Recharge and mineralization of groundwater of the Upper Cretaceous aquifer in Orontes basin, Syria

Abstract

Chemical and isotopic data of groundwater of the Upper Cretaceous aquifer in the Orontes basin, Syria, have been used to assess the groundwater geochemistry, the origin of groundwater recharge and groundwater residence time. The chemical data indicate that dissolution of evaporite minerals is the main process controlling groundwater mineralization. The composition of stable isotopes δ¹⁸O and δ²H, together with ¹⁴C activity, reflect the existence of three different groups: (a) groundwater in the Coastal Mountains with δ¹⁸O of –6.65‰, quite similar to modern-day precipitation, and high ¹⁴C (>50 pmC); (b) groundwater in the unconfined aquifer of the Hama Uplift, which has δ¹⁸O of –5.52‰ and ¹⁴C near 20 pmC, and is recharged locally; and (c) groundwater from the confined aquifer of the Homs Depression, which is characterized by more depleted δ¹⁸O,, –8.01‰, and low ¹⁴C (<7 pmC), and might be recharged in the northern piedmont of the Anti-Lebanon Mountains. The distinctive isotope signatures of the latter two groups indicate different recharge elevations and palaeoclimatic effects. The low recharge rate of the groundwater in the Hama Uplift aquifer, and the even slower recharge rate in the Homs Depression aquifer, are reflected by groundwater ¹⁴C residence times of 5 and over 22 Ka BP, respectively.

Editor D. Koutsoyiannis

Citation Al-Charideh, A., 2013. Recharge and mineralization of groundwater of the Upper Cretaceous aquifer in Orontes basin (Syria). Hydrological Sciences Journal, 58 (2), 452–467.

Résumé

Les données chimiques et isotopiques des eaux souterraines de l'aquifère du Crétacé supérieur dans le bassin de l'Oronte, en Syrie, ont été utilisées pour déterminer les processus générant la composition chimique, l'origine de la recharge, et le temps de séjour des eaux souterraines. Les données chimiques montrent que la dissolution des minéraux évaporitiques est le processus principal contrôlant la minéralisation des eaux souterraines. La composition en isotopes stables δ¹⁸O et δ²H, ainsi que l'activité en ¹⁴C, reflètent l'existence de trois groupes différents : (a) les eaux souterraines des montagnes côtières avec δ¹⁸O de –6,65‰, assez semblable aux précipitations des temps modernes, et un ¹⁴C élevé (>50 PMC), (b) les eaux souterraines de l'aquifère libre du soulèvement de Hama, avec un δ¹⁸O de –5,52‰ et un ¹⁴C de près de 20 PMC et qui sont rechargées localement, et (c) les eaux souterraines de l'aquifère captif de la dépression de Homs, qui sont caractérisées par un appauvrissement accru en δ¹⁸O (–8,01‰) et un ¹⁴C bas (<7 PMC), et qui peuvent être rechargées dans le piémont nord des montagnes de l'Anti-Liban. Les signatures isotopiques distinctes entre les deux derniers groupes indiquent une altitude de recharge et des effets paléoclimatiques différents. La baisse du taux de recharge des eaux souterraines dans l'aquifère du soulèvement de Hama, et le taux de recharge plus prononcée dans l'aquifère de la dépression de Homs, se traduisent respectivement par des temps de résidence des eaux souterraines ¹⁴C de 5 Ka BP et e plus de 22 Ka BP.

Citation Al-Charideh, A., 2013. Recharge and mineralization of groundwater of the Upper Cretaceous aquifer in Orontes basin (Syria). Hydrological Sciences Journal, 58 (2), 1–16.

