Hydrogeological framework

According to surficial observations, and due to their subaerial position for several million years, all the carbonate rocks cropping out in the area are deeply and intensely karstified (Fig. 4). Although the spring emerges from the Upper Pleistocene Red Beds, the source volume of the spring consists mainly of rocks belonging to the Mesozoic bedrock. Due to the complex geological and tectonic setting of the area uphill of Yperia Krini and its surroundings, the hydrogeological system is a typical karstic one. As a consequence, the basin and mechanism that feed the spring cannot easily be defined. In order to constrain this problem and to better define the aquifer system that feeds the spring, the hydraulic parameters and the hydro-chemical characteristics have been investigated.

Climatic data

In order to define the replenishment conditions of the aquifer, we used monthly values of precipitation and mean temperature obtained from the pluvio-metric station of Sotirio, for the period 1971-1995 (Y.E.B. Larissa). This station is located about 15 km north of Velestino, at an altitude of 47 m a.s.l.

From the distribution of the temperature of the air, averaged monthly and annually (Fig. 5a) for the years 1973-1995, the mean annual value for the whole period is 14.5°C with variations between 12.8°C (1974 and 1975) and 17°C (1995). By averaging the temperatures of each month (Fig. 5b), the coolest month is January, with a mean temperature of 4.8°C, while the warmest month is July with a mean of 26.1°C.

Fig. 4. Karstic features within the carbonate rocks of Mount Malouka confirming the high infiltration coefficient.
Fig. 5. Graphs representing (a) mean monthly and mean annual temperatures of the air and (b) mean monthly temperatures recorded at the pluviometric station of Sotirio for the period 1973-1995.

Similarly, from the distribution of the monthly (Fig. 6a) and annual (Fig. 6b) precipitations for the years 1971-1995, the mean annual precipitation is 420 mm with annual variation between 726 mm (1982) and 220 mm (1977). During the dry subper-iod 1984-1990, the mean annual precipitation was 325 mm, representing a reduction of 22.5% with respect to the mean annual value of the whole period.

The maximum, mean and minimum values of the monthly averaged precipitation for the same period are represented in Figure 6c. The driest month of the year is August, with mean rainfall of 12.6 mm, while the wettest month is November, with 72.3 mm. The comparison of Figure 5b with Figure 6c indicates that the dry season is from June to September. During the other months, the recharge of the humidity of the soil and the consequent replenishment of the aquifer occur.

Water discharge of the spring

In order to define the hydraulic behaviour of the aquifer, monthly measurements of water discharge of the spring for the period 1973-1995 (Y.E.B. Larissa) were used. Both annual and pluriannual variations of water discharge are clearly evident in Figure 7. The mean annual discharge of the spring is 241 l/s (Fig. 8), showing three maxima during 1973, 1983 and 1987, with 335, 320 and 317 l/s, respectively, and a minimum value of 75 l/s in 1994, which corresponds to only 30% of the mean annual discharge for the whole period.

However, if we consider the subperiod 1973 -1988, the minimum annual value is never lower than 230 l/s, while during the period 1989-1994 there is a continuous decrease of the mean annual discharge of the spring, with a strong reduction from 227.5 l/s to only 75.2 l/s, which also corresponds to a strong reduction of water stored in the reservoir.

By comparing the mean annual values of water discharge with those of precipitation (Fig. 8) for the period 1973-1995, we can observe that during the years 1973-1988, both positive and negative variations of the rain always induce comparable variations of the spring water discharge, though with some hysteresis. In contrast, from 1988 to 1994, while the distribution of precipitation is more or less uniform, though lower than average, the annual discharge is abated as much as the 86.5%, that is from nearly 220 l/s to 30 l/s during the summer of 1994. However, the observed rainfall reduction does not justify such a large contraction of the water discharge (Coutagne 1968). Possible explanations will be discussed later.

