Introduction
Previous studies have shown qualitatively that urban air pollution and industrial air pollution suppress precipitation-forming processes in convective clouds (Rosenfeld 1999, 2000). The pollution aerosols serve as small cloud condensation nuclei (CCN) that form large concentrations of small cloud droplets, which in turn suppress the drop-coalescence and warm-rain processes as well as ice precipitation (Rosenfeld 2000; Borys et al. 2003; Andreae et al. 2004) and so prolong the time required to convert the cloud water that exists in small drops into large hydrometeors that can precipitate. Borys et al. (2003) showed that the addition of as little as 1 μg m−3 of anthropogenic sulfate aerosols to a clean background can reduce the orographic snowfall rate in the Colorado Rocky Mountains by up to 50%. The suppression is stronger in shallower clouds with warmer top temperatures. Satellite observations showed that pollution can completely shut off precipitation from clouds that have temperatures at their tops >−10°C (Rosenfeld 1999, 2000; Rosenfeld and Woodley 2003). Therefore, it is expected that the greatest rain suppression will take place in regions that are dominated by clouds having relatively warm tops and short lifetimes downwind of major urban areas. Because of their short life, such clouds are more sensitive to slowing down of the conversion of cloud water to precipitation, whereas long-living clouds would eventually convert their water into precipitation regardless of the conversion rate.
Based on the above considerations, Givati and Rosenfeld (2004) expected that the main effect would be, therefore, the suppression of the orographic component of the precipitation, which would be manifested as a reduction in the orographic enhancement factor Ro, where Ro is defined as the ratio between the precipitation amounts at the hills and at the upwind lowland. Such conditions are abundant, especially on the west coast of continents in the subtropics and midlatitudes, where the precipitation over the hills is a major source for the scarce water there. Givati and Rosenfeld (2004) quantified the rainfall losses over hills downwind of major coastal urban areas in California and Israel.
Their main analysis tool was the time series of Ro based on annual precipitation from rain gauges downwind of major urban areas and their intercomparison with rain gauges in a similar geographical setting away from the urban areas at a direction perpendicular to the rain-bearing winds. The underlying assumption was that small-particulate air-pollution emissions, acting as CCN, have increased with the growth of the urban areas, resulting in a decrease in Ro with time. The suppression rate was found to be 15%–25% of the annual precipitation in hilly areas in California and Israel (Givati and Rosenfeld 2004). The suppression occurs mainly in the relatively shallow orographic clouds within the cold air mass of cyclones and not in the warm air mass. The Ro time series for relatively pristine areas crosswind to the polluted areas did not show any trend with time and so serve as controls to the polluted areas (e.g., areas in northern California).
The most likely alternative explanation for the reduction in Ro is a decreasing trend in the cross-mountain component of the low-tropospheric wind velocity and moisture flux during rain events. Givati and Rosenfeld (2004) applied a radiosonde regression model to predict the daily rain amounts in hilly stations. The model showed no difference between the measured rain and the predicted rain in central Israel (the southern seeded area). This result reflects that the relevant meteorological conditions during rain days did not change systematically along the years, and the observed trends in Ro are likely caused by nonmeteorological factors such as anthropogenic air pollution.
Enhancement of precipitation by cloud-seeding operations has been suggested statistically in many studies around the world in the last half-century although physical scientific proof is still lacking (Silverman 2001). The authors are not familiar with publications documenting successful seeding of winter convective cloud. In contrast, there are many reports of precipitation enhancement by orographic cloud seeding (e.g., Elliott 1986).
In terms of cloud microphysical effects, cloud seeding and air pollution are the opposite sides of the same coin. The opposite action to pollution aerosols suppressing precipitation by adding a large number of small CCN is the addition of giant CCN (i.e., hygroscopic seeding) for enhancing precipitation. In the case of glaciogenic seeding, as was done in Israel, the aim was to enhance ice precipitation, whereas air pollution, by reducing cloud drop size, also suppresses ice precipitation (Rosenfeld 1999, 2000; Rosenfeld and Woodley 2003).
Characterization of the study areas
Dynamical characterization of the rain clouds in Israel
Rainfall in Israel occurs mainly in the convective clouds that develop during winter in postfrontal cold air mass that interacts with the relatively warm surface of the Mediterranean Sea (Goldreich et al. 2004). This mechanism leads to the formation of convective clouds over sea. The land surface during winter storms is much colder than the sea. Therefore, the convective clouds lose their energy source, mature, and eventually dissipate inland. The clouds that form inland and are not orographic are usually synoptically forced and hence deep and long living, and therefore they are not susceptible to aerosols. Little new generation of convection occurs over lowlands, because of the lack of surface heating during the winter storms conditions. This description is supported by the findings of Sharon and Kutiel (1986), who showed that the rain intensity in Israel decreases from the coastline inland, regardless of topography. The convection over sea is often intense and is associated with thunderstorms. The maturation of these convective clouds when they move inland is seen in the findings of Yair and Levin (1994), who showed that the number of thunderstorms decrease from the coast inland, regardless of the elevation. Such deep convective clouds have smaller susceptibility to aerosol effects on rainfall amounts than do the shallower clouds (Phillips et al. 2002; Segal et al. 2004) that form when the air ascends over the hills inland. Rosenfeld (1986) showed that the radar-detected echo-top heights average, in Israel, about 5 km. The top heights decrease by about 400 m when moving from sea to the coastal plain and decrease by an additional 400 m over the hills.
Microphysical characterization of the rain clouds in Israel
Lahav and Rosenfeld (2000) characterized the microphysical properties of clouds in Israel using aircraft measurements and satellite images. They monitored clouds separately over sea and inland to document potential differences between the two areas. They found in several cases that the convective clouds over sea were microphysically maritime, that is, they had 200–350 cloud drops per cubic centimeter. At the same time, clouds at 10–20 km over land had a more microphysically continental nature, with local concentration exceeding 1000 drops per cubic centimeter near cloud base. The satellite-retrieved effective radius of the cloud droplets was smaller by about 2–3 μm for the same depth over land. The clouds over the sea exceeded the 14-μm precipitation threshold (Rosenfeld and Gutman 1994) at around −10°C isotherm; inland, they reached it at the −18°C isotherm. This trend of decreasing effective radius for the same cloud depth continued farther east over Jordan. It was reported for the first time by Rosenfeld and Lensky (1998), who showed a case of clouds exceeding the 14-μm precipitation threshold at the −10°C isotherm over sea off the Israeli coast. The cloud tops barely reached that threshold over Israel and were mostly below it over Jordan.
