(PDF) Current Desalination of the Sea of Azov and Its Correlation with Long-Term Variations in Atmospheric Circulation

MAIK “Nauka

/Interperiodica”0690

INTRODUCTIONThe spatial and temporal structure of the Sea of

Azov oceanographic fields, its salinity

in particular, ismainly controlled by variations in the water balanceelements, especially variations in the river runoff

andwater exchange with the Black Sea, which are associ-ated with changes in atmospheric circulation (AC).

Atmospheric circulation is a principal climate-form-ing factor. It is most variable in time; its long-term vari-ations are periodic and are responsible for the relevantvariability of hydrometeorological and oceanographiccharacteristics. If reliable relationships in the atmo-sphere–hydrosphere–biosphere system are established,the above fact could be used for evaluating the long-period variations not only of the hydrometeorologicaland oceanographic conditions in the Sea of Azov but ofits bioresources as well.

MATERIALS AND METHODSThe characteristics of atmospheric circulation used

in this study are based on the typification of synopticprocesses developed by G.Ya. Vangengeim in the 1930s[2] and improved by A.A. Girs, K.V. Kondratovich [6–8],and other researchers. The data arrays on the mainforms of atmospheric circulation (western

, eastern

,and northern

) have been prepared by specialists ofthe Arctic and Antarctic Research Institute. This studyis based on the data on the annual and monthly fre-quency of the above forms of atmospheric circulationfor 1891–1972 [6, 7], as well as the relevant indices for1973–1998 partly published in [5, 9, 14]. In addition,data on the water balance components and salinity ofthe Azov and Black seas for 1923–1985 [5] and theobservation data of Goskomgidromet on the Don andKuban river runoff (at the gauge lines of Razdorskaya

and Krasnodar, respectively) for 1912–1999 were used.The data arrays on water salinity in the open part of theSea of Azov are based on materials of the seasonalexpedition surveys performed by the HydrologicalInvestigations Laboratory of the Azov Research Insti-tute of Fisheries (1960–1999), as well as on the relevantdata obtained by the State Oceanographic Institute andthe Azov–Black Sea Research Institute of Fisheries andOceanography during the first years after the Don Riverregulation (1952–1959) and in the period, which is con-ventionally taken as the period of natural river runoff(1923–1951).

Correlation between long-term variations in the fre-quency of the atmospheric circulation forms, water bal-ance elements, and the salinity of the Azov and Blackseas was chiefly studied with the help of genetic andstructural relationships between them. Statistical meth-ods were used: the correlation and regression analyses,normalized difference integral curves of the modulecoefficients, linear trend estimates, and others.

EFFECT OF ATMOSPHERIC CIRCULATION ON WATER BALANCE ELEMENTS

AND SALINITY

A specific feature of the Sea of Azov is a closedependence of its hydrological regime, particularlysalinity, on river runoff, the volume of which is mainlygoverned by the amount of atmospheric precipitation(AP) falling onto the rivers' catchment areas duringcold seasons. Investigations of such relationships per-formed by the author as early as in the 1970s haveshown that the Kuban River runoff increases with sum-mer air temperature increasing above the norm (inten-sification of thawing of snow and glaciers in the moun-tains), whereas the runoff of rivers in the eastern part of

Current Desalination of the Sea of Azov and Its Correlation with Long-Term Variations in Atmospheric Circulation

Yu. M. Gargopa

Azov Research Institute of Fisheries, ul. Beregovaya 21/2, Rostov-on-Don, 344007 Russia

Received January 23, 2001

Abstract

—Correlation is established between the long-term variations in the frequency of the atmospheric cir-culation forms, water balance elements, and the Sea of Azov water salinity. It is found that the river runoff intothe sea and the sea freshwater balance increase and the sea salinity decreases in the periods, when northern andwestern forms of atmospheric processes develop; in the periods with a greater frequency of the eastern type ofatmospheric macroprocesses, the situation is reverse. It is also found that the effect of atmospheric circulationon the sea salinity tends to strengthen, whereas the effect of the human-induced decrease in river runoff tendsto diminish. The current desalination of the Sea of Azov down to 10.5‰ is shown to be mainly due to the devel-opment of western and northern forms of atmospheric circulation in the cold season of a year during the last10–15 years.

