done by authors to hide results that they do not want you to see. In effect, B23 performed the same act as Gagin and Neumann (1981) did when the latter authors did not report the results of random seeding of the south target of Israel-2, results that they did not want the world to see, and results that would have raised so many questions.
Regrettably, one-sided citing (a form of deception) is widely observed in Amer. Meteor. Soc. journals and in J. Weather Modification articles on cloud seeding/weather modification:
https://cloud-maven.com/journal-citing-practices-in-a-controversial-domain-cloud-seeding/
B23 practiced one-sided citing (defined by Schultz 2009) in their article concerning the Israel-1 and Israel-2 experiments. Inexplicably, our groundbreaking work (e.g., R88, RH88, who pointed out how anomalous the Israeli cloud reports were compared to other clouds, and RH95a) went uncited by B23. Our work, in toto, can be said to have anticipated the both the null result of decades of operational seeding of Lake Kinneret (Kessler et al. 2006, Sharon et al. 2008) and the null “primary” result of Israel-4 reported by B23.
B23 repeatedly misled/deceived readers, the “consumers” of journal science, concerning Israel-1 and Israel-2. If there is something different than what was done by B23 than what is described by the FTC above its not apparent.
Nor did B23 cite Wurtele (1971), Silverman (2001) or mention the critical airborne cloud measurements by one of Israel’s own leading scientists, Levin 1992, 1994, and Levin et al. 1996). The latter measurements were the first cloud ice measurements in Israel since Gagin (1975). Those new, independently acquired cloud ice measurements supported the conclusions in R88, RH88, and those in RH95a, all which contravened the many HUJ experimenters’ fictitious reports of “ripe for seeding” clouds whose tops could ascend to ~-20°C without precipitating.
Later measurements of cloud properties via satellite would also confirm the independent cloud measurements and assessments; that the clouds of Israel formed precipitation far more readily and at much higher cloud top temperatures (Ramanathan et al. 2001) than the HUJ experimenters could discern over many decades.
In 2015, the HUJ cloud researchers discovered that “sea spray” in the Mediterranean makes the cumuliform clouds invading Israel precipitate more efficiently and at the high cloud top temperatures like those reported in R88 (Freud et al. 2015). We can be quite sure that Mediterranean Sea spray has been occurring and affecting clouds that move into Israel for millions of years, and of course, did so during the 1970s when the HUJ scientists were performing their aircraft and radar cloud studies. Yet, they could not detect, or did not report, on those clouds that would have erased most of their seeding potential.
The shame of one-sided citing in B23 is that the authors could have added a single sentence following their repeated claims of rain increases in Israel-1 and -2: “However, these results, and the cloud reports that gave the statistical results credibility, have been questioned/overturned,” followed by a string of citations.
But B23 could not bring themselves to do that.
b) Why should we care about one-sided citing?
Knowledgeable readers of a specific topic like this writer will know that an article has been skewed to deliberately mislead readers due to omissions of contrary findings that go against what the authors assert. But less informed readers will not know, and their knowledge will be truncated regarding an important public policy, as when their state or local government considers a cloud seeding program. They will want to know the unabridged findings about the Israeli experiences as a tale of caution about accepting claims by promoters of seeding that have not been closely scrutinized by outside experts.
Moreover, “one-sided citing” sullies the reputations of all the authors even those who may not have agreed with doing it, and likewise sullies the reputations of institutions represented by the authors who practice it by suggesting that those institutions do not uphold standard science practices by those who work there. It also damages the authors whose work goes uncited since one’s impact in science is measured by citation metrics. Finally, even the journal in which one-sided citing occurs can be considered to have been damaged since unreliable findings have been published in it.
Nevertheless, it would appear that reviewers, editors, and journal management do not care so much about this issue. No statement in our Amer. Meteor. Soc. ethics statement addresses the question of the pernicious practice of one-sided citing as seen in B23. Its intellectually dishonest to omit relevant findings for your science audience just because you don’t like them
c) Who’s responsible for “one-sided” citing in journals?
“One-sided” citing, specifically as observed in B23, is due to poor peer reviews of manuscripts by seeding partisans or reviewers ignorant of the literature they are supposed to know. However, it is also due to those that do know the literature but do not get those manuscripts to review. For example, even though I would be deemed an expert on Israeli clouds, weather, cloud seeding, and on cloud microstructure, I was inexplicably not asked to review a manuscript in my specialty; that by B23 which would have made these comments unnecessary.