Key words:

• geochemistry
• isotopes
• dating
• Orontes basin
• Syria

Mots clefs:

• géochimie
• isotopes
• datation
• bassin de l'Oronte
• Syrie

1 INTRODUCTION

Both chemical and isotopic data are effective tools for solving various problems in hydrology, in particular in arid and semi-arid regions (IAEA 1980, 1983, Love et al. 1994, Clark and Fritz 1997, Cook and Herczeg 2000). Groundwater is an important source of drinking water for much of the population in Syria. Continuously increasing abstraction of groundwater resources to meet rising agricultural and domestic needs, leads to a growing deficit of water. The drawdown of piezometric levels and progressive degradation of water quality are the consequences of such intensive exploitation (Rasoul Agha 1999). In fact, this increase of groundwater abstraction from the several basins in Syria often requires exploitation of palaeo-groundwater that typically has depleted values of stable isotopes (relative to modern precipitation) and low ¹⁴C activities proving that no significant renewal of water has occurred recently (Selkhozpromexport 1979, Ferronsky et al. 1991, Wagner and Geyh 1999, Kattan 2002, Al-Charideh and Abou-Zakhem 2009, Al-Charideh 2011b). Similar isotopic patterns in groundwater are also observed in many basins of the Nubian Sandstone in northern Africa and in the arid zones of the Middle East (Gat and Issar 1974, Bajjali et al. 1997, Abd El Samie and Sadek 2001, Vengosh et al. 2007, Zouari et al. 2011). However, limestone and dolomitic rocks of Upper Cretaceous age constitute a highly productive aquifer in the highlands and mountain ranges of western Syria (Fig. 1). In the areas where they outcrop, which have a sub-humid Mediterranean-type climate, these carbonate rocks receive relatively abundant recharge and provide the principal reservoir of freshwater supply for the region. However, this aquifer extends downward and further east to the more arid parts of Syria, where it is mostly covered by a thick sequence of marls.

Fig. 1 Location map showing the main topographic features bounding the study area.

This work presents the primary results of a multiple tracer study focused mainly on the deep Upper Cretaceous aquifer; several samples taken from the outcropping aquifer were assumed to serve as reference data for the present-day recharge. However, preliminary radiocarbon and stable isotope determinations suggested the presence of fossil groundwater in a number of wells. Therefore, a multi-tracer approach (¹⁸O, ²H, ¹³C, ³H and ¹⁴C, as well as the main and some trace chemical constituents) was used to develop an initial schematic geo-hydrological model of groundwater recharge, circulation and chemical evolution. The specific objectives of this work are: (a) to identify the different recharge regimes; (b) to determine the dominant geochemical processes which influence groundwater chemical composition; (c) to determine its subsurface residence time; and, (d) to schematize groundwater flow patterns and the nature of aquifers in selected important hydrogeological regions of the basins. It is worth mentioning that the primary results from this study could be useful for other basins in Syria and in Middle East countries for understanding the recharge process and planning water resources management, in particular for the Upper Cretaceous aquifer.

2 STUDY AREA

The Orontes basin in western Syria extends over an area of about 21 600 km² (Selkhozpromexport 1979). The basin is bound to the west by the Coastal Mountains that reach 1500 m a.s.l., in the south by the northern Anti-Lebanon Mountains (2000 m), in the east by the northern Palmyrides Range of mountains (1300 m) and in the north by the Aleppo Plateau (Ponikarov 1967). The central sector of the study basin, the so-called Hama Uplift and Homs Depression, are relatively flat and situated at 300–400 m (Fig. 1).

The present climate of the study area is semi-arid Mediterranean in the west and changes to arid continental in the east. The N–S trending mountain ridge along the Dead Sea Fault acts as an orographic barrier that can be clearly seen in the rainfall pattern (Fig. 1). Thus, the annual precipitation ranges from 800 mm/year along the narrow western slopes of the Coastal Range to less than 100 mm/year east of the Hama Uplift and Homs Depression. Due to the small precipitation values, it follows that groundwater resources in the Orontes basin are replenished at low rates and there is consequently a high risk of over exploitation by pumping from wells. Since 1980, groundwater extraction from the Upper Cretaceous and Neogene aquifers system has increased greatly from 1000 to 1900 hm³/year in 2003, as a result of the growth in water needs (GCHS 2003, SDWC 2008).

The geological formations of the Orontes basin consist of marine carbonate deposits ranging in age from Upper Jurassic to Pliocene, with terrestrial sedimentation increasing from the Miocene onwards as the area was subjected to uplift accompanied by regression of the sea (Fig. 2(a)).

Fig. 2 (a) Simplified geological map of the study area; (b) schematic cross-section A–A′; and distribution of the sampling sites, nos. 1–40. Key: 1: Jurassic, 2: Upper Cretaceous aquifer (limestone and dolomites), 3: Upper Cretaceous (clay and limestone of Senonian to Maestrichtian), 4: Palaeogene (chalks and marls), 5: Neogene (conglomerates and limestone aquifer), 6: Basalt (Miocene/Pliocene) aquifer, 7: Quaternary, 8: Main faults, 9: Wells, 10: Cities.