Cloud seeding in northern Israel
Two cloud-seeding experiments were carried out in northern Israel during the period of 1961–75. The seeding was done by a seeder aircraft with silver iodide acetone burners flying along a seeding line parallel to the coastline (see Fig. 1). The first Israeli experiment (Israel-1, 1961–67) had a two-target crossover design of north versus south. The seeding was randomized for each day to be done in the northern or in the southern target area. The overall result was 15% enhancement in the north and south combined, which is statistically significant (Gabriel 1967; Gagin and Neumann 1974, 1981; Gabriel and Rosenfeld 1990). Additional analyses showed that most of the effect of Israel-1 was contributed from the northern target area (Gagin and Neumann 1981; Rosenfeld 1997).
In Israel-2 (1970–75), the seeding line of the northern target area was shifted eastward inland so that a northern control area could be defined between the coastline and the seeding line. This allowed the evaluation of the seeding effect for the northern target alone, although the experiment was conducted in a crossover design as in Israel-1. Analysis of the Israel-2 experiment for the north alone showed statistically significant enhancement of 13% (Gagin and Neumann 1981) while having no effect in the south or when analyzed as a crossover experiment (Gabriel and Rosenfeld 1990).
After the two experiments achieved positive results in the north, the seeding in northern Israel continued after 1975 in operational mode, that is, seeding on all days with potential rain clouds. Nirel and Rosenfeld (1995) evaluated this operational seeding and found only a 6% enhancement for the period of 1975–90. Their main analysis was based on the assumption of stability in the historical target/control ratio. They identified a decreasing trend in that ratio but offered no physical explanation (Nirel and Rosenfeld 1995). When they repeated the analysis of the seeding effect assuming an extrapolated decreasing trend of the not-seeded target/control ratio into the operational seeding period, the operational seeding effect was calculated at 11% (Nirel and Rosenfeld 1995).
Cloud seeding in southern Israel
During the Israel-1 and Israel -2 experiments, the seeded days in the north were the unseeded days in the south and vice versa (the southern part of the seeding experiments is actually from a geographical point of view the center of Israel. In reference to the north it was called “south” in the seeding experiment). The seeding line in Israel-1 was over the sea parallel to the coastline, from abeam Ashqelon to abeam Natanya (see Fig. 1). In Israel-2 the seeding line was shifted to the coastline and extended southward to the full length of the Gaza Strip. Israel-3 was conducted between 1975 and 1995 as a randomized experiment in the south only. The seeding line at the southern half was shifted eastward to the line of Ashqelon–Beer Sheba. This left S7 as a control area (see Fig. 1).
The statistical analysis for the southern seeding area in Israel indicated that seeding did not enhance the rainfall in the southern target area during Israel-2 (Gabriel and Rosenfeld 1990). The positive effect in Israel-1 for the combined north and south target areas appeared to be contributed from the north. When the south was considered alone, it did not show enhancement (Rosenfeld and Nirel 1996). The formal analysis of the Israel-3 experiment resulted in no discernable change of the precipitation as a result of seeding (Rosenfeld 1998). Understanding the reasons for the north–south differences of the seeding effects was a prime objective of several studies. Rosenfeld and Farbstein (1992) postulated that desert dust advected from the North African, Sinai, and Negev deserts was responsible for the seeding ineffectiveness in the south, by naturally “seeding” the clouds there by ice nuclei. They have shown that the seeding effect of Israel-2 in the north was almost fully obtained during one-half of the days, on which no dust was observed in the synoptic observations in Israel. Herut et al. (2000) have shown that rainfall in the south was basic, on average, and rainfall in the north was mostly slightly acidic. The basic nature of the rain in the south was caused by its high alkaline content. Geochemical analysis pointed to the Sahara, Sinai, and Negev deserts as the source area of the high alkalinity of the rain. Levi and Rosenfeld (1996) have shown that the content of desert dust in the rain was an order of magnitude larger in the south than in the north. The aerosols brought with southwesterly winds were rich in desert dust, which acted as ice nuclei at relatively high temperature. The south is an area of large gradient of isohyets, which is the climatological manifestation of being affected mainly by the southern margins of the rain-cloud systems. The hypothesis that desert dust affected the southern margins of the rain cloud system but washed down farther north was tested by Rosenfeld and Nirel (1996), who analyzed the seeding effect in the north, stratified by the southward extent of the rain-cloud system. They found that rainfall was enhanced in the north only in situations in which it extended to the southern part of the southern area. That result had a higher significance level than any other statistical analysis throughout the Israeli experiments (Rosenfeld and Nirel 1996).
Air pollution in Israel
Air pollution arrives in Israel on rain days with airflow from eastern Europe through the eastern Mediterranean Sea. The flow curves cyclonically and typically arrives in Israel at low levels as a southwesterly wind that picks up considerable additional air pollution from the Israeli coastal plain. A major pollution source in northern Israel is Haifa Bay, a densely populated area (over 1.5 million people) that contains power plants, a refinery, and much industry (Israel Ministry of Environment 2003). The pollution source in central Israel is the Tel Aviv metropolitan area, which contains industry and a dense population (over 2 million people).
In Israel, as in other places in the world, rapid technological development, improvement in the standard of living, and increased population density have brought in their wake pollutant emissions from both stationary and mobile sources. Vehicle density has risen from 34 cars in 1954 to over 230 today per 1000 people, with the number of cars reaching 2 million. Diesel consumption increased from 692 × 103 tons in 1970 to 2565 × 103 tons in 2003 (Israeli Bureau of Statistics 2004). Those emissions [very small (<0.1 μm) aerosol particles] are as efficient as biomass smoke particles at acting as CCN (Lammel and Novakov 1995). A diesel car produces several orders of magnitude more such particles per mile than a gasoline car with the same fuel consumption (Pierson and Brachaczek 1983; Williams et al. 1989; Lowenthal et al. 1994; Weingartner et al. 1997; Maricq. et al. 1999). Furthermore, the total petroleum products in Israel increased from 2764 × 103 tons in 1970 to 8944 × 103 tons in 2003 (Israeli Bureau of Statistics 2004), and the emissions of all pollutants have increased since 1980 (Israel Ministry of Foreign Affairs 1998). The increase was primarily due to the increasing reliance on diesel fuels for trucks and commercial vehicles (Fletcher et al. 1999). For example, levels of particulate matter of less than 10 μm in diameter in Tel Aviv and Jerusalem are on par with, or may even exceed, the average in the Los Angeles, California, region (Fletcher et al. 1999).