The reviewers of B23 manuscript were either ignorant of the literature they were supposed to be knowledgeable about or were cloud seeding partisans that also desired that the “other side” of the story for Israel-1 and Israel-2, as represented in the peer-reviewed literature by R88, RH88, RH95a, RH97a, b, c, d, e, Silverman (2001), Wurtele (1971) and Levin’s cloud measurements (e.g., Levin et al. 1996), be hidden from the journal readers.
At the top of the “responsibility pyramid” for one sided citing in journal articles, however, must reside the editor of the journal who chose the reviewers that allowed this to happen. Whomever this was at the J. Appl. Meteor. Climate, should not be allowed to be an editor who disburses cloud seeding manuscripts again.
d) Concluding remarks on one-sided citing
While all the B23 authors are technically responsible for its misleading content, one suspects some were likely “drug along” by stronger author personalities or authors who have funding power over them. As is done in Geophys. Res. Letts., the actual contributions of each author to this article should have been listed so we can truly know who was responsible for providing one-sided histories for Israel-1 and Israel-2 and other misleading statements.
We know, too, seeding partisans at the HUJ that have cost their own country so much will not let the “primary” null result of B23 stand; there will be “secondary” and “tertiary” stratifications of Israel-4 data perhaps designed to mislead the INWA into another randomized cloud seeding experiment or to resume operational seeding of Lake Kinneret.
It will be critical that if a new experiment is conducted at the behest of the HUJ seeding partisans, that outside, independent experts conduct it! It is also critical that prior to a new experiment that new airborne measurements of the clouds of Israel also be undertaken by outside, independent and experienced researchers in view of the problems that researchers at the HUJ have had over several decades, right up to today, in reporting ice particle concentrations in their clouds and their clouds’ actual seeding potential.
The major question we must now confront to avoid further science mischief by HUJ cloud seeding researchers, is how was it that they were not aware of the natural state of their clouds, namely, that clouds with tops warmer than -10°C that regularly rained, a finding that seriously limits cloud seeding potential? To date, no explanation has been put forward. And what evidence will they skew or miss in a likewise manner in the inevitable Israel-4, “secondary” results article?
=========================
Lastly, a note of scientific etiquette for B23 and young researchers: B23 cite the work of French et al. (2018) in demonstrating cloud seeding efficacy via the use of mm-wavelength radar.
The first use of mm-wavelength radar of the type used by French et al. (2018) was used by the Cloud and Aerosol Group at the University of Washington in a “proof of concept” experiment (Hobbs et al. 1981). Scientific etiquette means citing those that went first (Schultz 2009) Thus, a citation to the Hobbs et al. (1981) article should have preceded that of French et al. 2018)8. Our experiment proved that cloud seeding works in limited situations as in those described by French et al. (2018).
=================FOOTNOTES=====================
1The Israeli experiments have had several names over their history. We use the latest terms for them here, e.g., Israel-1, etc.
2Pressure was applied in 1986 on the HUJ researchers by the Israeli experiments’, “Chief Meteorologist,” Mr. Karl Rosner, who began a letter writing campaign to have the important results of seeding in the south target published by Prof. Gagin. Mr. Rosner told Professor Hobbs and myself in a Seattle visit that he felt that Prof. Gagin’s co-author, Jehuda Neumann, was “drug along” as a co-author of Gagin’s papers.
3This author believes that it is critical that a certified copy of the list of random decisions for Israel-2 be compared against those days used in the experiment. The remarkably unlikely random draw described by Gabriel and Rosenfeld (1990) could be explained if the original list was violated by the experimenters: draws were made and assigned to “seed” days when heavy storms were forecast.
4Rosenfeld (1989) in an unpublished HUJ report argued that the divergent apparent effects of cloud seeding were real.
5The findings of Kessler et al. were challenged by seeding partisans at the HUJ and who claimed that “air pollution” had decreased rain as much as cloud seeding had increased it after 1990. While this was a convenient explanation, it was not found credible by many subsequent independent investigators, including by Kessler et al. (2006).
6Ben-Yehuda and Oliver-Lumerman’s 2020 book, Fraud and Misconduct in Research, should be required reading for B23. Ben-Yehuda and Oliver-Lummerman are professors at the HUJ.