INTERACTION BETWEEN CONTINENTAL WATER AND THE ENVIRONMENT

WATER RESOURCES

Vol. 29

No. 6

2002

CURRENT DESALINATION OF THE SEA OF AZOV AND ITS CORRELATION 691

the near-Azov area increases in the case of a relativelylow temperature in winter (retention of snow in thecatchment area and deep freezing of soil contribute tothe formation of a high spring flood). However, the cor-relation coefficients reflecting these relationships arenot high (

= 0.26 and –0.42, respectively) althoughclose to statistically significant level. The amounts ofprecipitation falling in different zones of the above riverbasins during the cold season (November throughMarch) and the period of precipitation accumulation(October through April) have the decisive effect on theriver runoff

(

= 0.70–0.96)

According to [12], themany-year (1880–1965) variations in the Don annualrunoff at the gauge line in Kalach-on-Don correlate mostclosely with similar variations not only in the amount ofwinter precipitation of the current year (

= 0.41) but alsoin the amount of precipitation from August throughOctober of the preceding year (

= 0.56).

Some equations developed on this basis are pre-sented in Table 1. Their reliability is assessed accordingto [1, 13].

The relationship between the annual values of theSea of Azov salinity and the river runoff from its basinin 1960–1986 was characterized by a relatively closenegative correlation (

= –0.50…–0.61) [3]. It was max-imal in the case when the independent variable was pre-sented by the total river runoff from June of the preced-ing year through May of the current year. The abovecorrelation weakened in the direction from northeast tosouthwest, i. e., from Taganrog Bay (

= –0.76…–0.94) tothe Arabatskaya Spit and Kerch Strait (

= –0.21…–0.33).This regularity reflects a diminishing freshening effectof the Don and Kuban river runoff and growing salina-zing effect of the inflowing Black Sea water. Because ofa rather small volume of Taganrog Bay (25.4 km

),which is commensurable with the annual river runoffvolume of the Don disgorging in the bay, the long-term

variations in the bay salinity correspond to similarvariations in the annual river runoff into the Sea ofAzov (

= –0.84 …–0.86). This correlation is the clos-est (

= –0.92…–0.94) for the mean annual values ofsalinity in the central part of Taganrog Bay. As for thewhole Sea of Azov, whose volume (323 km

) is 9–10 times as large as the total annual inflow of riverwater, the long-term variations in its salinity are to thegreatest degree correlated with the total river runoff forthe current year and 4–5 preceding years. In the seaareas adjoining Taganrog Bay and Arabatskaya Spit,the variations in salinity are most closely correlatedwith the total river runoff for 4–5 and 6–7 years, respec-tively.

The formation of salinity during the current phase ofthe Sea of Azov desalination (1987–1999) is on thewhole characterized by the above regularities. How-ever, the degree of correlation between the salinity in thesea and especially in Taganrog Bay with the river runoff isslightly lower (

= –0.82…–0.85 and –0.60…–0.67,respectively). This is likely to be first associated withchanges in the climate-forming processes in the atmo-sphere. They were favorable for strengthening the cor-relation between the salinity and other components ofwater balance and diminishing the effect of human-induced reduction of river runoff.

As a result, statistical models approximating therelationships described above were developed. Somemodels are presented in Table 2.

At the first stage of the study, the relationshipsbetween the forms of atmospheric circulation, waterbalance elements, and salinity of the Azov and Blackseas were analyzed with the use of materials [5] for theperiods of conventionally natural (1923–1951) and reg-ulated (1952–1982) river runoff.

Under natural conditions, an increase in river runoffcan be associated with a higher frequency of

and

Table 1.

Parameters

and

of linear equations approximating the dependence of the natural (reconstructed) runoff of theKuban River, km

, and the rate of streamflow in its tributaries, m

/s, on atmospheric precipitation in the cold season

and inthe period of its accumulation

, as well as characteristics of the equations reliability (here and in Tables 2 and 3,

is thecoefficient of correlation;

is the Fisher transformation testifying to the nonrandomness of

value, given that

≥

2.5;

isthe root-mean-square error of the equation;

is the root-mean-square deviation of the predicted value;

is the criterionof accuracy of the equation (if it is equal to 0.50 or 0.51–0.80, the quality of equations is good or satisfactory, respectively)

River, gauge line Gauging station YearsPeriod

A B R Z S

Q AP

Kuban, Krasnodar Krasnodar 1912–1975 X–IX A 4.4 0.02496 0.80 8.0 1.6 0.59

Dondukovskaya 1937–1942, 1946–1972 X–IX A 4.8 0.02548 0.83 5.0 1.5 0.56

Verkhnie Tuby 1937–1939, 1945–1972 X–IX A 5.9 0.00896 0.82 6.2 1.4 0.57

Arkhyz 1937–1972 I–XII X 9.0 0.01470 0.77 5.9 1.3 0.64

Khamyshki 1928–1942, 1944–1972 I–XII A 8.3 0.01055 0.96 11.5 0.5 0.28

Adagum, Krymsk Krymsk 1944–1972 XI–III X 28.0 0.01565 0.85 6.5 11.4 0.53

Belaya, Severnyi ridge Verkhnie Tuby 1937–1938, 1946–1963 I–XII X 45.2 0.09756 0.95 7.0 7.3 0.33

