CC BY
1
УДК 551.576
Gekkieva S.
Federal State Budgetary Institution «High-mountain Geophysical Institute»,
Nalchik, Russia
EVALUATION OF THE EFFECTIVENESS OF AN ARTIFICIAL INCREASE IN PRECIPITATION ON THE EXAMPLE OF THE STAVROPOL TERRITORY
Abstract
With the development of agro-industrial complex, industry and public utilities, with the growth of cities and population, as well as a significant decrease in rainfall, the need for fresh water has increased significantly. The issue of replenishing natural water resources has become an urgent problem. One of the options for solving this problem was the construction of new reservoirs, but the climatic trend of decreasing precipitation significantly reduces the efficiency of these systems. In this regard, they cannot meet the demand for fresh water due to the small replenishment of water reserves due to precipitation. As a result of numerous studies in different countries, devices and technologies have been developed for seeding clouds with ice-forming nuclei in order to increase the amount of precipitation.
Keywords:
precipitation, artificial increase in precipitation, active influences, statistical method.
Currently, there are several dozen research and operational projects on artificial regulation of precipitation, carried out in various countries of the world - Australia, USA, Canada, Israel, Italy, India, China, which indicates a significant interest in the problem and its practical significance. But many projects do not include research phases, since such work is from the very beginning aimed at obtaining additional precipitation, so research work is extremely fragmentary. Among the many issues on active impacts that require additional study is the problem of a possible redistribution of liquid precipitation in the territories adjacent to the experimental areas. Some scientists believe that an increase or decrease in the amount of precipitation in one area will entail fluctuations in the amount of precipitation in areas located downwind of the impact areas in which cloud seeding was carried out [1]. All this necessitates a more thorough analysis. To study the issue of possible redistribution of liquid precipitation in the territories adjacent to the experimental territories, the Stavropol Territory was chosen. The Stavropol Territory is one of the most important agricultural regions of the Russian Federation. The leading branch of the region's economy is the agro-industrial complex, which provides more than 30% of the profit. The Stavropol Territory is located in the southwest of Russia and has a unique landscape structure, which combines the features of the Russian Plain and the Greater Caucasus. Recently, global warming has been noted, which has its own characteristics in different regions, and the Stavropol Territory is at risk, often finding itself in difficult weather and climatic conditions. The lack of moisture in the meter layer and severe frosts lead to extremely negative consequences for the industry. To regulate the acute problem, it was decided to start experimental work on the territory of the Stavropol Territory in order to artificially increase precipitation. In the present study, when choosing the adjacent territory, we proceeded from the structure of the test sites of the Stavropol Service. The territories located to the east of the area of active impacts were taken as the adjacent territory. The method of historical regression was used for statistical analysis of the redistribution of precipitation. It should be noted that when choosing this method, all the requirements that determine the possibility of its use were met [2-6]. The results of the research carried out in this direction are given below.
The paper presents a statistical analysis developed on the basis of the method of historical regression and
designed to evaluate the work on the artificial increase in precipitation over a large area [7-10]. The territories for evaluation were chosen so that they meet the basic requirements for their further comparison by the method of historical regression. The main requirements include the following: similar physical and geographical characteristics; dimensions close to it in terms of the area of influence, the density of the ground-based precipitation network close to it, the length of the series of precipitation observations approximately the same as it; stable in time correlation of precipitation between the compared territories for the longest possible number of years before the onset of active impacts. The following territories were taken as control territories: Mineralnye Vody, Kislovodsk, Karachaevsk. Adjacent territories: Nalchik, Prokhladny, Terek, Kamennomostskoye, Mozdok, Yuzhno-Sukhokumsk, Terekli-Mekteb, Kochubey. The choice of control stations and the assessment of the results of impacts are made separately for each of the territories. An assessment is also performed separately for each season of work. The presence of a high correlation in the amount of precipitation between the compared territories and the absence of the influence of active influences in the control territories allows the use of this method. The basic data on which the method is based are averaged precipitation data for the spring-summer season (May-August) from individual weather stations. Precipitation averaging makes it possible to reduce the coefficients of precipitation variation. The calculation period was chosen from 1970 to 1985, because Since 1986, in the territory of the Stavropol Territory, work has begun on active influences on convective clouds in order to artificially increase precipitation, in connection with which the natural regime of atmospheric precipitation was violated. An analysis of the temporal course of precipitation over the entire region shows that the amount of precipitation throughout the territory has quasi-periodic fluctuations with cycles from 2 to 6 years.