7I suggested the use of our vertically pointed, mm-wavelength radar for cloud seeding use to Prof. Larry Radke and Peter Hobbs, after seeing virga signatures pass overhead of that radar, realizing that creating lines of seeding in supercooled cloud layers that passed over such a radar could prove the viability of cloud seeding in a new way. I also carried out this experiment as flight scientist/meteorologist in the seeding/monitoring aircraft. However, I was not credited for this idea by Prof. Hobbs in the Science article.
==================REFERENCES====================
Ben-Yehuda, N., and A. Oliver-Lumerman, 2020: Fraud and Misconduct in Research. University of Michigan Press, 266pp. No doi.
Benjamini, Y, A. Givati, P. Khain, Y. Levi, D. Rosenfeld, U. Shamir, A. Siegel, A. Zipori, B. Ziv, and D. M. Steinberg, 2023: The Israel 4 Cloud Seeding Experiment: Primary Results. J. Appl. Meteor. Climate, 62, 317-327. https://doi.org/10.1175/JAMC-D-22-0077.1
Brier, G. W., and I. Enger, 1952: An analysis of the results of the 1951 cloud seeding operations in central Arizona. Bull. Amer. Meteor. Soc., 23, 208-210. https://doi.org/10.1175/1520-0477-33.5.208
Broad, W. J., and N. Wade, 1982: Betrayers of the Truth: Fraud and Deceit in the Halls of Science. Simon and Schuster, 256pp. No doi.
Dennis, A. S., 1989: Editorial to the A. Gagin Memorial Issue of the J. Appl. Meteor., 28, 1013.
Foster, K. R., and P. W. Huber, 1997: Judging Science: Scientific Knowledge and the Federal Courts. The MIT Press, 333pp. No doi.
French, J.R., Friedrich, K., Tessendorf, S.A., Rauber, R.M., Geerts, B., Rasmussen, R.M., Xue, L., Kunkel, M.L. and Blestrud, D.R., 2018: Precipitation formation from orographic cloud seeding. Proc. Natl. Acad. Sci., 115, 1168-1173.
https://doi.org/10.1073/pnas.1716995115
Freud, E., H. Koussevitsky, T. Goren, D. Rosenfeld, 2015: Cloud microphysical background for the Israel-4 cloud seeding experiment. Atmos. Res., 158-159, 122-138. https://doi.org/10.1016/j.atmosres.2015.02.007
Gabriel, K. R., 1967a: The Israeli artificial rainfall stimulation experiment: statistical evaluation for the period 1961-1965. Vol. V., Proc. Fifth Berkeley Symp. on Mathematical Statistics and Probability, L. M. Le Cam and J. Neyman, eds., University of California Press, 91-113.
___________, 1967b: Recent results of the Israeli artificial rainfall stimulation experiment. J. Appl. Meteor., 6, 437-438.
https://doi.org/10.1175/1520-0450(1967)006%3C0437:RROTIA%3E2.0.CO;2
____________, and R. Baras, 1970: The Israeli rainmaking experiment 1961-1967 final statistical tables and evaluation. Tech. Rep., Hebrew University, Jerusalem, 47pp. No doi.
_____________, and Rosenfeld, D., 1990: The second Israeli rainfall stimulation experiment: analysis of precipitation on both targets. J. Appl. Meteor., 29, 1055-1067. https://doi.org/10.1175/1520-0450(1990)029%3C1055:TSIRSE%3E2.0.CO;2
Gagin, A. 1975: The ice phase in winter continental cumulus clouds. J. Atmos. Sci., 32, 1602-1614. https://doi.org/10.1175/15200469(1975)032%3C1604:TIPIWC%3E2.0.CO;2
Gagin, A., 1980: The relationship between depth of cumuliform clouds and their raindrop characteristics. J. Rech. Atmos., 14, 409-422. No doi.
Gagin, A., 1981: The Israeli rainfall enhancement experiments. A physical overview. J. Wea. Mod., 13, 1-13. https://doi.org/10.54782/jwm.v13i1.41
Gagin, A., 1986: Evaluation of “static” and “dynamic” seeding concepts through analyses of Israeli II and FACE-2 experiments. In Precipitation Enhancement–A Scientific Challenge, Meteor. Monog.,21, No. 43, Amer. Meteor. Soc., 63-70.
https://doi.org/10.1175/0065-9401-21.43.63
_______., and K. R. Gabriel, 1987: Analysis of recording gage data for the Israeli II experiment Part I: Effects of cloud seeding on the components of daily rainfall. J. Appl. Meteor., 26, 913-926.
https://doi.org/10.1175/1520-0450(1987)026%3C0913:AORRDF%3E2.0.CO;2
_______, and J. Neumann, 1974: Rain stimulation and cloud physics in Israel. Climate and Weather Modification, W. N. Hess, Ed., Wiley and Sons, New York, 454-494. No doi.