Teberda, Teberda Arkhyz 1934–1935, 1952–1970 X–IX A 19.5 0.01938 0.79 4.5 1.4 0.61

692

WATER RESOURCES

Vol. 29

No. 6

2002

GARGOPA

types of atmospheric circulation in the cold season (

=0.69 and 0.49, respectively). The higher frequency of

form and particularly of

form produces a nega-tive effect on river runoff (

= –0.48 and –0.71, respec-tively). In the period of regulated river runoff, thedegree of their influence, particularly of the latter type,is statistically insignificant (

= –0.30 and –0.11,respectively). With the development of

and

forms,the amount of precipitation slightly decreases (

= –0.22and –0.33, respectively) and in the case of the

typedomination, increases (

= 0.51).

Statistical analysis of the effect of atmospheric pro-cesses on the river runoff in 1983–1998 and throughoutthe period of runoff regulation (1952–1998) confirmedthe regularities found for 1952–1982 and discussedabove. For the whole period of the Kuban river regula-tion, the most statistically significant is the positiveeffect of an increase in the number of days with

and

forms in the cold season (

= 0.41 and 0.43,respectively); during the last 16 years, macroprocessesof the types of

and

were most essential forriver runoff (

= 0.56 and 0.53, respectively). The devel-opment of

form above the norm led to a decrease inthe Kuban runoff (

= –0.65). The effect of the atmo-spheric processes in the cold season on the Don Riverrunoff and on the total inflow of river water into the Seaof Azov was statistically insignificant.

The assessment made for the periods of 1923–1951and 1952–1982 showed that under natural conditions,the correlation between salinity and the annual fre-quency of

and

forms is positive (

= 0.25–0.38); thewestern type of macroprocesses causes a decrease insalinity (

= –0.52) in both natural and regulated runoffperiods. Under the conditions of regulated runoff,

form also contributes to desalination (

= –0.49). The

form contributes to an increase in salinity in both peri-ods—of natural and regulated runoff (

= 0.38 and 0.58,respectively). Under the conditions of regulated runoff,salinization is most intense in situations, when

E, E

,and

forms develop in the cold season (

= 0.64,0.52, and 0.49–0.56, respectively). Salinity decrease ismaximal when

W, N

, and particularly

W + N

formsdevelop in the cold season (

= –0.52, –0.54, and –0.64,respectively). Under natural conditions, the effect of

W,N

, and

forms is also negative (

= –0.44, –0.33,and –0.50, respectively); the effect of

and

forms is positive (

= 0.27 and 0.41, respectively).

Statistical analysis of atmospheric processes in thesucceeding years and during the entire period of avail-able observations (1952–1998) confirmed the conclu-sions based on the data for the period since the start ofthe river runoff regulation in the Sea of Azov basin untilthe early 1980s. For 1983–1998, the closeness of corre-lation between the salinity and different types of atmo-spheric macroprocesses is much higher: as for

formcontributing to salinization (

= 0.81), so for

and

(

W) forms contributing to the Azov waterfreshening (r = –0.70…–0.84). The steadiness andgrowing closeness of the relationships described abovewere supported by checking against an independent16-year series (1983–1998), which proved the reliabil-ity of statistic equations approximating the relationshipbetween the salinity and the atmospheric circulationforms in 1952–1982. Only in three cases (1983, 1992,and 1998), the calculated values of salinity exceededthe observed values by 0.7–0.9‰, the admissible errorbeing equal to 0.55‰; in other years, the differencebetween calculated and observed values was muchsmaller than the admissible error. With this, the averagevalue of salinity calculated for the whole verification

Table 2. Parameters of linear equations approximating the dependence of the salinity of the Sea of Azov Ssi, the sea properSp.si, and Taganrog Bay Sbi in the given year i versus the Don and Kuban total annual runoff in the given i and preceding(i +…+ i – n) years