Before starting a statistical analysis of the effect of redistribution of precipitation, it is necessary to make sure that the length of the statistical data series is sufficient to obtain statistically significant results. To assess the sufficiency of data series to obtain statistically significant conclusions, a t-test is conducted, assuming that the samples simulating the general populations {Nsi}and {Nci} consist of normally distributed random variables with parameters Ns = MNsi, S2 = DNsi and Nc = MNci, S2 = DNci, respectively, where M is the mathematical expectation and D is the variance of the considered parameters. Using Student's t statistics to compare the averages of two independent normal populations, a one-sided confidence (with a confidence level of 1-a) interval (-TO, (Ns - Nc)k) is constructed for the value Ns - Nc. In this case, the critical value (Ns - Nc)k is related to the upper quantile ta (calculated using normal approximation) of the a level of Student's distribution (44) with (ns+nc-2) degrees of freedom by the relation:
Hence, for the effect of artificial increase in precipitation that interests us, we get a critical for T-test the effect of artificial increase in precipitation Ek(%):
amount of precipitation during the years of active influences and before the influences respectively; ns and nc -The number of years of active influences and the number of years of observation to active influences (i.e., a year has been adopted for the experimental unit). To conduct a t-test, it is necessary to compare the actual value of efficiency E with the critical Ek found by formula (2). If E > Ek, then the effectiveness of the artificial increase in precipitation is considered significant at the chosen significance level a=0.05, and the number of years before the artificial increase in precipitation and during the period of active impacts is sufficient for a statistically significant conclusion that the increase in precipitation is due to active impacts, and not natural variability. If E < Ek, then we can assume that active impacts did not lead to significant changes in precipitation, i.e. the average
(1)
where t2a - upper quantile level 2a distribution of the value|t|; Ss and Sc- The average deviation of the
statistical changes are comparable to the natural variability of precipitation and there is no evidence to suggest that the effect of increased precipitation is present. On the basis of statistical analysis, it was found that the length of the statistical data series (1970-1997) is sufficient to obtain statistically significant results.
To assess the effect of redistribution of the amount of precipitation, the relationship between the amount of precipitation in the control (x) and the adjacent territory (y) obtained using regression analysis was used:
y = 0,7x + 38, (3)
The coefficient of correlation between precipitation in the compared territories was obtained r = 0.8. To estimate the correlation coefficient г- of the general population, we use the formula:
1 — V2 1 — T2
r--<rr <r +—■=- , (4)
где n - объем выборки.
It is necessary to test the hypothesis about the significance of the sample correlation coefficient. That is, at a significance level of 0.05, we will test the null hypothesis that the general correlation coefficient is equal to zero Hq: rr=0.
As a criterion for testing the null hypothesis, a random variable is taken Tobs. = = 4,93.
Wn-2 1-r2
(5)
According to the significance level of 0.05 and the number of degrees of freedom k, we find the critical point of the two-sided critical region from the Student's distribution table tcr.:
tcr (0,05;16)=2,12.
Поскольку Tobs..>tcr, нулевую гипотезу отвергаем. Следовательно, выборочный коэффициент корреляции значимо отличается от нуля, т.е. осадки на прилегающей и контрольной территории коррелированы.
Since Tobs..>tcr, we reject the null hypothesis. Therefore, the sample correlation coefficient is significantly different from zero, i.e. precipitation in the adjacent and control areas are correlated. The error in determining the expected amount of precipitation in the adjacent territory according to the regression equation (1) is equal to the standard deviation from the regression equation:
s = stVl — r2, (6)
s =±11,4(mm).
Table 1 shows data on additional precipitation amounts calculated according to equation 3 and actual precipitation in the compared territories for each year of active interventions. An analysis of the results of seasonal impact efficiency for the Stavropol Territory as a whole (Table 1) showed that over 10 years of work seasons, 6.8 - 48.8 mm of additional water was received, which amounted to 11-19% of natural precipitation.