_______, and _______, 1976: The second Israeli cloud seeding experiment–the effect of seeding on varying cloud populations. Proc. 2nd WMO Scientific Conf. on Weather Modification, Boulder, 195–204. World Meteor. Organization, 41 Ave. Giuseppe Moffs, Geneva 2, Switzerland.) No doi.
________, and ________, 1981: The second Israeli randomized cloud seeding experiment: evaluation of results. J. Appl. Meteor., 20, 1301-1311.
https://doi.org/10.1175/1520-0450(1981)020%3C1301:TSIRCS%3E2.0.CO;2
Givati, A., and Rosenfeld, D., 2005: Separation between cloud-seeding and air pollution effects. J. Appl. Meteor. Climate, 44, 1298-1314.
https://doi.org/10.1175/JAM2276.1
Goldsmith, P., Gloster, J., and C. Hume, 1976: The ice phase in clouds. Preprint Volume, Intern. Conf. on Cloud Physics, Boulder, Amer. Meteor. Soc., Boston, MA. 163-167.
Hobbs, P. V., and A. L. Rangno, 1978: A reanalysis of the Skagit cloud seeding project. J. Appl. Meteor., 17, 1661–1666.
https://doi.org/10.1175/1520-0450(1978)017%3C1661:AROTSC%3E2.0.CO;2
_________, and _________, 1979: Comments on the Climax randomized cloud seeding experiments. J. Appl. Meteor., 18, 1233-1237.
https://doi.org/10.1175/1520-0450(1979)018%3C1233:COTCAW%3E2.0.CO;2
_________, and __________, 1985: Ice particle concentrations in clouds. J. Atmos. Sci., 36, 2523-2549. https://doi.org/10.1175/1520-0469(1985)042%3C2523:IPCIC%3E2.0.CO;2
_________, and ___________, 1990: Rapid development of ice particle concentrations in small polar maritime cumuliform clouds. J. Atmos. Sci., 47, 2710-2722.
https://doi.org/10.1175/1520-0469(1990)047%3C2710:RDOHIP%3E2.0.CO;2
_________, Lyons, J. H., Locatelli, J. D., Biswas, K. R., Radke, L. F., Weiss, R. W., Sr., and A. L. Rangno, 1981: Radar detection of cloud-seeding effects. Science, 213, 1250-1252. https://doi.org/10.1126/science.213.4513.1250
Kerr, R. A., 1982: Cloud seeding: one success in 35 years. Science, 217, 519–522. https://doi.org/10.1126/science.217.4559.519
Kessler, A., A. Cohen, D. Sharon, 2006: Analysis of the cloud seeding in northern Israel. A report submitted to the Israel Hydrology Institute and the Israel Water Management of the Ministry of Infrastructure, In Hebrew with an English abstract). 117pp. No doi.
Koenig, L. R., 1963: The glaciating behavior of small cumulonimbus clouds. J. Atmos. Sci., 20, 29-47.
https://doi.org/10.1175/1520-0469(1963)020%3C0029:TGBOSC%3E2.0.CO;2
__________, 1977: The rime-splintering hypothesis of cumulus glaciation examined using a field-of-flow cloud model. Quart. J. Roy. Meteorol. Soc., 103, 585-606.
https://doi.org/10.1002/qj.49710343805
__________, 1984: Further theoretical studies of the role of splintering in cumulus glaciation, Quart. J. Roy. Meteorol. Soc., 110, 1121-1141.
https://doi.org/10.1002/qj.49711046621
Levin, Z., 1992: The role of large aerosols in the precipitation of the eastern Mediterranean. Paper presented at the Workshop on Cloud Microphysics and Applications to Global Change, Toronto. (Available from Dept. Atmos. Sci., University of Tel Aviv). No doi.
_______, 1994: The effects of aerosol composition on the development of rain in the eastern Mediterranean. WMO Workshop on Cloud Microstructure and Applications to Global Change, Toronto, Ontario, Canada, World Meteor. Org., 115-120. No doi.