S, ‰Q, km3

A B R Z Sy Sy /σyperiod years

1960–1986

Ssi X–IX i + … + i – 5 17.51 –0.028227 –0.96 8.3 0.16 0.19

Ssi VI–V i + … + i – 5 17.31 –0.032741 –0.96 8.3 0.15 0.19

Sp.si X–IX i + … + i – 5 17.76 –0.027869 –0.96 8.3 0.15 0.19

Sp.si VI–V i + … + i – 5 17.58 –0.032316 –0.96 8.3 0.15 0.19

Sbi X–IX i 12.79 –0.148744 –0.84 4.1 0.71 0.46

Sbi VI–V i 13.66 –0.174691 –0.86 4.1 0.70 0.43

1987–1999

Ssi X–IX i + … + i – 5 16.67 –0.028513 –0.82 3.7 0.29 0.39

Ssi VI–V i + … + i – 5 16.68 –0.028538 –0.84 4.3 0.27 0.38

Sp.si X–IX i + … + i – 5 17.13 –0.029124 –0.83 3.7 0.28 0.39

Sp.si VI–V i + … + i – 5 17.14 –0.02913 –0.85 4.0 0.26 0.37

WATER RESOURCES Vol. 29 No. 6 2002

CURRENT DESALINATION OF THE SEA OF AZOV AND ITS CORRELATION 693

series exceeds the respective observed value by 0.16and for 1983–1997, only by 0.08‰. This suggests thatin the last 20 years, the effect of atmospheric climate-forming processes on the variability of the water bal-ance structure and salinity of the Sea of Azov hasbecome stronger, whereas the effect of the human-induced decrease in river runoff has diminished. In therelatively high-water years of 1987–2000, the volumeof river runoff averaged 34.6 km3/yr, which exceeds themany-year average only by 1 km3 (3%). In 1983–2000,the total river runoff of the Don and Kuban (33.0 km3/yr)was on average even below (by 0.6 km3/yr) its normalvalue. During the period of the Sea of Azov desalina-tion (1977–2000), the river runoff (34.5 km3/yr) alsodiffered only slightly from the runoff in the period ofsalinization (1952–1976). It should be mentioned thatthe strengthening effect of atmospheric macrocircula-tion processes on the Sea of Azov salinity was earliermentioned in [9], where the anthropogenic, geophysi-cal, and cosmic factors governing the circulationepochs and variations in the Caspian Sea level wereanalyzed. For the whole period of regulated runoff, thecloseness of statistical correlation between the Sea ofAzov salinity and the circulation forms (E, N, W, andW–N) is also quite satisfactory (r = 0.73, –0.62, –0.58,and –0.73, respectively), the lag between the change ofthe sign of the circulation form frequency anomaly andthe relevant change in the salinity being 1–6 years.

Table 3 presents some equations approximating thedescribed relationships between the salinity and theintensity of different types of atmospheric processes forthe period of regulated river runoff.

CORRELATION BETWEEN LONG-TERM VARIATIONS IN ATMOSPHERIC CIRCULATION, WATER BALANCE ELEMENTS, AND SALINITY

The reliability of the cause-and-effect and structuralrelationships between atmospheric circulation, riverrunoff, and the Sea of Azov salinity, which are deter-

mined in this work and described above, is confirmedby the correlation of their variations during many years.

Analysis of variations in the annual frequency ofoccurrence of the atmospheric circulation forms overthe period of 1891–1998 showed that the frequency ofW form demonstrates a negative trend; E form has apositive trend, and N form has a slightly pronouncedtendency to a decrease. The most distinct and steadytransition to the negative (as compared with the norm)zone of the annual, spring, autumn, and winter W fre-quency occurred in the early 1950s, but especially sig-nificant and steady shift to this zone was observed sincethe early 1960s until the second half of the 1980s. Atapproximately the same time, the E form frequencypassed into the positive zone, an increase in E fre-quency being the most significant and steady since themid 1960s until the early 1990s.

Within the period of regulated river runoff, the low-est frequency of W was observed in 1968–1978 and thehighest frequency of E, in 1966–1986. In the latterperiod, N frequency was mostly (81% of cases) belownormal. Since the late 1930s until the beginning of thesecond half of the 1960s, the frequency of macropro-cesses of N form exceeded its norm in most years(76% of cases); in the period of regulated runoff, thisexcess was especially significant in 1956–1960, 1962–1966, and 1968–1974.

As for the last 12–13 years, the frequency of W andN forms in the cold season exceeded the norm, whereasthe frequency of E form was below the norm.

