Groundwater Monitoring in Saint-Petersburg: Past and Present
Corresponding author: Dr. Ernst Zaltsberg, Interenvironment Ltd., 1101-131 Torresdale Ave., Toronto, ON M2R3T1, Canada. Tel: +1-416-739-7963; E mail: firstname.lastname@example.org
Saint-Petersburg is the second largest Russian city with a population of approximately 5.2 million and with a total area of 605.8 km2. Founded in 1703, it was the Russian capital from 1712 till 1918. The City is located in the delta of the Neva River which flows westward from the Lake Ladozhs- koye to the Finnish Gulf of the Baltic Sea.
Numerous Neva’s tributaries and natural and dug channels flow through the City that is called «the Northern Venice». The City of Saint-Petersburg is incorporated in the North- West Federal District that includes nine municipal towns and 21 townships. The total population of the District is 5.4 million and the total area is 1,439 km2.
Early groundwater monitoring
The first systematic studying on the groundwater regime commenced in the 19th century, and the first publication on this subject appeared as early as in 1867. Based on the lim- ited data at the dug wells, Illish  made a conclusion that shallow groundwater elevation within the City were defined by water level elevations in the Neva River and its tributar- ies, instead of depending on precipitation. One year later
another article on groundwater was published in the same journal by Pell . The paper was illustrated by a map enti- tled “A map of soils and groundwaters in Saint-Petersburg”. It was based on water table measurements conducted at differ- ent times in 107 dug wells, soil excavations and ditches. The depth to the water table was shown by different colors. Each color was dedicated to the specific depth interval (0.0_0.3 m; 0.3_0.6 m; etc., with the maximum depth interval of 3.3_3.6 m). The areas with hydraulic connection between ground- water and surface water have been delineated. Pell  also indicated some specific areas where water table fluctuations depended on precipitation rather than the surface water lev- el fluctuations.
Therefore, both researchers focused their attention on the problem of hydraulic connection between groundwater and surface water that was fully understandable. The Neva Riv- er experienced frequent and unpredictable, sometimes cat- astrophic floods caused by strong west-eastern winds from the Finnish Gulf. Each flood resulted in the rapid and damag- ing groundwater table rise in the areas adjacent to the Neva River and its tributaries.
For further studying of this problem, 16 groundwater monitoring wells were installed in the Saint-Petersburg downtown area in 1877 under the supervision of the military engineer A.Tillo. The distance between the wells and the nearest sur- face water bodies ranged from 32 to 638 m. Water tables, pre- cipitation and air temperature were measured daily from June 1, 1877 to June 1, 1878. The accuracy of groundwater level measurements was 2mm; simultaneously, water level mea- surements in the Neva River were conducted. All monitoring results obtained were thoroughly analyzed and then published .
Comparing groundwater and surface water fluctuations, Til- lo  came to the conclusion that hydraulic connections be- tween them were observed in the narrow strip along surface water bodies where average annual water table elevations did not exceed 0.5 m above sea level (a.s.l.). In the areas where groundwater level elevations were higher than this mark, hy- draulic connections between groundwater and surface water were not observed. Tillo  pointed out that his conclusion was based on the one year observations only when there were no significant floods in the Neva River. He also suggested that even during significant floods when water level elevations in the Neva River could rise to 2.0_2.5 m a.s.l., they would not in- fluence groundwater fluctuations in the areas with the average annual water table elevations above 3.0 m a.s.l.
Groundwater temperature measurements conducted by Til- lo  indicated the groundwater “warming” effect associated with City’s buildings. Groundwater temperature in 12 wells installed in backyards ranged from 3.7 to 7.5 oC while the tem- perature in 4 wells installed in basements ranged from 9.9 to 16.9 oC.
The next round of groundwater monitoring was initiated by Professor Luboslavsky and conducted in 1892_1925 in the park of the Forestry Academy located at the northern suburb of the City at a distance of several kilometers from the Neva River. The water table was measured daily in several monitor- ing wells installed in shallow Quaternary deposits. One well was equipped with the recorder. In addition, precipitation, air temperature and soil temperature in the unsaturated zone were measured on the daily basis. Unfortunately, almost all measurements were lost during the Second World War. Only some of them along with conclusions have been published in a short article . Analyzing water table fluctuations and pre- cipitation, Voronina  indicated that “the influence of precip- itation on water table fluctuations is evident especially consid- ering their annual values”.