The rock formations constituting the major aquifers in the area include the dolomite and limestone of the Upper Cretaceous, in particular those of Cenomanian and Turonian age (thickness 350–500 m). Palaeogene deposits, composed of chalks, limestone and marls occur but do not constitute important aquifers (Fig. 2(b)). The Neogene and Quaternary deposits are composed of sandstone, conglomerate, gravel and basalt and provide important aquifers.

The main aquifer in the study area comprises the strongly fissured and karstified massive strata of the limestone and dolomite of Cenomanian to Turonian ages, with marl, sandstone and gypsum intercalation in the lower part (Brew et al. 2001, SDWC 2008). The upper part of this aquifer is exposed at the surface in the vicinity of Hama and along a big anticlinal structure of the northernmost limb of the northern Palmyrides (Fig. 2(b)). However, the major part of this aquifer, in the Homs Depression, is confined and located between two aquitards of clays and marls: Aptian-Albian deposits serve as the regional bottom, and Senonian, Palaeogene and Pliocene deposits, of up to 250 m thickness, form the top east and southeast of the Homs Depression and Hama Uplift (Fig. 2(b)). The overlying Senonian to Maestrichtian marly and chalky limestone and dolomite, partly containing flints and gypsum, may form a combined aquifer system with the Cenomanian-Turonian rocks of high to medium productivity in the Homs Depression. The Cretaceous aquifer is known for its great hydraulic conductivity, and it is the most reliable aquifer not only in the Orontes basin, but also throughout Syria, in terms of storage capacity and the huge discharge of issuing springs (Al-Charideh 2011a).

Groundwater recharge is concentrated in the high rainfall areas along the narrow western slopes of the Coastal Range, and the northeast part of the Anti-Lebanon Mountains, where the rainfall is between 500 and 800 mm/year; roughly half of the rainfall may recharge the mountain aquifer systems (Selkhozpromexport 1979). According to rather poorly founded information on the piezometric surface, the unconfined groundwater of the Hama Uplift flows from the west towards the Al-Ghab rift valley, where it feeds the major karstic springs. However, the groundwater in the deep confined aquifer of the Homs Depression flows from the southwest, where the recharge could be related to the Anti-Lebanon Mountains, to the northeast. In addition, the main fault direction, S–N, might act as barrier to groundwater flow between the unconfined and confined aquifer systems, and cause the lowering of piezometric head by 40 m in the confined aquifer (Hatem 1995).

Finally, the main and the only drainage in this area is the Orontes River, which starts in the Bekaa Valley of Lebanon and flows in a northerly direction along the east side of the mountain ridge across the Al-Ghab graben structure, where it finally makes a sharp turn west to the Mediterranean. Major karstic springs are located in the Bekaa Valley at altitudes of 690 m, and discharge into the Orontes, maintaining a baseflow even in summer. The average flow of the Orontes is reported by the Ministry of Irrigation as 6.9 m³/s, or 217 hm³/year.

3 SAMPLING AND ANALYTICAL METHODS

A total of 40 groundwater samples were collected from active pumping wells (200 to 1650 m in depth) during September 2007. The sampling event was repeated in May 2008 (Fig. 2(a)). No significant difference was observed in either the chemical or the stable isotopes analyses of the first and the second sampling events.

Physical and chemical parameters were measured in the field, including: temperature (T, °C), pH, electrical conductivity (EC) and total alkalinity (i.e. HCO₃ ⁻). The total dissolved inorganic carbon (TDIC) for 16 groundwater samples was precipitated in the field from about 120 L of groundwater using the well-known procedure of IAEA (Tamers and Scharpensel 1970). The chemical analyses of major ions such as Cl⁻, SO₄ ^2-, NO₃ ⁻, Ca²⁺, Mg²⁺, Na⁺, K⁺ were conducted by liquid-ion chromatography. The charge balance was less than ±5% for all the analysed samples.