The results of analysis of the linear trend of long-term variations in the Don and Kuban natural (recon-structed) runoff since 1911 until the late 1980s areindicative of a slight tendency toward a decrease in therunoff (by 1.7 km3, or 9%). This tendency is strongerfor the period up to 1976 (6.7 km3, or 16%). Under theconditions of regulated river runoff, the total runoff ofthe Don and Kuban featured a positive trend (4–5 km3,or 12%) in the first case and a negative one (5.3 km3, or

Table 3. Parameters of linear equations approximating the dependence of S, ‰, in the given year versus the total number ofdays with different forms of atmospheric circulation in the cold season of the given year i and preceding years (i + … + i – n)

Type of atmospheric circulation Years A B R Z Sy Sy /σy

1952–1998

W i + … + i – 5 14.02 –0.0062 –0.59 4.5 0.34 0.43

E i + … + i – 3 8.31 0.0085 0.73 6.2 0.29 0.36

C i + … + i – 2 13.85 –0.0143 –0.58 4.4 0.34 0.43

C + W i + … + i – 2 15.14 –0.0102 –0.73 6.2 0.29 0.37

1983–1998

W i + … + i – 5 13.68 –0.0054 –0.78 3.8 0.28 0.35

E i + … + i – 5 8.96 0.0043 0.81 4.1 0.26 0.33

C i + … + i – 3 13.58 –0.0117 –0.70 3.1 0.30 0.38

C + W i + … + i – 5 14.47 –0.0043 –0.84 4.9 0.25 0.32

694

WATER RESOURCES Vol. 29 No. 6 2002

GARGOPA

13%) in the second case. The last 13–14 years understudy were characterized by the phase of enhanced rateof streamflow, particularly in the Kuban River. Thisphase contributed to the formation of a positive trend inthe Don and Kuban total runoff (up to 6.8 km3, or 19%)in the period of regulated runoff and a slight tendencyto its increase (1.1 km3, or 3%) in 1911–1999. In otheryears of regulated river runoff, a slightly marked ten-dency to the runoff decrease (1.1 km3, or 3%) wasobserved. It was most significant (12.3 km3, or 31%) in1952–1976, when the Don runoff decreased by 9.2 km3

(36%) and that of the Kuban, by 3.3 km3 (23%).

Until the 1950s, the evaporation from the surface ofthe Sea of Azov (as from all southern seas of Russia)have been increasing; then, it started to decrease withapproximately the same rate. Variations in the atmo-spheric precipitation had a tendency to increase in the20th century [11]. Analysis of data (although incom-plete) on the rates of precipitation and evaporation indi-cates that the above tendencies persisted during the last10–15 years under study.

Analysis of the linear trends in the long-term varia-tions in the Sea of Azov salinity yielded positive trendswithin the periods of conventionally natural runoff(0.7‰), the early 1920s–the end of the 1960s (1.6‰),the mid-1970s–mid-1980s (2.4–2.5‰), and the end ofthe observation series under study (1.3–1.4‰). Thesalinity of Taganrog Bay also featured positive trends inthe above periods (0.9, 1.6, 3.4, 2.9, and 2.0‰, respec-tively). An almost double decrease in the trend value for1922–2000 (as compared with its maximal value for theperiod from 1922 to the middle of the 1970s–1980s) isassociated with a steady desalination of the Sea of Azovafter 1976, when its maximal salinity was observed.

Under the conditions of regulated river runoff, thelong-term variations in salinity feature two main peri-ods of salinization (1952–1976) and desalinization(since 1977 until present time), as well as several char-acteristic periods of smaller duration. In particular,these are the period (1949–1953) of a decrease in riverwater inflow into the Sea of Azov and an increase in itssalinity (from 11.3 up to 12.4‰), the 13-year period(1954–1966) of an increased river runoff and decreased(to 10.8‰) salinity, and the succeeding 10-year period(1967–1976) of the sharply defined negative anomalyof the river water inflow into the Sea of Azov and therelevant positive anomaly of the sea salinity (saliniza-tion up to 13.8‰, the absolute maximum for the wholeobservation series). An increase in the freshwater run-off by 15–16 km3/yr (as compared with the precedingextremely low-water (23 km3/yr) five-year period(1972–1976)) caused a rapid decrease in salinity to10.9‰ in 1982. The process of desalination at theexpense of the increased river runoff discontinued in1983–1986 and then recommenced at the expense of anincrease in the Kuban River runoff. As a result, the Seaof Azov salinity has changed from 10 to 11‰ duringthe last 8 years under study.