Considering daily water table fluctuations and precipitation, Voronina  indicated that the water table rise depended on the water content in the unsaturated zone immediately prior to precipitation. She also pointed out the influence of atmospher- ic pressure on water table fluctuations during winter seasons and the dependence between the spring water table rise and the snow cover thickness at the end of the winter season.
Groundwater monitoring after 1945
The most extensive groundwater monitoring and associated investigations commenced after the Second World War. By 1980 the monitoring network in the City consisted of ap- proximately 250 wells completed in various Quaternary and underlying deep Cambrian deposits. Quaternary water bear- ing formations included technogenic (urban) deposits, sands, loams, lacustrine clays, upper fractured till and sandy layers in massive till. They are encountered at the ground surface and have the total thickness of up to 140_160 m.
Groundwater monitoring was conducted within the upper seg- ment of these deposits and the depth of the monitoring wells usually ranged from 2_5 to 15_20 m.
Groundwater monitoring was also conducted within the ar- tesian Cambrian (Gdovsky) aquifer. It consists of sands and sandstones; its top is located at a depth of 140_160 m from the ground surface.
Shallow groundwaters within the City are responsible for basement flooding, water inflows into construction excava- tions and ditches and existing underground structures. In or- der to avoid unpredictable groundwater flooding, knowledge of the shallow groundwater regime is of paramount interest to construction companies and various City’s Departments re- sponsible for safe and efficient maintenance of underground structures and systems.
For many years comprehensive groundwater monitoring and studying of groundwater regime and balance in the City of Saint-Petersburg and its vicinity were conducted by the North- West Hydrogeological Station under the direction of prom- inent Russian hydrogeologists B.Archangelsky, P.Gass and N.Shvedchikova.Based on the long term monitoring results, several generalized hydrogeological maps for the territory of the City have been compiled. They included the following:
- The 1967 map of the shallow groundwater regime
(scale 1: 50,000)
The following areas were shown on the map:
- The areas with the natural groundwater regime which usually coincide with City’s suburbs and big parks. The water table regime is defined by precipitation and air tem- perature. Depending on lithology of shallow water bearing formations, seasonal and long term water table fluctua- tions range from 1_2 to 4_5 m;
- The areas with the partially disturbed groundwater re- gime that is caused by the medium density housing devel- opment and the presence of underground systems. The seasonal and long term groundwater fluctuations range from 0.5_1.0 to 1.5_2.0 m and depend on both the meteoro- logical and urban factors; and
- The areas with the disturbed groundwater regime which are located in the historical City’s downtown. Seasonal and long term water table fluctuations are flattened and usually in the order of 0.1_0.5 m. This regime is caused by dense urban development, the presence of numerous underground systems and structures, the asphalt and
concrete cover and impermeable embankments.
Within the areas with the natural and partially disturbed regimes two subareas were delineated:
- the subarea with the groundwater regime defining by meteorological factors and the absence of hydraulic con- nection between groundwater and surface water; and
- the subarea with hydraulic connection between groundwater and surface water. This subarea is stretched along the Finnish Gulf, the Neva River and its tributar- ies. The width of this subarea depends on its litholog- ical setting and ranges from 100 to 300 m. The annual average water table elevations within the subarea do not exceed 2 m a.s.l. Therefore, the previous conclu- sions regarding the size of this subarea made by Il- lish , Pell  and Tillo , were further clarified.
- The 1975 map of maximum water table elevations
- Maximum water table elevations in shallow Quaternary deposits were shown on this map. The water table mea- surements at approximately 300 shallow monitoring wells (both existing and abundant) have been utilized for the map’s compilation. The observation period in each well ranged from a few months to 25_27 years. Short observa- tion periods at some wells have been extended by means of correlation with measurements at corresponding long term observation wells; and
- In addition, single measurements at several thousand shallow exploration geotechnical wells installed during the high water table periods (springs and falls) were in- corporated into a map.