Method
Overall Ro trends
To separate cloud-seeding and air-pollution effects, this study focused mainly on the control and target seeding areas in northern Israel, because this is the location for which cloud seeding was reported to have increased the rainfall. The first step is to quantify the combined effects of air pollution and seeding without attempting to separate them.
As was done by Givati and Rosenfeld (2004) in California and in central Israel (in the Judea and Samaria hills, the southern part of the seeding area), highly correlated pairs of hills–coast rain gauges were selected to achieve a good representation of the control and the target areas in the north. The stations in each cluster were selected according to the principles of topographical similarity (meaning that the stations in each cluster should be around the same elevation) and equal precipitation measuring periods (see locations in Fig. 2). The main response parameter that is used is the annual time series of Ro. Northern Israel was divided into six target subareas and three control subareas (see Fig. 1) as was done in the cloud-seeding experiments (Gagin and Neumann 1974, 1981).
To evaluate the trends in Ro, the ratio of precipitation between clusters of hilly rain gauges and clusters of upwind lowland rain gauges was measured. Therefore, the clusters of stations used here in the south are the same clusters that were used to evaluate the trends in Ro in Givati and Rosenfeld (2004), but they do not mach exactly the clusters that were used in the formal evaluation of the cloud-seeding experiments (see the cluster and stations detail in Table A9 in the appendix) The results are presented in section 4a.
Separation between seeded and unseeded trends
As was mentioned before, the seeding line was to the west of the coastline during Israel-1 in both north and south, on the coastline during Israel-2 and Israel-3 for the south, and east of the coastline during Israel-2 and the operational seeding for the north. According to this arrangement, the north coastal plain stations could have been under the influence of seeding during only the 6 years of Israel-1. When we exclude this period, the trends for the seeded and unseeded conditions did not change. However, the following considerations suggest that it is not necessary to omit the years in which the coastal stations were in the seeded area:
The coastal and plain stations are very close to the seeding line, so that the seeding material there is less likely to spread and affect as many clouds as farther downwind, and the just-seeded clouds have less time to respond to the seeding than farther downwind.
Givati and Rosenfeld (2004) showed that the orographic clouds were the most sensitive to the precipitation suppression effects of the air pollution, as quantified by the trends in Ro. They showed that no change was evident in the lowland stations both in Israel and California [e.g., no change was evident between the lowland stations of San Bernadino and Los Angeles, Sacramento and San Francisco (in California), and pairs of lowland stations in coastal and the inland plains of central Israel]. We invoke here our suggestion that rain enhancement by cloud seeding is the other side of the same coin of the sensitivity of the precipitation efficiency of orographic clouds to aerosols. The lack of the apparent pollution effects in the lowlands suggests that the precipitation there occurs from clouds that are not as sensitive to the aerosol source over land, because they were originated by convection or synoptic forcing over sea and are relatively mature when advected from the sea inland.
Support to the last point is obtained from the results of the analyses, presented later in this study, that show little indicated seeding effects in the low-elevation subtargets inland and substantial seeding effects in the hilly areas, suggesting that the orographic clouds are those that mainly responded positively to the seeding in northern Israel.
Using as controls those stations that are “contaminated” by seeding that produces a seeding effect of the same sign but smaller than in the target area would cause the evaluation to detect the seeding effect as being smaller than the real one. Therefore, if an effect is still detected with formal statistical significance, it can be interpreted as an underestimate of the real effect.
The years 1950–60 were the preexperimental years and so were considered as the nonseeded period. The years since 1975 are the operational period and so were considered as seeded. In the south, experimental randomized seeding continued until 1995. During the period of 1961–75 (Israel-1 and -2 experiments), some of the days were seeded and some were unseeded, according to a random allocation that was inherent to the experiment. The separate time series contain annual precipitation amounts for the seeded and unseeded days separately (from November to April of each year).
Results
Overall Ro trends
Figure 3 displays the time series of Ro (the ratio of target–control annual rainfall) for the hilly clusters of N1 (Fig. 3a) and N3 (Fig. 3b) against the plains cluster of C2 (see locations in Fig. 2 and clusters details in Table A1 in the appendix). The control area of C2 has the highest daily and annual correlation with the clusters of N1 and N3, and thus this specific cluster was used for measuring the hill/plains ratio for the north (see correlations in Table A7 in the appendix). The high correlation is required for using the coastal station to predict the “natural” rainfall in the hilly station. The high correlation is also essential for assuring that the mountain stations that were selected are indeed downwind of the coastal stations.
The hilly clusters are located downwind of the pollution sources in the bay of Haifa. It can be seen in Fig. 3 (and also in Table A1 of the appendix) that a decreasing trend of 10%–15% in Ro occurred along the years in those target areas. The decrease is statistically significant. This decrease in Ro is similar to those found in hilly areas in central Israel and in California (Givati and Rosenfeld 2004). In contrast, little change was found in the target/control ratio for the low-elevation clusters of stations in the areas of N2, N6 (the area of N6 has topographic diversity, and so we divided it into two areas: N6a and N6b), and N8. This is in line with the lack of orographic enhancement factor between these subareas and the upwind coastal control clusters (see locations in Fig. 2 and the station details and results in Table A1 of the appendix) and is evident in Fig. 4, in which the absence in the orographic enhancement is associated with lack of significant trend in Ro in those areas along the years.
An even stronger decrease was already found by Givati and Rosenfeld (2004) in the southern seeded area in the Judea hills in area S3 (Fig. 1) as shown in Fig. 5, which is reproduced in this paper. Both the northern (the upper Galilee hills) and the southern mountainous seeded areas (the Judea hills) show an overall decreasing trend.