Until 1968, the character of atmospheric circulation(positive anomaly of the E form frequency in mostyears) contributed (on the whole) to a decrease in thesalinity, notwithstanding anthropogenic withdrawals ofriver runoff. Thereafter, a sharply defined negativeanomaly of the macroprocesses of W, N, and N + Wforms, as well as the exceptional development of Eform, led to the Sea of Azov salinization. Since 1977,the period of the Sea of Azov desalination began. How-ever, in its first five years and until 1986 including, thefrequency of W, N, and W + N forms was below thenorm, despite the tendency for growth; the E form fre-quency was well above its norm. K.V. Kondratovich[9], N.S. Sidorenkov and V.I. Shveikina [14], who stud-ied the macrocirculation reasons for the Caspian Sealevel rise and the variations in the climate regime of theVolga Basin and the Caspian Sea, explain the abovephenomenon in the following way. In these years, avariant of E form (defined by G.Ya. Vangengeim)started to develop. In winter, this form entails the local-ization of the Mediterranean cyclones over the waterareas of the Azov, Black, and Caspian seas. Thesecyclones bring cloudy weather and an increasedamount of precipitation, which, in their turn, contributeto a decrease in evaporation and increase in river runoff.Indeed, analysis of the typical maps (compiled byA.S. Girs [6–8]) of distributions of air pressure, tem-perature, and precipitation at different types of atmo-spheric macroprocesses allows us to suppose that thevariant of the eastern form E, at which the probabilityof the positive anomaly of winter precipitation is 50–70%, was likely to be observed at that time.

The available data on the frequency of differentforms of atmospheric circulation in the cold season of ayear (1987–1998) shows that in four cases, macropro-cesses of the W type dominated and in six cases, W + Nor N + W dominated. At this time, the N form frequencyexceeded the norm or was equal to it (in six cases).Such atmospheric processes are particularly favorablefor the formation of river runoff and, consequently, forthe maximal increase in freshwater balance and maxi-mal decrease in the Sea of Azov salinity. The E form,which causes the reverse effect, absolutely dominatedonly in one case and was combined with W and N formsin three cases. Thus, in 11 years out of 12, the atmo-spheric circulation contributed to the Sea of Azovdesalination (to 10–11‰) in 1993–2000.

It should be noted that anthropogenic factors alsomade a certain contribution to the current desalinationof the Sea of Azov. A particular factor is a significantreduction of the river water withdrawals downstream ofthe calculation gauge lines in deltaic areas, especially inthe basin of the Kuban River: a considerable increase inthe Kuban river runoff since 1987 led to an unusualexpansion of the zones of freshened water in the easternpart of the Sea of Azov. One of the reasons for increas-ing the surface river runoff in the basins of the Kuban’stributaries is the destructive deforestation in the West-ern Caucasus during the last decade, when the decrease

WATER RESOURCES Vol. 29 No. 6 2002

CURRENT DESALINATION OF THE SEA OF AZOV AND ITS CORRELATION 695

in the Kuban river runoff within the reach from Krasn-odar to the mouths of the Kuban branches averaged1.1 km3/yr, i. e., twice as low as that in the preceding1965–1986. Irreversible withdrawals of the Don runoffalso decreased and made about 6.5 km3/yr (total for thewhole basin) during the last 3–4 years. An increased(due to natural and anthropogenic factors) rate ofstreamflow in the cold seasons of 1977–2000, and par-ticularly of the last 12–13 years, played the most impor-tant role in the Sea of Azov desalination. The year of1997 is distinguished for a very high water in the KubanRiver, which caused an extraordinary (for the condi-tions of regulated river runoff) decrease in the Azovwater salinity. On the whole, the total runoff of the Donand Kuban rivers in the autumn-and-winter season onlyin the gauge-lines of Razdorskaya and Krasnodarincreased by almost 4 km3/yr, or by one third. Desalina-tion of the most part of Sivash Bay at the expense ofwater discharged from the Severo-Krymskii (the North-Crimean) canal also played a certain role in the Sea ofAzov desalination, because the Sivash water enteringthe Sea of Azov produces, in the warm season, not thesalinazing but, probably, a slight and short-lived desali-nating effect on the adjoining sea area.