_______, E. Ganor, and V. Gladstein, 1996: The effects of desert particles coated with sulfate on rain formation in the eastern Mediterranean. J. Appl. Meteor., 35, 1511-1523.
https://doi.org/10.1175/1520-0450(1996)035%3C1511:TEODPC%3E2.0.CO;2
_________., N. Halfon, and P. Alpert, 2010: Reassessment of rain enhancement experiments and operations in Israel including synoptic considerations. Atmos. Res., 97, 513-525.
http://dx.doi.org/10.1016/j.atmosres.2010.06.011
MacCready, P., 1952: Results of cloud seeding in central Arizona, winter 1951. Bull. Amer. Meteor. Soc., 23, 48-52.
Mason, B. J., 1982: Personal Reflections on 35 Years of Cloud Seeding. Contemp. Phys., 23, 311-327. http://doi-org/10.1080/00107518208237084
Mossop, S. C., 1985: Secondary ice particle production during rime growth: the effect of drop size distribution and rimer velocity. Quart J. Roy. Meteor. Soc., 111, 1113-1124. http://DOI-org/10.1002/qj.49711147012
National Research Council-National Academy of Sciences, 1973: Weather and Climate Modification: Progress and Problems, T. F. Malone, Ed., Government Printing Office, Washington, D. C., 258 pp.
Neumann, J., K. R. Gabriel, and A. Gagin, 1967: Cloud seeding and cloud physics in Israel: results and problems. Proc. Intern. Conf. on Water for Peace. Water for Peace, Vol. 2, 375-388. No doi.
Nirel, R., and D. Rosenfeld, 1995: Estimation of the effect of operational seeding on rain amounts in Israel. J. Appl. Meteor., 34, 2220-2229.
https://doi.org/10.1175/1520-0450(1995)034%3C2220:EOTEOO%3E2.0.CO;2
Ramanathan, V., P. J. Crutzen, J. T. Kiehl, and D. Rosenfeld, 2001: Aerosols, climate and the hydrological cycle. Science, 294, 2119-2124.
https://doi-org/10.1126/science.1064034
___________, 1979: A reanalysis of the Wolf Creek Pass cloud seeding experiment. J. Appl. Meteor., 18, 579–605.
https://doi.org/10.1175/1520-0450(1979)018%3C0579:AROTWC%3E2.0.CO;2
___________, 1986: How good are our conceptual models of orographic clouds? In Precipitation Enhancement–A Scientific Challenge, R. R. Braham, Jr., Ed., Meteor. Monographs, 43, Amer. Meteor. Soc., 115-124. (An invited paper.) https://doi.org/10.1175/0065-9401-21.43.115
___________, 1988: Rain from clouds with tops warmer than -10 C in Israel. Quart J. Roy. Meteor. Soc., 114, 495-513.
https://doi-org/10.1002/qj.49711448011
__________, 2000: Comment on “A review of cloud seeding experiments to enhance precipitation and some new prospects.” Bull. Amer. Meteor. Soc., 81, 583-585. No doi
__________, and P. V. Hobbs, 1980a: Comments on “Randomized seeding in the San Juan mountains of Colorado.” J. Appl. Meteor., 19, 346-350.
https://doi.org/10.1175/1520-0450(1980)019%3C0346:COCSIT%3E2.0.CO;2
__________, and _________, 1980b: Comments on “Generalized criteria for seeding winter orographic clouds”. J. Appl. Meteor., 19, 906-907.
https://doi.org/10.1175/1520-0450(1980)019%3C0906:COCFSW%3E2.0.CO;2
__________, and _________, 1981: Comments on “Reanalysis of ‘Generalized criteria for seeding winter orographic clouds’”, J. Appl. Meteor., 20, 216.
https://doi.org/10.1175/1520-0450(1981)020%3C0216:COOCFS%3E2.0.CO;2
__________, and _________, 1983: Production of ice particles in clouds due to aircraft penetrations. J. Climate Appl. Meteor., 22, 214-232.
https://doi.org/10.1175/1520-0450(1983)022%3C0214:POIPIC%3E2.0.CO;2
__________, and _________, 1987: A re-evaluation of the Climax cloud seeding experiments using NOAA published data. J. Climate Appl. Meteor., 26, 757-762.
https://doi.org/10.1175/1520-0450(1987)026%3C0757:AROTCC%3E2.0.CO;2
__________, and _________, 1988: Criteria for the development of significant concentrations of ice particles in cumulus clouds. Atmos. Res., 22, 1-13. No doi.