Trend analysis under seeded and unseeded conditions—Northern Israel
Figure 6 shows the trend lines under seeded (the broken line and empty points) and unseeded (the solid line and full points) conditions for the ratio between the clusters of stations in N1 (Fig. 6a) and N3 (Fig. 6b) against the coastal cluster of C2. Both the seeded and the unseeded trend lines in the two clusters decreased along the years, but the broken line that represents the ratio under seeded conditions is shifted upward by 12%–14% with respect to the solid line that represents the unseeded condition. (More detail is presented in Table A6 of the appendix.) During the randomized seeding period of 1961–75 the target/control ratio of the seeded days was greater by 12.4% than for the unseeded days in N1 and by 14.4% in N3 (the calculation for the distance between the lines was made according to the average value of the periods 1961–74). Adding the years before 1961 to the unseeded regression and the years after 1975 to the seeded regression did not change the regression lines much. The regressions shown in Fig. 6 are for these full seeded and unseeded datasets. Implementation of the Student’s t test (statistical test for difference between groups) shows that the differences between the seeded and unseeded regression trend lines are significant at the levels of 6% for N1 and 3% for N3. This result means that without cloud seeding the precipitation amount in those two target areas might have been 12%–14% less than what they are today, or a loss of about 100 mm yr−1 in the hilly areas, with annual precipitation near 800 mm. The increase in the orographic enhancement factor from cloud seeding does not fully compensate for the continuing decrease from air pollution. Cloud seeding for rain enhancement is aimed at accelerating the conversion of cloud droplets into precipitation particles, whereas air pollution has the opposite effect of suppressing the coalescence of cloud droplets into raindrops and the formation of ice hydrometeors. Rosenfeld (2000) showed the pollution suppression effect on ice formation in places such as Australia.
As can be seen in Fig. 6, the ratio for the seeded conditions at the end of the measured period is still lower by 14%–15% than the unseeded ratio at the beginning of the measurement period. This result is in agreement with the overall ratio for all days without partitioning by seeding, as shown in Fig. 3.
Cloud seeding can fundamentally increase precipitation amount above the natural values, as was actually done in the early randomized cloud-seeding experiments. After all, this goal was the original purpose of these experiments. If we assume that the natural conditions have not changed during the research period, as suggested by the radiosonde analysis (Givati and Rosenfeld 2004), a decrease of a given amount in precipitation efficiency from air pollution implies a same percentage increase in the rain enhancement potential by introduction of aerosols that have the countereffect. In such case, there is at least a potential of an additional 14% for rain enhancement to restore to the values of the natural ratio that existed in the early 1950s. This finding further implies that cloud seeding as practiced in Israel is far from being optimal, by targeting and/or microphysical effects.
Seeding as done in Israel is likely to target only a fraction of the clouds for which it is intended, whereas air pollution is likely to affect a greater fraction, or even all clouds if it is long-range transported pollution from Europe. This consideration alone can explain why air pollution would dominate the overall combined effect. Next, we have to consider what would be the combined effect of pollution and seeding on the same cloud. Air pollution and large concentrations of small CCN have been already observed (Rosenfeld 2000) and simulated (Khain et al. 2001) to suppress the formation of ice hydrometeors in clouds. Air pollution may contain also additional ice nuclei, but if the precipitation is reduced by the air pollution it means that the precipitation suppression effect resulting from the increased concentrations of small CCN dominated the possible precipitation enhancement effect resulting from the increased concentrations of the ice nuclei. Therefore, the polluted clouds become more “seedable”; that is, the same cloud-seeding operations may result in a greater rain enhancement in more polluted clouds. Whether the net effect would be positive or still negative remains an open question here.
Figure 7 displays the time series of Ro for the plains and valley clusters of N2, N6, and N8 against the control areas of C2 and C4. No significant trend in the Ro under seeded and unseeded conditions was found between the low-elevation target clusters and the control areas of C2 and C4. The ratio was stable along the years since the 1950s until today (more detail is presented in Table A6 of the appendix). One exception might be a statistically insignificant positive seeding effect in N6 south, which is located just downhill of the highest mountains of the upper Galilee. The seeding effect might still persist there, but it is not associated with a decreasing trend of the unseeded conditions. It is apparent that the effects of cloud seeding and air pollution in the low-elevation areas are more limited because of the lack of an orographic enhancement to the rainfall. This result strengthens the suggestion that orographic clouds are the most susceptible to precipitation enhancement from cloud seeding and to precipitation suppression from air pollution.
Trend analysis under seeded and unseeded conditions—Southern Israel
A similar analysis of the trends of Ro during the seeded and unseeded conditions was done for the Judean hills, where a strong overall decreasing trend was indicated. Figure 8 shows the trend lines for the ratio under seeded and unseeded conditions in the southern seeded area of Israel between hilly clusters of stations of the Judea hills (S3 in Fig. 2b) against the Judea plains (S2 in Fig. 2b) and versus a cluster at the coastal plain farther upwind (S1 in Fig. 2b).
As was found in Givati and Rosenfeld (2004), the Ro between the hilly cluster to the plains and the coast is decreasing, but it can be seen in Fig. 8 that, unlike what is shown for the north in Fig. 6, here there are almost no differences between the trend lines that represent the seeded and unseeded conditions. According to Fig. 8c, no trend was found for the ratio between the Judea plains to the coastal plain under seeded and unseeded conditions, and also no difference was found between the lines (additional detail is presented in Table A8 of the appendix). These results suggest that the seeding did not appear to have affected the clouds either over the plain or over the hills downwind the seeding line. Those findings are consistent with the previous statistical analyses for Israel-1 (Rosenfeld 1997), Israel-2 (Gabriel and Rosenfeld 1990), and Israel-3 (Rosenfeld 1998) that showed no rainfall enhancement in the south target area, probably because of the effects of desert dust already seeding naturally these clouds.
This observation suggests that the air-pollution suppression effect on orographic precipitation has a dominant role over the ice nuclei positive effects. This is consistent with the observation of Borys et al. (2003) that the small CCN suppress precipitation mainly by slowing down the riming of the ice crystals that fall through the supercooled orographic cloud.
Conclusions
The results presented in this paper show for the first time the opposite effects of deliberate and inadvertent human actions to alter precipitation processes. The effect of growing urban and industrial air pollution since the 1950s has caused an inadvertent decrease in the orographic enhancement factor between the hills and the upwind plains (as was found in Israel and California). Without the air pollution, no trend would have probably occurred in the ratio, as was found to be the case in relatively pollution-free areas in Israel and California (Givati and Rosenfeld 2004). Cloud seeding with silver iodide was found to enhance precipitation, especially where the orographic enhancement factor was the largest. Likewise, the pollution effects reduced precipitation by the greatest amount in the same regions. Shallow and short-living orographic clouds are particularly susceptible to such impacts. The orographic clouds respond in opposite ways to the different kinds of aerosols that we inject into them. This sensitivity suggests that attempts to alter winter precipitation should be concentrated on orographic clouds.