Doubtlessly, the Sea of Azov salinity depends on thedirection and intensity of the water exchange betweenthe Azov and Black seas through Kerch Strait. Thisexchange varies depending on the atmospheric pro-cesses described above and on the sea levels, whichwere high during the last quarter of the 20th century;probable anomalies of the wind direction and speedover the area in question could also affect the regime ofsea salinity. To evaluate the above changes, moredetailed investigations, field studies including, areneeded. Nevertheless, in [5], for example, it is notedthat in the period of the spring-and-summer flood in therivers and the following level rise in the Sea of Azov,southern winds and calms contribute to the freshwaterretaining in the sea, thus reducing its salinity. Northernand northwestern winds facilitate freshwater carryingout of the sea, thus increasing its salinity. In the periodsof the autumn and winter low water in rivers and theSea of Azov lowered level, winds of the northern direc-tion decrease the sea salinity by preventing the inflowof the Black Sea water, whereas southeastern and west-ern winds are favorable for this inflow and increase thesalinity. With this, the salinazing effect of winds is moststrong in the medium- and particularly low-water peri-ods. For example, in the extremely dry period of 1972–1978, the process of the Sea of Azov salinization wasassociated with frequent and intense advections of theBlack Sea water, whose influence could be traced evenin the near-sea part of Taganrog Bay. We may supposethat under the conditions of the enhanced river waterinflow into the sea (annual and particularly in theautumn-and-winter season) described above, the fresh-ening effect of the northern air transfer was strong,whereas the salinazing effect of southeastern and west-ern winds was weak. The traces of the Black Sea water

advections through the Kerch Strait, which wereobserved during the oceanographic surveys of the Seaof Azov, are presently found in decreasing frequencyand have the form of well defined frontal zones of con-siderable dimensions. Most often, especially in the last8 years, they are found in the form of “lenses” of trans-formed Black Sea water (for example, such lenses werefound in the eastern and southeastern parts of the seaproper in 1994 and 1996).

One more important fact explaining the climatic rea-sons for the periods of the Sea of Azov salinization andfreshening is noted in [4].

Unlike the Sea of Azov water salinity, the Black Seasalinity (average for a layer of 0–200 m) increases withthe development of N form of atmospheric circulation(r = 0.75–0.89) in both periods, those of conventionallynatural (1923–1951) and regulated (1952–1985) riverrunoff. The effect of W form on salinity is found to bepositive in the period of regulated runoff (r = 0.68–0.75) and negative (r = –0.46…–0.51) in the period ofnatural runoff. The development of E form causes adecrease in salinity under the conditions of natural(r = –0.30…–0.34) and especially of regulated(r = –0.80…–0.85) river runoff. In this case, the effectof atmospheric circulation depends not only on the pro-cesses in the cold season but, to even greater degree, onthe processes in the course of the whole current yearand preceding 4–7 and more (8–20) years. This is firstassociated with the fact that the freshwater balance ofthe Black Sea increases in the years of the enhanced Eform frequency (r = 0.45–0.59) and decreases in the years

–10

19701950 1990 Year–20

20Σ(k – 1)/Cv

1234

Fig. 1. Normalized difference integral curves of the modulecoefficients of frequency of W, E, N, and W + N atmosphericcirculation forms (1–4, respectively) in cold seasons for thegiven year and five preceding years. Here and in Figs. 2 and3, K is the module coefficient equal to the ratio of the fre-quency of atmospheric circulation forms, sea water salinity,and river runoff in the given year to the whole-period meanvalues of the above characteristics; k – 1) is the alge-braic sum of deviations of the atmospheric circulationforms frequency, river runoff, and salinity in particularyears from the norms; Cv is the coefficients of variation forthe atmospheric circulation forms frequency, sea salinity,and river runoff.

(∑

696

WATER RESOURCES Vol. 29 No. 6 2002

GARGOPA

of prevailing N and W processes (r = –0.49…–0.47and –0.40…–0.47, respectively). In the first case, theBlack Sea level rises and in the second case, it drops [10].For the Sea of Azov level, the relationships are reverse.Therefore, in the years when E form prevails, the rela-tive position of the levels of these seas is likely to createthe hydrodynamic prerequisites for intensification ofthe Black Sea water advections and the Sea of Azovsalinization, whereas in the years with enhanced fre-quency of N and W macroprocesses, the inflow of waterfrom the Black Sea into the Sea of Azov drasticallydecreases, and the Sea of Azov goes into the period ofdesalination. The most dramatic examples of this inter-action are the extreme salinization of the Sea of Azov in1972–1976 and its equally unusual desalinization last-ing since 1993 until the present time.

The long-term variations in the W frequency and theSea of Azov mean annual salinity are to a significantdegree asynchronous (64% of cases). The degree ofasynchronism for the combined form of W + N in thecold season and the Azov water salinity is higher (75%

of cases). Less significant (57% of cases) is the asyn-chronism of the long-term variations in the frequencyof the meridional-type processes and the Sea of Azovsalinity; as for E form, the degree of synchronism ishigh (79% of cases) (Figs. 1–3).