Table 1
Estimation of the seasonal effectiveness of impacts for the Stavropol Territory in 1986-1996
Season Actual rainfall, mm Precipitation estimate, mm Impact efficiency, mm Relative efficiency, %
1986-1987 75,67 65,2 10,5 12
1987-1988 84,8 74,3 12,5 11,4
1988-1989 1Q2 53,2 48,8 19
1989-199Q 89,7 86,43 3,27 10
199Q-1991 82,6 62,8 19,8 13
1991-1992 11Q 100,3 9,7 11
1992-1993 84,3 68,2 16,1 12
1993-1994 78,6 71,8 6,8 11
1994-1995 56,5 45,5 11 12,4
1995-1996 47,3 40,3 7 11,7
Also, to assess the significance of the impact effect in a single experiment (for one season), the three-delta method [8] is used, if | Ay | >o(y), then one can speak of the impact effect with a confidence probability of 68%; if | Ay | >2o(y) - the confidence level is 95%; if |Ay| >3o(y) - the confidence probability is almost 100% [8,9]. In the practice of work on the artificial increase in precipitation, a confidence level of 68% is considered satisfactory, i.e. condition |Ay|>o(y) [10].
Thus, when statistically evaluating the annual results of the work, we separately obtain that the amount of additional precipitation per season should be |Ay | >11.4 mm in order to talk about a significant effect of redistribution of precipitation with a confidence level of 68%. When statistically assessing the results of the work for the entire period, we find that the amount of additional precipitation of 5.7 mm is significant at a significance level of 0.5.
Conclusions
Total for the period (1986-1997) data were analyzed for six control points and eight points in the adjacent territory. Table 1 shows that when considering the totality of sown and control areas without stratification of experiments, a positive effect is observed. The results obtained on the basis of the historical regression method and the three-delta method confirm that the amount of additional precipitation during the active season should be Ay>11.4 mm to speak of a significant redistribution effect of precipitation with a confidence probability of 68%. The only exception is one active season (1988), when the year itself is wet enough, and where a relative efficiency of 1.7 exceeds expected rainfall. It is difficult to relate this fact to the redistribution of precipitation due to active impacts in neighboring territories.
References:
1. Koloskov B.P., Korneev V.P., Shchukin G.G., Methods and equipments of cloud, rain and fog modification. - St Petersburg: RSHU Publishers, 2012.123p.
2. Abshaev M.T., Klimovskaya L.N. On the influence of anti-hail work on the intensity and amount of precipitation // Proceedings of HMGI. Issue 33,1976. 54p.
3. Gmurman V.S. Theory of Probability and Mathematical Statistics. - M.: Higher school, 1972. 368p.
4. Mitropolsky A. K. Technique of statistical calculations. Moscow: Fizmatgiz, 1961. 97p.
5. Shipilov O. The use of statistical methods to assess the effectiveness of work to increase precipitation. Hydrometeorology. Obninsk: VNIIGMI-MCD, 1983. 28p.
6. Bartishvili I.T. et al. Assessment of the impact of anti-hail protection on the precipitation regime of the protected and control territories of southern and eastern Georgia / // Proceedings of the Main Geophysical Observatory. Issue 497, 1986.47p.
7. Bartishvili Ya.T., Vatyan M.R., Kapanadze N.I., Odikadze M.Ya. / Investigation of the effect of anti-hail protection on the precipitation regime of the Central part of South Georgia // In the book: Materials of the Al l -Union seminar on the physics of the formation of hail processes and active influences on them. - M.: Gidrometeoizdat, 1988.114p.
8. Abshaev M.T., Abshaev A.M. Analysis of options for increasing rainfall in the local territory. Proceedings of HMGI, Issue 100, 2017.76p.
9. Vatiashvili M.R. Changes in the precipitation regime in the Central Caucasus region during anti-hail protection // Science. Innovation. Technology. 2019. No.2. P.88-96.
10. Gekkieva S. Statistical estimation of the effect of precipitation redistribution using the historical regression method. To cite this article: IOP Conf. Ser.: Mater. Sci. Eng.1083 012096. 2021.
© Gekkieva S., 2023