This result also suggests that the conceptual model on which the Israeli cloud-seeding experiments was based is not exactly as postulated. The seeding was originally aimed at the convective clouds that formed over sea and the coastal plain, with the intent of nucleating ice crystals and forming graupel earlier in the cloud life cycle (Gagin and Neumann 1974). It appears that cloud seeding did not enhance the convective precipitation, however, but rather increased the orographic precipitation, probably by the Bergeron–Findeisen process (Bergeron 1935).
The lack of enhancement of the convective clouds in Israel might be explained by their tendency to mature and dissipate inland during the winter storms. Seeding of mature convective clouds cannot affect them much. The lack of enhancement is also consistent with the microphysically maritime nature of the convective clouds. It appears to be caused mainly by the natural hygroscopic seeding by sea spray in the winter storms that enhance the warm precipitation (Rosenfeld et al. 2002) as well as promoting the formation of ice hydrometeors followed by ice multiplication (Hallett and Mossopp 1974). These suggestions are supported by the results of glaciogenic cloud seeding in Tasmania that targeted a hilly area by seeding along an upwind coastline. The seeding in Tasmania was shown to enhance precipitation from the stratiform orographic clouds but not from the convective clouds (Ryan and King 1997).
This conclusion potentially resolves the microphysical questions put forth by Rangno and Hobbs (1995), who asserted that cloud seeding as done in Israel could not have possibly caused the statistically documented rain enhancement from the convective clouds there.
The meaning of the results that are presented here is that statistical evaluations of nonrandomized seeding efficacy on orographic precipitation without taking into account the pollution effects will likely lead to erroneous results and misleading conclusions. This fact can explain why such models that estimated the seeding effects in northern Israel based on historical comparisons (Nirel and Rosenfeld 1995) showed decreases in the apparent seeding effect along the years.
The case that was analyzed here is not unique. Many places in the world, such as California, are likely influenced by those opposite effects. The effect of air pollution on orographic precipitation was documented and quantified also in California, but it was not separated from the possibly positive effect of decades of glaciogenic cloud seeding of the orographic clouds there.
Acknowledgments
This study was partially funded by the Israeli Water Commission. The authors thank Dr. W. L. Woodley for valuable discussions and for reviewing the manuscript.
REFERENCES
Andreae, M. O., D. P. Rosenfield, A. A. Costa, G. P. Frank, K. M. Longo, and M. A. F. Silva-Dias. 2004. Smoking rain clouds over the Amazon. Science 303:1337–1342.
Bergeron, H. J. R. 1935. On the physics of clouds and precipitation. Proc. Fifth Assembly UGGI, Vol. 2, Lisbon, Portugal, UGGI, 156–160.
Borys, R. D., D. H. Lowenthal, S. A. Cohn, and W. O. J. Brown. 2003. Mountain and radar measurements of anthropogenic aerosol effects on snow growth and snowfall rate. Geophys. Res. Lett. 30.1538, doi:10.1029/2002GL016855.
Elliott, D. R. 1986. Review of wintertime orographic cloud seeding. Precipitation Enhancement—A Scientific Challenge, Meteor. Monogr., No. 43, Amer. Meteor. Soc., 87–101.
Fletcher, E., G. Ginsberg, and Y. Garb. 1999. Road transport, environment and equity in Israel. ADVA Executive Summary, 14 pp. [Available from ADVA Center, P.O.B. 36529, Tel Aviv, Israel, 61364, or online at http://www.adva.org/transport.html.].
Gabriel, K. R. 1967. The Israeli artificial rainfall simulation experiment. Statistical evaluation for the period 1961–1965. Weather Modification Experiments, Vol. 5, Proc. Fifth Berkeley Symp. on Mathematical Statistics and Probability, L. M. Lecam and J. Newman, Eds., University of California Press, 91–113.
Gabriel, R. K. and D. Rosenfeld. 1990. The second Israeli Rainfall Stimulation Experiment: Analysis of precipitation on both targets. J. Appl. Meteor. 29:1055–1067.
Gagin, A. 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.
Gagin, A. and J. Neumann. 1981. The second Israeli randomized cloud seeding experiment: Evaluation of the results. J. Appl. Meteor. 20:1301–1321.
Givati, A. and D. Rosenfeld. 2004. Quantifying precipitation suppression due to air pollution. J. Appl. Meteor. 43:1038–1056.
Goldreich, Y., H. Moses, and D. Rosenfeld. 2004. Radar analysis of cloud systems and their rainfall yield in Israel. Isr. J. Earth Sci. 53:63–76.
Hallett, J. and S. C. Mossopp. 1974. Production of secondary ice particles during the riming process. Nature 249:26–28.
Herut, B., A. Starinsky, A. Kats, and D. Rosenfeld. 2000. Relationship between the acidity and chemical composition of rainwater and climatological conditions along a transition zone between large deserts and Mediterranean climate, Israel. Atmos. Environ. 34:1281–1292.
Israeli Bureau of Statistics 2004. Transport, communication and energy budget. Report on Transport Communication.
Israel Ministry of Environment 2003. The environment in Israel districts—Report for the year of 2000. The Jerusalem Institute for Israel Studies Rep., 229 pp.
Israel Ministry of Foreign Affairs 1998. Israel’s Air Resources Management Program. MFA Newsletter, 4 Aug 1998, 229 pp. [Available online at http://www.mfa.gov.il/mfa/facts%20about%20israel/health%20-%20social%20services/israel-s%20air%20resources%20management%20program.].
Khain, A. P., D. Rosenfeld, and A. Pokrovsky. 2001. Simulating convective clouds with sustained supercooled liquid water down to −37.5°C using a spectral microphysics model. Geophys. Res. Lett. 28:3887–3890.
Lahav, R. and D. Rosenfeld. 2000. Microphysics characterization of the Israeli clouds from aircraft and satellites. Proc. 13th Int. Conf. on Clouds and Precipitation, Reno, NV, ICCP, 732–735.
Lammel, G. and T. Novakov. 1995. Water nucleation properties of carbon black and diesel soot particles. Atmos. Environ. 29:813–823.
Levi, Y. and D. Rosenfeld. 1996. On ice nuclei, rainwater chemical composition and static cloud seeding effects in Israel. J. Appl. Meteor. 35:1494–1501.
Lowenthal, D. H., B. Zielnska, J. C. Chow, J. G. Watson, M. Gautam, D. H. Ferguson, G. R. Neuroth, and K. D. Stevens. 1994. Characterization of heavy-duty diesel vehicle emissions. Atmos. Environ. 28:731–743.