CONCLUSIONS

The river runoff into the Sea of Azov and the seafreshwater balance increase and the sea salinitydecreases in the years with a higher frequency of W andN forms of atmospheric circulation. Domination of Eform produces the reverse effect. With this, the rever-sals of the sign of anomalies of atmospheric processeslead (by 1–6 years) the respective changes in the Sea ofAzov water balance and salinity. Macroprocesses of Wand E forms, which have long-period components intheir variation structure, generate the most prolongedand steady (up to 12 years) periods of negative and pos-itive anomalies of the river water inflow into the Sea ofAzov and sea salinity, whereas N form is responsiblefor the relevant negative anomalies of shorter duration(mostly 6–7 years). The last 20 years are characterizedby an enhanced effect of macrocirculation atmosphericprocesses on salinity and by a relatively weak impact ofthe human-induced reduction of river runoff. The cur-rent desalination of the Sea of Azov down to 10.5‰ onaverage is mainly due to the development of W and Nforms of atmospheric circulation in the cold seasons ofthe last 10–15 years. The likely retention of the positiveanomaly of W form is a reason to believe that the vari-ations in the Sea of Azov salinity would be within 10–11.5‰ during the nearest decades.

REFERENCES

1. Apollov, B.A., Kalinin, G.P., and Komarov, V.D., Kursgidrologicheskikh prognozov (Course in HydrologicalForecasts), Leningrad: Gidrometeoizdat, 1974.

2. Vangengeim, G.Ya., Meteorol. Gidrol., 1938, no. 3,p. 38.

3. Gargopa, Yu.M., Gidrobiol. Zh., 1988, vol. 24, no. 6,p. 63.

4. Gargopa, Yu.M., Proc. Int. Sci. Conf. Problemysokhraneniya ekosistem i ratsional’nogo ispol’zovaniyabioresursov Azovo-Chernomorskogo basseina (Prob-lems of Preserving Ecosystems and Rational Use ofBioresources in the Azov–Black Sea Basins), Rostov-on-Don, 2001, p. 42.

5. Gidrometeorologiya i gidrokhimiya morei SSSR(Hydrometeorology and Hydrochemistry of the Seas ofthe USSR), Simonov, A.I. and Al’tman, E.N., Eds.,Leningrad: Gidrometeoizdat, 1991, vols. 4 and 5; 1992,vol. 3.

6. Girs, A.A., Mnogoletnie kolebaniya atmosfernoi tsirku-lyatsii i dolgosrochnye gidrometeorologicheskie prog-nozy (Long-Term Variations in Atmospheric Circulationand Long-Range Forecasts), Leningrad: Gidrometeoiz-dat, 1971.

19701950 1990 Year–10

20Σ(k – 1)/Cv

Fig. 2. Normalized difference integral curves of the modulecoefficients of the (1) Don and (2) Kuban river runoff and(3) water inflow into the Sea of Azov for the given year andfive preceding years.

19701950 1990 Year–10

20Σ(k – 1)/Cv

Fig. 3. Normalized difference integral curves of the modulecoefficients of the mean annual values of Ss, Sp.s, and Sb (1–3,respectively).

WATER RESOURCES Vol. 29 No. 6 2002

CURRENT DESALINATION OF THE SEA OF AZOV AND ITS CORRELATION 697

7. Girs, A.A., Makrotsirkulyatsionnyi metod dolgos-rochnykh meteorologicheskikh prognozov (Macrocircu-lation Method for Long-Range Meteorological Forecast-ing), Leningrad: Gidrometeoizdat, 1974.

8. Girs, A.A. and Kondratovich, K.V., Metody dolgos-rochnykh prognozov pogody (Methods for Long-RangeWeather Forecasting), Leningrad: Gidrometeoizdat,1978.

9. Kondratovich, K.V., Vodn. Resur., 1994, vol. 21, no. 6,p. 623.

10. Lappo, S.S. and Reva, Yu.A., Meteorol. Gidrol., 1997,no. 12, p. 63.

11. Nikolenko, A.V., Vodn. Resur., 1997, vol. 24, no. 3,p. 261.

12. Rakhmanov, V.V., Tr. GMTs SSSR, 1973, no. 114.13. Rozhdestvenskii, A.V. and Chebotarev, A.I., Statis-

ticheskie metody v gidrologii (Statistical Methods inHydrology), Leningrad: Gidrometeoizdat, 1974.

14. Sidorenkov, N.S. and Shveikina, V.I., Vodn. Resur., 1996,vol. 23, no. 4, p. 401.

(PDF) Current Desalination of the Sea of Azov and Its Correlation with Long-Term Variations in Atmospheric Circulation - DOKUMEN.TIPS (2024)