It is well known that the shape of a hydrogram is mainly related to the area extent of the karstic aquifer and to its hydraulic parameters (e.g. Maillet 1905; Mangin 1975; Soulios 1991; Padilla et al. 1994; Bogli 1994). In order to estimate (i) the volume of the groundwater stored within the aquifer and (ii) the volume of water emerging from the spring, the Maillet (1905) method is applied:

where Qt is the water discharge of the spring (m3/s) at the time t from the beginning of the dry period, and Q0 is the water discharge of the spring (m3/s) at the beginning of the depletion (t = 0).

Equation 1 was applied to the discharge of the dry seasons within the investigated period (Fig. 6a). Due to the relatively small number of data, only a rough estimate of the hydrogeological parameters (Q0, a) for each year can be obtained. Mean values of Q0 and a for the whole period 1973-1995 are 0.086 m3/s and 0.002 day"1, respectively. According to several authors (e.g. Drogue 1992; Bonacci 1987; Soulios 1985), a low a value indicates a slow drainage velocity of the karstic system, and that the water movement within the aquifer occurs via many microfractures and few wide conduits. At first glance, this inference seems to contradict the widespread karstic phenomena observed at the surface within the carbonate outcrops of the investigated area (Fig. 4).

Moreover, according to the Maillet's formula, it is possible to estimate the dynamic storage capacity, W0, of the reservoir in the saturated zone at the beginning of the dry season (t = 0), corresponding to about 3.7 x 106 m3, and this is the volume that is progressively depleted in the case of no recharge. When this volume is completely exhausted, the spring becomes typically temporary, with water discharge occurring only when precipitation occurs.

A karstic system can be theoretically recharged by (i) infiltration of precipitation, (ii) lateral underground leakage, or even (iii) episodic additional recharge from superficial waters. Following the case, the system reacts accordingly and the discharge commonly follows some fixed law distribution (Mangin 1971, 1975). By analysing the cumulative classified values of the discharge, it is thus possible to understand the behaviour of the system, for example the recharge and discharge (Mangin 1975; Soulios 1985; Bonacci 1993).

From the cumulative distribution of the monthly values of water discharge for the years 1973-1995 (Fig. 9), it is possible to recognize if additional and temporal water outlets exist in the karstic system. Indeed, the curve in Figure 9 is clearly segmented showing two important slope variations, at about

Fig. 6. Graphs representing monthly (a) and annual (b) precipitation recorded at the pluviometric station of Sotirio, for the period 1971-1995; (c) maximum, mean and minimum monthly precipitation for the same period.
Fig. 7. Graphs representing monthly water discharge recorded at Yperia Krini spring for the period 1973-1995.

90 1/s and 280 l/s. The first variation is probably related to the drilling of new bore-holes that extract part of the water from the karstic aquifer, thus diminishing the water discharge at Yperia Krini. The second variation is possibly due to the water surplus of the karstic system that probably generates temporary outlets (Campi & Dragoni 2000).

Hydrochemical data

More information is available from the chemical analyses, particularly concerning the provenance and quality of the water (e.g. Hem 1985; Freeze & Cherry 1979; Hounslow 1995; Stumm & Morgan 1981; Lloyd & Heathcote 1985; Matthes 1982). In the present work, 12 samples with monthly frequency have been collected from the Yperia Krini spring in the years 1995 and 1996. The results of the statistical analyses are represented in Table 1. The measured pH values indicate a neutral to slightly basic water, while a total hardness of 36.5-37.8°F indicates hard water, typical of the groundwater of limestone formations (Aminot 1974; Bakalowicz 1977).

The absolute and relative values of ion content generally give useful information about the litho-logical characteristics of the rocks permeated by underground waters (Truesdell & Jones 1974; Plummer et al. 1976). The chemical analyses of the samples indicate a Ca/Mg ratio between 3.2 and 4.2, which confirms the importance of calcium-rich materials, such as carbonate rocks, within the aquifer (Hem 1985; Christopher 1992). Similar

Fig. 8. Graph comparing the mean annual water discharge (Q) recorded at Yperia Krini spring and the mean annual precipitation (P) recorded at the pluviometric station of Sotirio, for the period 1973-1995.
Fig. 9. Cumulative distribution of water discharge of Yperia Krini spring during the period 1973-1995.

conclusions can be inferred from the positive values of the index of saturation of calcite, dolomite and aragonite (Table 1).