Maricq, M. M., R. E. Chase, D. H. Podsiadlik, and R. Vogt. 1999. Vehicle exhaust particle size distributions: A comparison of tailpipe and dilution tunnel measurements. Society of Automotive Engineers Paper 1999-01-1461, 30 pp.
Nirel, R. and D. Rosenfeld. 1995. Estimation of the effect of operational seeding on rain amounts in Israel. J. Appl. Meteor. 34:2220–2229.
Phillips, V. T. J., T. W. Choularton, and A. M. Blyth. 2002. The influence of aerosol concentrations on the glaciation and precipitation of a cumulus cloud. Quart. J. Roy. Meteor. Soc. 128:951–971.
Pierson, W. R. and W. W. Brachaczek. 1983. Particulate matter associated with vehicles on the road II. Aerosol Sci. Technol. 2:1–40.
Rangno, A. L. and P. V. Hobbs. 1995. A new look at the Israeli cloud seeding experiments. J. Appl. Meteor. 34:1169–1193.
Rosenfeld, D. 1986. The dynamic characteristics of cumuliform clouds and cloud systems and their effect on the rainfall precipitated by them. Ph.D. thesis, The Hebrew University of Jerusalem, 142 pp.
Rosenfeld, D. 1997. Comment on “A new look at the Israeli cloud seeding experiment.”. J. Appl. Meteor. 36:260–271.
Rosenfeld, D. 1998. The third Israeli randomized cloud seeding experiment in the south: Evolution of the result and review of all three experiment. Preprints, Conf. on Cloud Physics and 14th Conf. on Planned and Inadvertent Weather Modification, Everett, WA, Amer. Meteor. Soc., 565–568.
Rosenfeld, D. 1999. TRMM observed first direct evidence of smoke from forest fires inhibiting rainfall. Geophys. Res. Lett. 26:3105–3108.
Rosenfeld, D. 2000. Suppression of rain and snow by urban air pollution. Science 287:1793–1796.
Rosenfeld, D. and H. Farbstein. 1992. Possible influence of desert dust on seedability of clouds in Israel. J. Appl. Meteor. 31:722–731.
Rosenfeld, D. and G. Gutman. 1994. Retrieving microphysical properties near the tops of potential rain clouds by multispectral analysis of AVHRR data. J. Atmos. Res. 34:259–283.
Rosenfeld, D. and R. Nirel. 1996. Seeding effectiveness—The interaction of the desert dust and the southern margins of rain cloud systems in Israel. J. Appl. Meteor. 35:1502–1510.
Rosenfeld, D. and I. M. Lensky. 1998. Satellite based insights into precipitation formation processes in continental and maritime convective clouds. Bull. Amer. Meteor. Soc. 79:2457–2476.
Rosenfeld, D. and W. L. Woodley. 2003. Closing the 50-year circle: From cloud seeding to space and back to climate change through precipitation physics. Cloud Systems, Hurricanes, and the Tropical Rainfall Measuring Mission (TRMM), Meteor. Monogr., No. 51, Amer. Meteor. Soc., 59–80.
Rosenfeld, D., R. Lahav, A. P. Khain, and M. Pinsky. 2002. The role of sea-spray in cleansing air pollution over ocean via cloud processes. Science 297:1667–1670.
Ryan, B. F. and W. D. King. 1997. A critical review of the Australian experience in cloud seeding. Bull. Amer. Meteor. Soc. 78:239–254.
Segal, Y., A. Khain, M. Pinsky, and D. Rosenfeld. 2004. Effects of hygroscopic seeding on raindrop formation as seen from simulations using a 2000 bin spectral cloud parcel model. Atmos. Res. 71:3–34.
Sharon, D. and H. Kutiel. 1986. The distribution of rainfall intensity in Israel, its regional and seasonal variations and its climatological evaluation. J. Climatol. 6:277–291.
Silverman, B. A. 2001. A critical assessment of glaciogenic seeding of convective clouds for rainfall enhancement. Bull. Amer. Meteor. Soc. 82:903–923.
Weingartner, E., H. Burtscher, and U. Baltensperger. 1997. Hygroscopic properties of carbon and diesel soot particles. Atmos. Environ. 31:2311–2327.
Williams, D. J., J. W. Milne, S. M. Quigley, and D. B. Roberts. 1989. Particulate-emissions from in-use motor vehicles. Atmos. Environ. 23:2647–2661.
Yair, Y. and Z. Levin. 1994. Lightning disintegration in clouds and into the ground in thunderstorms in Israel. Meteor. Isr. 3:20–28.
APPENDIX
Description of the Additional Tables
This appendix contains additional information about the graphs that are presented in this paper. The appendix includes nine tables (Tables A1 –A9). It follows the same structure and principles as in Givati and Rosenfeld (2004). The tables help to show the full picture of the time series that are presented. The tables contain any of the following information: station/cluster name—the rain gauge name or the cluster name; lat, lon—the geographical latitude and longitude (in decimal degrees); elev—the elevation of the rain gauges in meters above sea level; avg yearly precip—the average annual amount of precipitation at each station for the tested years (mm yr−1); E/S—the ratio (between the precipitation amounts at the hills and at the upwind lowlands) in the end of the time series divided by the ratio in the beginning gives the change, as calculated using the regression line, which is the trend along the years; P value—statistical significance of the trend corresponding to the Student’s t-test statistic, quantifying the chance that there is no trend; regression eqn—the regression equation that is calculated for the trend; E/S precip—same as E/S, but for individual rain gauge time series, not pairs of station ratios; cluster—groups of rain gauges that represent the average annual precipitation in a selected geographical area; seeded E/S—the ratio in the end of the time series divided by the ratio in the beginning under seeded conditions; unseeded E/S—the ratio in the end of the time series divided by the ratio in the beginning under unseeded conditions; pct diff between seeded and unseeded conditions—the difference in percentages between the annual ratio of precipitation under seeded and unseeded conditions; ratio between clusters—the annual ratio of precipitation between clusters of stations.
Map of Israel that shows the two experimental areas and subareas for Israel-1 (1961–69) and Israel-2 (1969–74), the operational seeding in the north (since 1975), and Israel-3 in the south (1975–95). The target and control areas are divided into subareas N1–N8 in the northern target, S1–S7 in the south target, and C1–C4 for the northern control area; S7 served as a control for Israel-3. The seeding line in Israel-2 north was shifted inland. The hatched area in the north represents the catchment of the Lake of Galilee.