Since 1975 the map of maximum groundwater levels has been further verified and updated. In 1989 the map of maximum groundwater levels (scale 1:25,000) with probability of 5% was compiled by the Northwest Geological Survey. 20 years later this map was revisited again and the latest monitoring data were incorporated into it.
- The 1970s geological-hydrogeological maps at the depth of
10, 25, and 50m from the ground surface (scale 1: 50, 000).
- They were based mainly on the bore whole log data from more than 100,000 exploration geotechnical wells drilled at various time within the City’s territory. At each specific depth the main aquifers and aquitards have been delineat- ed; and
- Piezometric heads typical for main water bearing forma- tions at each specific depth were indicated at each of three maps.
For areas with natural and partially disturbed regimes repre- sentative wells with observation periods not less than 20–25 years were selected. Using the multiple correlation method, the
equation for predicting the winter minimum and spring maxi- mum water tables were calculated for each well .
For predicting the winter minimum level, the following predic- tors have been utilized: the maximum level in the previous fall and the average winter air temperature in December–Febru- ary. For predicting the spring maximum groundwater level two other predictors have been used namely the minimum previous winter level and winter precipitation in December– February or December-March. Using these equations, seasonal forecasts of the extreme groundwater levels were calculated, mapped and distributed among various Municipal and City Departments and construction companies which utilized them in planning their upcoming operations.
The results of long term groundwater level monitoring have been used for calculating groundwater balance components such as groundwater inflow and outflow, groundwater infil- tration and evapotranspiration. Such calculations have been conducted for numerous experimental sites within the City with various geological settings and types of the groundwater regimes .
General information on hydraulic properties of shallow Qua- ternary deposits that is necessary for these calculations was derived from pumping and water inflow tests conducted at the shallow monitoring wells. In addition, about 500 express tests were performed in shallow dug test pits. Each test consisted of water inflow into the pit and the following water level decline measurements. All results were statistically processed, and the average hydraulic conductivity value and its standard devia- tion for each water bearing deposit have been calculated and mapped.
Close correlation was established between the density of hous- ing development and the long term and seasonal groundwater fluctuations within various Quaternary water bearing depos- its. These equations were used for forecasting shallow ground- water levels in the areas of the proposed housing development.
Of special interest is hydraulic head monitoring within the deep Cambrian (Gdovsky) artesian aquifer. First water sup- ply wells were installed in this aquifer in the 1860s under the supervision of Professor A. Inostrantsev. At that time the piezometric head was above ground surface at elevation of 4 m a.s.l. In the following decades especially after 1945, water from this aquifer was widely used for cooling industrial equip- ments. In 1965 the water withdrawal from the aquifer reached its maximum of 36,000 m3/day, in 1975 the withdrawal rate was 32,000 m3/day . In the 1970s hydraulic head elevations of the Cambrian (Gdovsky) aquifer in the central part of the City were 65-70 m below sea level (b.s.l.) and the total area of the cone of depression was about 20,000 km2.
Starting from the 1980s, the Soviet economy began to decline, the industrial production in Saint-Petersburg was significantly shrunk and groundwater extraction from the Cambrian aqui- fer was reduced accordingly.
As a result, piezometric heads started to recover at a rate of ap-
proximately 0.5-1.5 m/year. In 2005 piezometric heads were observed at elevations 22.6_25.2 m b.s.l. The relatively slow recovery rate is due to the fact that the water withdrawal from the aquifer is still continuous for the municipal water supply beyond the City’s boundaries.
Table 1. The results of Fourier analysis by Zaltsberg .
Cyclicity of groundwater level fluctuations
Long term groundwater level monitoring allowed identifica- tion of the cyclic (harmonic) components in water table fluctu- ations using the classical Fourier analysis . Some results of this analysis are given in Table 1.
|Long term series analyzed||Duration of harmonics in years|
|1. The City of Saint-Petersburg
The average annual groundwater level in Quaternary deposits
|Well 160* (1946-1980)||3||5||9||26|
|Well 170* (1933-1980)||3||5||10||26|
|Well 860* (1946- 1980)||3||5||9||27|
|2. The North-West Federal District
The average annual groundwater level in Ordovician lime- stones
|Well 1002* (1933-1980)||4||11||16||27|
|Well 1009* ( 1933-1980)||5||8||16||25|
|3. The average annual water level in Lakes|
|4. The average annual flow in the Neva River (1860-1980)||6||11||29|
|The meteorological station|
|Annual precipitation (1900-1980)||5||6||11||33|
|Average annual air temperature|
|Average air temperature in the|
|winter season (1950-1980)||11|
Notes: * Measurements are partially reconstructed.