According to the triangular diagram of Piper (1953) and to the classification of Davis & DeWiest (1967), the water of the Yperia Krini spring is of calcium bicarbonate type (Ca-HCO3).

Recharge area

According to the hydrogeological and hydroche-mical investigations, the Yperia Krini spring behaves as a typical karstic aquifer. Although the Yperia Krini spring emerges from the Pleistocene terrigenous Red Beds, the bulk of the water volume originates a few metres below, from the Mesozoic limestones of the Malouka Mount. As mentioned above, the carbonate rocks of the Chalk-odonio Massif are clearly connected to the hydro-geological system. Nevertheless, even if we assume the maximum infiltration values proposed for karstified rocks in Greece corresponding to about 50% of precipitation (Marinos 1975; Soulios 1985; Manakos & Dimopoulos 1995), the total areal extent of these two major outcrops, which is about 13 km2, is clearly not sufficient to supply the water discharge observed at the spring. In fact, the 465 mm/a of mean annual precipitation for the years 1971-1984, that is before the dry period and before the widespread pumping of the underground waters, can account for only 95 l/s

Table 1. Minimum and maximum values of the major ionic concentrations in the water of Yperia Krini

Min.

Max.

Temperature of water (°C)

16.6

17.3

pH

7.5

7.9

Electrical conductivity (mS/cm at 20°C)

580

603

TDS (mg/l)

368

380

Total hardness (°F)

36.5

38

Ca++ (mg/l)

114

120

Mg++ (mg/l)

17

22

K+ (mg/l)

14.5

15.0

Na+ (mg/l)

0.58

0.70

(HCO3)2 (mg/l)

410

420

(SO4)2 (mg/l)

7.2

13

Cl2 (mg/l)

7.1

10.4

(NO3)2 (mg/l)

3.5

6.0

Ca++/Mg++

3.2

4.5

Saturation Index:

Calcite

+0.832

+0.850

Dolomite

+ 1.083

+1.190

Aragonite

+0.670

+0.830

Monthly data for the period 1995-1996.

Monthly data for the period 1995-1996.

of water discharge at the spring, against the measured average of 286 l/s. Indeed, consistent with the assumed infiltration coefficient, it is easy to calculate that the required replenishment area is about 38 km2.

According to the geological and tectonic framework of the region, a possible replenishment area contributing to the aquifer of Yperia Krini is represented by Partes-Pyrgaki Hill, just south of the Chalkodonio Massif (Fig. 2). Indeed, under a thin cover of Quaternary deposits, the two carbonate masses are in direct hydraulic contact, though slightly displaced by the Righeo Fault, which has downthrown the southern block a few tens of metres (Caputo 1995). These additional 17.8 km2 make a total replenishment area of about 31 km2, which is comparable to the expected one, especially if we consider further possible minor contributions to the hydrogeological balance given by different non-carbonate superficial lithologies, such as the Red Beds or other terrigenous rocks.

As concerns the system analysis, the low value of the coefficient of recession, a, is possibly related to the presence of cavities within the unsaturated zone of the aquifer system, whose number and importance progressively decrease with depth (saturation zone). Alternatively, the behaviour of the hydrogeo-logical system is probably due to the evident hydraulic bottleneck existing between Mount Malouka and the Chalkodonio Massif (Fig. 2), represented by the carbonate succession which is only a few metres thick and is interbedded with rocks of low permeability. These geometric characteristics of the hydraulic system explain the permanent and generally uniform behaviour of the spring. Indeed, most of the water is contained within the reservoirs of Chalkodonio Massif and Pyrgaki-Partes Hill, but its transfer to Mount Malouka, and subsequently to the spring, is evidently controlled by the above-mentioned bottleneck.

Was this article helpful?

0 0

Post a comment