Citation: Journal of Applied Meteorology 44, 9; 10.1175/JAM2276.1
Map of Israel that shows the two experimental areas and subareas for Israel-1 (1961–69) and Israel-2 (1969–74), the operational seeding in the north (since 1975), and Israel-3 in the south (1975–95). The target and control areas are divided into subareas N1–N8 in the northern target, S1–S7 in the south target, and C1–C4 for the northern control area; S7 served as a control for Israel-3. The seeding line in Israel-2 north was shifted inland. The hatched area in the north represents the catchment of the Lake of Galilee.
Citation: Journal of Applied Meteorology 44, 9; 10.1175/JAM2276.1
Map of Israel that shows the two experimental areas and subareas for Israel-1 (1961–69) and Israel-2 (1969–74), the operational seeding in the north (since 1975), and Israel-3 in the south (1975–95). The target and control areas are divided into subareas N1–N8 in the northern target, S1–S7 in the south target, and C1–C4 for the northern control area; S7 served as a control for Israel-3. The seeding line in Israel-2 north was shifted inland. The hatched area in the north represents the catchment of the Lake of Galilee.
Citation: Journal of Applied Meteorology 44, 9; 10.1175/JAM2276.1
Clusters of homogeneous rain gauges in the target areas (N1–N8 for the north and S3 for the south) and control areas (C2 in the north and S1–S2 in the south) in the (a) northern and (b) southern seeded areas in Israel. The gauge clusters were selected based on topographical similarity and equal precipitation measuring periods.
Citation: Journal of Applied Meteorology 44, 9; 10.1175/JAM2276.1
Clusters of homogeneous rain gauges in the target areas (N1–N8 for the north and S3 for the south) and control areas (C2 in the north and S1–S2 in the south) in the (a) northern and (b) southern seeded areas in Israel. The gauge clusters were selected based on topographical similarity and equal precipitation measuring periods.
Citation: Journal of Applied Meteorology 44, 9; 10.1175/JAM2276.1
Clusters of homogeneous rain gauges in the target areas (N1–N8 for the north and S3 for the south) and control areas (C2 in the north and S1–S2 in the south) in the (a) northern and (b) southern seeded areas in Israel. The gauge clusters were selected based on topographical similarity and equal precipitation measuring periods.
Citation: Journal of Applied Meteorology 44, 9; 10.1175/JAM2276.1
Recent trends in the ratio of annual precipitation (Ro) during 1950–2001 for (a) a cluster of rain gauges at elevations of 500–700 m in subarea N1 and for (b) a cluster at elevations of 700–950 m in subarea N3 in northern Israel. Note the statistically significant decrease of 10%–15% in the ratios with respect to the C2 control area. Station details are in the appendix.
Citation: Journal of Applied Meteorology 44, 9; 10.1175/JAM2276.1
Recent trends in the ratio of annual precipitation (Ro) during 1950–2001 for (a) a cluster of rain gauges at elevations of 500–700 m in subarea N1 and for (b) a cluster at elevations of 700–950 m in subarea N3 in northern Israel. Note the statistically significant decrease of 10%–15% in the ratios with respect to the C2 control area. Station details are in the appendix.
Citation: Journal of Applied Meteorology 44, 9; 10.1175/JAM2276.1
Recent trends in the ratio of annual precipitation (Ro) during 1950–2001 for (a) a cluster of rain gauges at elevations of 500–700 m in subarea N1 and for (b) a cluster at elevations of 700–950 m in subarea N3 in northern Israel. Note the statistically significant decrease of 10%–15% in the ratios with respect to the C2 control area. Station details are in the appendix.
Citation: Journal of Applied Meteorology 44, 9; 10.1175/JAM2276.1
Lack of orographic enhancement factor and respectively no significant trend in the Ro along the years between the low-elevation clusters in the targets areas of (a) N8 (−180 m), (b) N2 (27 m), (c) N6a (240 m), and (d) N6b (230 m) with respect to the control areas of C2 and C4.
Citation: Journal of Applied Meteorology 44, 9; 10.1175/JAM2276.1
Lack of orographic enhancement factor and respectively no significant trend in the Ro along the years between the low-elevation clusters in the targets areas of (a) N8 (−180 m), (b) N2 (27 m), (c) N6a (240 m), and (d) N6b (230 m) with respect to the control areas of C2 and C4.
Citation: Journal of Applied Meteorology 44, 9; 10.1175/JAM2276.1
Lack of orographic enhancement factor and respectively no significant trend in the Ro along the years between the low-elevation clusters in the targets areas of (a) N8 (−180 m), (b) N2 (27 m), (c) N6a (240 m), and (d) N6b (230 m) with respect to the control areas of C2 and C4.
Citation: Journal of Applied Meteorology 44, 9; 10.1175/JAM2276.1
Trend in the annual ratio of precipitation (Ro) between clusters of rain gauges in the Judean hills (S3) vs the upwind Judean Plains (S2). The small P values show that the decreasing trend is statistically highly significant. Station details are in the appendix.
Citation: Journal of Applied Meteorology 44, 9; 10.1175/JAM2276.1
Trend in the annual ratio of precipitation (Ro) between clusters of rain gauges in the Judean hills (S3) vs the upwind Judean Plains (S2). The small P values show that the decreasing trend is statistically highly significant. Station details are in the appendix.
Citation: Journal of Applied Meteorology 44, 9; 10.1175/JAM2276.1
Trend in the annual ratio of precipitation (Ro) between clusters of rain gauges in the Judean hills (S3) vs the upwind Judean Plains (S2). The small P values show that the decreasing trend is statistically highly significant. Station details are in the appendix.
Citation: Journal of Applied Meteorology 44, 9; 10.1175/JAM2276.1
Change in the target/control ratio of annual precipitation during 1950–2001 for the seeded (broken line and open circles) and unseeded (solid line and filled circles) conditions between stations in the target areas of (a) N1 (western upper Galilee, 500–700 m) and (b) N3 [highest area of the upper Galilee to the east of (a); 700–950 m] to the control area C2 (northern coast). The seeded trend line is shifted upward with respect to the unseeded line by 12.4% in N1 and by 14.2% in N3.