Stable isotope contents of the water samples were measured by isotope ratio mass-spectrometry, using the CO₂ equilibration technique for δ¹⁸O (Epstein and Mayeda 1953) and isotopic exchange with H₂ gas for δ²H (Coplen et al. 1991). The isotope contents are reported in the usual δ notation relative to the Vienna Standard Mean Ocean Water (V-SMOW) standard. Typical analytical uncertainty of the reported values is about ±0.1‰ for δ¹⁸O and ±1.0‰ for δ²H. The δ¹³C contents of total dissolved inorganic carbon were measured using isotope ratio mass spectrometry and reported in ‰ versus V-PDB (Vienna Pee Dee formation, North California, USA) (Coplen 1996) with a typical analytical uncertainty of ±0.1‰.

The analyses of tritium (³H) were determined by scintillation counter (Quantulus 1220) after electrolytic enrichment. The results are expressed in Tritium Units (TU) (one TU is equivalent to one ³H atom per 10¹⁸ atoms of hydrogen) with the analytical uncertainty in the order of ±0.3 TU at 2σ.

The ¹⁴C results are reported as per cent modern carbon (pmC); the degree of uncertainty depends on the quantity of analysed material, the count time and the activity of the sample. All chemical and environmental isotopes were analysed at the laboratories of the Atomic Energy Commission of Syria.

4 RESULTS AND DISCUSSION

4.1 Chemistry of groundwater

Physical and chemical characteristics of groundwater from the Upper Cretaceous aquifer systems (Table 1), show important variations with respect to increasing depth, progressive confinement from west to east and the geological facies variation of the aquifer. Groundwater temperature varies from 16 to 47°C in the study area. The total dissolved solids (TDS) values of groundwater samples range from 300 to 4000 mg/L (Table 1). The waters are weakly mineralized in the western part and highly mineralized in the deeper part of the aquifer in the northeast. The increase in TDS values of the groundwater can be explained by the chemical reactions that take place during water–rock interaction, resulting in the dissolution and precipitation of mineral phases. The majority of the groundwater samples have pH values of between 6.76 and 8.47, with a mean value of 7.3, indicating that the waters are generally neutral to slightly alkaline.

Table 1 Mean chemical composition and field data of groundwater samples from the Upper Cretaceous aquifer in the Orontes basin. See Fig. 2 for location of sites