__________, and _________, 1991: Ice particle concentrations in small, maritime polar cumuliform clouds. Quart J. Roy. Meteorol. Soc., 117, 207-241. https://doi.org/10.1002/qj.49711749710
__________, and _________, 1993: Further analyses of the Climax cloud-seeding experiments. J. Appl. Meteor., 32, 1837-1847.
https://doi.org/10.1175/1520-0469(1990)047%3C2710:RDOHIP%3E2.0.CO;2
__________, and _________, 1995: A new look at the Israeli cloud seeding experiments. J. Appl. Meteor., 34, 1169-1193.
https://doi.org/10.1175/1520-0450(1995)034%3C1169:ANLATI%3E2.0.CO;2
__________, and _________, 1997a: Reply to Rosenfeld. J. Appl. Meteor., 36, 272-276.
https://doi.org/10.1175/1520-0450(1997)036%3C0272:R%3E2.0.CO;2
__________, and _________, 1997b: Comprehensive Reply to Rosenfeld, Cloud and Aerosol Research Group, Department of Atmospheric Sciences, University of Washington, 25pp. http://carg.atmos.washington.edu/sys/research/archive/1997_comments_seeding.pdf
__________, and _________, 1997c: Reply to Dennis and Orville. J. Appl. Meteor., 36, 279. https://doi.org/10.1175/1520-0450(1997)036%3C0279:R%3E2.0.CO;2
__________, and _________, 1997d: Reply to Ben-Zvi. J. Appl. Meteor., 36, 257-259.
https://doi.org/10.1175/1520-0450(1997)036%3C0257:R%3E2.0.CO;2
__________, and _________, 1997e: Reply to Woodley. J. Appl. Meteor., 36, 253-254.
https://doi.org/10.1175/1520-0450(1997)036%3C0253:R%3E2.0.CO;2
___________, ___________, and L. F. Radke, 1977: Tracer and diffusion and cloud microphysical studies in the American River basin. Final Report to the Bureau of Reclamation, Final Report to the Division of Atmospheric Water Resources Management, Bureau of Reclamation, Contract 6-07-DR-20140), University of Washington, Seattle, WA, 98195.
Rosenfeld, D., 1980: Characteristics of Rain Cloud Systems in Israel Derived from Radar and Satellite Images. M. S. Thesis, The Hebrew University of Jerusalem, 129pp. (Available from the Department of Meteorology, Hebrew University of Jerusalem, Jerusalem, Israel).
_________, 1989: The divergent effects of cloud seeding uncer different physical conditions in Israeli 1 and 2 experiments. Dept. Atmos. Sci., Hebrew University of Jerusalem, May 1989. No doi.
___________, 1998: The third Israeli randomized cloud seeding experiment in the south: evaluation of the results and review of all three experiments. Preprints, 14th Conf. on Planned and Inadvertent Wea. Modif., Everett, Amer. Meteor. Soc. 565-568. No doi.
Rosenfeld, D., and H. Farbstein, 1992: Possible influence of desert dust on seedability of clouds in Israel. J. Appl. Meteor., 31, 722-731.
https://doi.org/10.1175/1520-0450(1992)031%3C0722:PIODDO%3E2.0.CO;2
Schultz, D. M., 2009: Eloquent Science: A practical guide to becoming a better writer, speaker, and atmospheric scientist. Amer. Meteor. Soc. pp412.
Sharon, D., A. Kessler, A. Cohen, and E. Doveh, 2008: The history and recent revision of Israel’s cloud seeding program. Isr. J. Earth Sci., 57, 65-69. https://DOI.org/10.1560/IJES.57.1.65.
Silverman, B. A., 2001: A critical assessment of glaciogenic seeding of convective clouds for rainfall enhancement. Bull. Amer. Meteor. Soc., 82, 903-923.
https://doi.org/10.1175/1520-0477(2001)082%3C0903:ACAOGS%3E2.3.CO;2
Wurtele, Z. S., 1971: Analysis of the Israeli cloud seeding experiment by means of concomitant meteorological variables. J. Appl. Meteor., 10, 1185-1192.
https://doi.org/10.1175/1520-0450(1971)010%3C1185:AOTICS%3E2.0.CO;2