Citation: Journal of Applied Meteorology 44, 9; 10.1175/JAM2276.1
Change in the target/control ratio of annual precipitation during 1950–2001 for the seeded (broken line and open circles) and unseeded (solid line and filled circles) conditions between stations in the target areas of (a) N1 (western upper Galilee, 500–700 m) and (b) N3 [highest area of the upper Galilee to the east of (a); 700–950 m] to the control area C2 (northern coast). The seeded trend line is shifted upward with respect to the unseeded line by 12.4% in N1 and by 14.2% in N3.
Citation: Journal of Applied Meteorology 44, 9; 10.1175/JAM2276.1
Change in the target/control ratio of annual precipitation during 1950–2001 for the seeded (broken line and open circles) and unseeded (solid line and filled circles) conditions between stations in the target areas of (a) N1 (western upper Galilee, 500–700 m) and (b) N3 [highest area of the upper Galilee to the east of (a); 700–950 m] to the control area C2 (northern coast). The seeded trend line is shifted upward with respect to the unseeded line by 12.4% in N1 and by 14.2% in N3.
Citation: Journal of Applied Meteorology 44, 9; 10.1175/JAM2276.1
Same as Fig. 6, but for the (a) southern and (b) northern parts of N6 (the Jordan valley to the east of the upper Galilee, averaging 235 m above sea level), with C2 as a control; for (c) the region of N2, the lower Galilee (N2, with C2 as its control); and for (d) the eastern shores of Lake of Galilee (N8, with C4 as its control). The low elevations of N2 (110–200 m above sea level) and N8 (130–200 m below sea level) result in none to negative orographic enhancement factor and in no discernible trends and differences in them with time or with seeding. Some positive seeding effect is noted at N6 south, but it does not reach significance, and no trend is discerned in the unseeded series. No effect or trend is noted for N6 north.
Citation: Journal of Applied Meteorology 44, 9; 10.1175/JAM2276.1
Same as Fig. 6, but for the (a) southern and (b) northern parts of N6 (the Jordan valley to the east of the upper Galilee, averaging 235 m above sea level), with C2 as a control; for (c) the region of N2, the lower Galilee (N2, with C2 as its control); and for (d) the eastern shores of Lake of Galilee (N8, with C4 as its control). The low elevations of N2 (110–200 m above sea level) and N8 (130–200 m below sea level) result in none to negative orographic enhancement factor and in no discernible trends and differences in them with time or with seeding. Some positive seeding effect is noted at N6 south, but it does not reach significance, and no trend is discerned in the unseeded series. No effect or trend is noted for N6 north.
Citation: Journal of Applied Meteorology 44, 9; 10.1175/JAM2276.1
Same as Fig. 6, but for the (a) southern and (b) northern parts of N6 (the Jordan valley to the east of the upper Galilee, averaging 235 m above sea level), with C2 as a control; for (c) the region of N2, the lower Galilee (N2, with C2 as its control); and for (d) the eastern shores of Lake of Galilee (N8, with C4 as its control). The low elevations of N2 (110–200 m above sea level) and N8 (130–200 m below sea level) result in none to negative orographic enhancement factor and in no discernible trends and differences in them with time or with seeding. Some positive seeding effect is noted at N6 south, but it does not reach significance, and no trend is discerned in the unseeded series. No effect or trend is noted for N6 north.
Citation: Journal of Applied Meteorology 44, 9; 10.1175/JAM2276.1
Trend lines under seeded and unseeded conditions for the ratio between the cluster of rain gauges over the hills of Jerusalem, 600–800 m, (S3) vs (a) the upwind plain just to the west of the hills, 100–200 m, (S2) and vs (b) a rain gauge cluster at the coast (S1). It is evident that (c) the similarity between (a) and (b) suggests that no relative trends occurred between S2 and S1. It can be seen that although the Ro of the hill to the plain and coast is decreasing, unlike in northern Israel (Fig. 6) here there is no difference between the lines that represent seeded conditions (broken line and open circles) and unseeded conditions (solid line and filled circles). It suggests that cloud seeding did not enhance precipitation in the hills over central Israel.
Citation: Journal of Applied Meteorology 44, 9; 10.1175/JAM2276.1
Trend lines under seeded and unseeded conditions for the ratio between the cluster of rain gauges over the hills of Jerusalem, 600–800 m, (S3) vs (a) the upwind plain just to the west of the hills, 100–200 m, (S2) and vs (b) a rain gauge cluster at the coast (S1). It is evident that (c) the similarity between (a) and (b) suggests that no relative trends occurred between S2 and S1. It can be seen that although the Ro of the hill to the plain and coast is decreasing, unlike in northern Israel (Fig. 6) here there is no difference between the lines that represent seeded conditions (broken line and open circles) and unseeded conditions (solid line and filled circles). It suggests that cloud seeding did not enhance precipitation in the hills over central Israel.
Citation: Journal of Applied Meteorology 44, 9; 10.1175/JAM2276.1
Trend lines under seeded and unseeded conditions for the ratio between the cluster of rain gauges over the hills of Jerusalem, 600–800 m, (S3) vs (a) the upwind plain just to the west of the hills, 100–200 m, (S2) and vs (b) a rain gauge cluster at the coast (S1). It is evident that (c) the similarity between (a) and (b) suggests that no relative trends occurred between S2 and S1. It can be seen that although the Ro of the hill to the plain and coast is decreasing, unlike in northern Israel (Fig. 6) here there is no difference between the lines that represent seeded conditions (broken line and open circles) and unseeded conditions (solid line and filled circles). It suggests that cloud seeding did not enhance precipitation in the hills over central Israel.
Citation: Journal of Applied Meteorology 44, 9; 10.1175/JAM2276.1
Table A1. Trend analysis of annual precipitation between the target areas and the control of C2 and C4 in northern Israel.
Table A2. Station information for clusters C1, C2, and C4.
Table A3. Trend analysis of annual precipitation of the target areas to the average value of all the control areas (C1, C2, and C4) in northern Israel.
Table A4. Station information for clusters N1, N2, N3, N6, and N8.
Table A5. Trend analysis between target and control areas (C2 and C4) under seeded and unseeded conditions in northern Israel.
Table A6. Trend analysis under seeded and unseeded conditions of the target areas to average value of all the control areas (C1, C2, and C4) in northern Israel.
Table A7. Annual and daily correlations for precipitation between the control and target areas in northern Israel.
Table A8. Trend analysis under seeded and unseeded conditions for the hill (S3*), plain (S2), and coast (S1) areas in the southern seeded area in southern Israel.
Table A9. List of stations in the southern coast (S1), southern plains seeded area (Judea plains; S2), and hilly southern seeded area (Judea hills; S3).