The following cycles (harmonics) were found in the long-term groundwater level fluctuations: 23-27, 16-17, 8-11, 4-5, and 3 years. The similar harmonics were observed in the long series of the hydrological data (the average annual water level in the nearby lakes, the annual flow in the Neva River). Harmonics of similar durations were found also in series of meteorolog- ical data. Such similarities should be expected because both groundwater level and surface water level fluctuations are de- fined mainly by meteorological conditions.
Fourier analysis allows defining several harmonics and the random component in the measurement series. Each harmon- ic is characterized by the amplitude, duration and the initial phase. Using these parameters, each harmonic specified and their sum could be extended beyond the observation period and, therefore, some prediction on the future annual average groundwater level with the lead time of 1-2 years could be made. Such forecasts have been calculated for the monitoring wells mentioned in Table 1, and their accuracy was satisfactory.
The maximum amount of groundwater monitoring works was conducted in Saint-Petersburg in the 1960s and 1970s. In the 1990s the state governed Soviet economy was transformed into the market economy. As a result, government’s funding for groundwater monitoring was significantly reduced. By 2005, the monitoring network in the North_ West Federal District consisted of about 70 shallow and deep monitoring wells .
It is interesting to compare the groundwater monitoring net- work in Saint-Petersburg and in some other cities. In the Rus- sian capital Moscow the number of monitoring wells was as follows: 369 in 1995, 280 in 2000, and 154 in 2008. In the big- gest Canadian City of Toronto there were 2 monitoring wells in 2010, while in the New York City there was only one monitor- ing well in the same year. Due to lack of the long-term monitor- ing data, there are no detailed hydrogeological maps for either Toronto or New York demonstrating the peculiarities of the groundwater regime within their territories. Such a deficiency has practical implication. For example, prior to commencing any construction project in Toronto, the significant amount of exploration work needs to be conducted for characterizing general geological/hydrogeological conditions at the con- struction site. In Saint-Petersburg this information could
be easily obtained from the previously compiled large scale hydrogeological maps based on the long-term ground- water monitoring results. During exploration work at the construction site in Saint-Petersburg the main efforts are fo- cused on solving site specific hydrotechnical problems rather than on gathering general hydrogeological information that is already available. As a result, a total cost of the construction work is significantly reduced and expenditures associated with map compilations have been returned in a relatively short period of time.
Using the huge previously collected groundwater monitoring data base, even the reduced groundwater monitoring network in Saint-Petersburg still satisfies the needs of various City’s services and construction and transportation companies, as well as provides valuable and timely information on the groundwater regime, balance and resources.
- Illish FS. Shallow groundwaters in Saint-Petersburg. J.Ar- chive of Forensic Medicine and Public Hygiene. 1867: 3_17.
- Pell AC. Soils and shallow groundwaters in Saint-Peters- burg. J Archive of Forensic Medicine and Public Hygiene. 1868: 8_24.
- Tillo AA. Groundwater level fluctuations in Saint-Peters- burg. Proceedings of the Imperial Russian Geographical Society. 1895, 29 (4): 1_37.
- Voronina O. Water table fluctuations and their dependence on meteorological factors. The Meteorological Bulletin. 1925, 6: 7_11.
- Zaltsberg E. Application of statistical methods for ground-water level forecasting. Nedra, Leningrad. 1976.
- Zaltsberg E. Evaluation of groundwater balance compo- nents at experimental sites in Leningrad. Proceedings of VSEGINGEO, Moscow. 1970, 27: 44-57.
- Golubeva D, Sorokina N. Protection of the environment. Use of natural resources and ecological safety in Saint-Pe- tersburg in 2005. The Committee for Environmental Pro- tection and Ecological Safety, Saint-Petersburg. 2006.