Site	T	pH	TDS	HCO3	Cl	NO3	SO4	Na	K	Mg	Ca
no.	(°C)		(mg/L)	(mg/L)	(mg/L)	(mg/L)	(mg/L)	(mg/L)	(mg/L)	(mg/L)	(mg/L)
Recharge area
1	19.5	7.36	401	286	11.9	4.4	8.4	7.3	0.9	28	54.0
2	16.2	7.48	367	255	7.3	14.4	5.6	4.5	0.3	19.9	60.2
3	16.7	7.30	163	105	7.0	11.4	1.9	11.6	1.6	8.9	15.1
4	16.0	8.37	155	108	6.0	3.5	1.7	16.1	1	8.1	10.5
5	22.1	8.47	202	130	11.7	6.4	2	31.1	2.4	8.1	10.2
Hama Uplift
6	20.0	7.13	430	288	14.8	11.7	12.1	9.4	0.9	28.9	64.0
7	20.9	7.20	532	263	64.9	39.2	30.2	55.2	0.7	20.3	58.0
8	36.8	7.00	1873	210	87.0	0.0	1105.2	85.7	8.3	107.9	266.0
9	27.2	6.76	1071	363	37.5	4.2	391.5	40.4	5.8	60.3	165.2
10	23.1	6.97	460	259	34.0	10.0	44.8	21.9	1.7	25.5	62.8
11	24.9	7.11	567	266	46.4	7.3	103.6	32.5	4.2	32.7	74.0
12	25.3	6.90	1476	356	152.2	2.2	564.4	108.4	5.6	77	208.9
13	23.6	6.95	493	239	40.2	20.1	70.1	25.1	2.5	33.7	62.1
14	24.4	6.90	740	281	54.3	16.2	191.3	44.1	3.8	42.5	106.1
15	30.0	7.23	608	191	72.6	2.1	187.1	47.4	4.9	29.4	72.9
16	23.4	6.93	1132	336	141.5	Nd	352.8	107.3	6.1	59.7	127.4
17	25.8	6.97	1031	241	125.4	1.0	376.4	97	7.1	52.1	129.3
18	26.2	7.16	355	192	36.4	8.1	26.7	29.6	2.4	20.8	39.0
19	26.7	7.82	532	206	117.9	35.1	20.8	50.3	2.4	39.4	56.6
20	26.4	7.50	364	202	33.2	11.5	24.9	27.5	2	21.9	41.1
21	22.5	7.15	602	408	19.5	27.4	5.6	11.8	1.4	36.2	92.4
22	21.0	7.30	564	331	61.8	12.7	21.1	30.6	1.2	34.5	70.3
23	20.7	7.18	580	333	77.4	7.2	22.4	35.8	1	37	66.0
24	21.0	7.70	461	184	70.3	61.5	21.1	58.4	2.6	25.3	37.3
Homs Depression
25	35.0	7.00	463	251	49.0	0.4	44.4	31.4	3.6	27.7	54.6
26	40.0	7.07	386	229	37.8	Nd	32.3	25	2.6	30.4	28.2
27	41.0	7.05	520	291	58.1	Nd	42.3	36.1	2.9	31.3	56.6
28	46.6	6.95	667	280	130.5	Nd	77.1	91	3.6	33.4	48.7
29	29.6	7.06	849	297	116.2	Nd	204.6	101.6	6.6	50	66.7
30	25.9	7.43	1280	320	214.8	Nd	379.6	158.5	8.8	61.7	130.1
31	26.4	7.64	732	368	61.9	Nd	115.4	46.7	4.1	44.4	86.8
32	20.2	7.30	3880	197	990.9	7.4	1221.2	571.1	37.8	191.8	292.0
33	43.2	7.23	1289	358	339.2	Nd	195.4	229.3	13.8	54.3	91.7
34	45.0	7.08	1423	210	120.2	2.2	699.7	92.3	12	73.3	206.4
35	38.3	6.80	958	226	92.7	Nd	382	71.2	9.4	49.6	123.3
36	40.0	7.32	909	256	282.5	Nd	95.1	164.9	5.3	35.7	61.8
37	29.8	7.31	1220	310	175.3	Nd	386.2	137	3.5	71.3	136.3
38	20.0	7.6	537	247	75.5	9.7	69.4	43.8	2.7	30.6	57.4
39	23.5	7.62	336	233	10.1	4.8	12.1	8.7	1	21.4	44.4
40	45.0	7.85	1298	318	492	1.6	41.9	344.3	10.4	30.7	43.6
Note: Nd: not detected.

The differences in ionic composition indicate that there are three major water types, and the dominant anion species of the water change systematically from HCO₃ ⁻ to SO₄ ^2- to Cl⁻ (Fig. 3). Groundwaters from the Coastal Mountains aquifer have a Ca-Mg-HCO₃ geochemical facies typical of karst environments, indicating limestone and dolomite dissolution during infiltration.

Fig. 3 Average concentration and geochemical facies of the dominant groundwater groups in the study area.

The groundwater samples from the unconfined Hama Uplift also present Ca-Mg-HCO₃ geochemical facies. However, the higher Mg²⁺ and SO₄ ^2- content of these waters compared with the Coastal Mountains area indicates enhanced dolomite and evaporite dissolution. The salinity of these waters ranges between 400 and 1800 mg/L. The groundwater samples from the more confined section in the Homs Depression mainly have Ca-Mg-HCO₃ geochemical facies with very variable Na⁺ and Cl⁻ concentrations where the salinity ranges between 400 and 3880 mg/L.

One groundwater sample (well 32, 200 m depth) is of interest due to its high salinity, dominated by magnesium, sodium, chloride and high sulphate concentrations. Due to the shallow depth of this well, the groundwater may be fed by a significant contribution from surface water of Wadi Hbeishoun during flood periods. The isotopic data, showing relative enrichment and a trace of tritium content support this suggestion.

The groundwater samples have concentrations of Na⁺ between 5 and 571 mg/L and of Cl⁻ between 6 and 990 mg/L (Table 1). The high concentrations of Na⁺ and Cl⁻ detected in the southeast of the Homs Depression, where the aquifer is more confined (well depths > 900 m) may suggest the dissolution of chloride salts formed as very thin salt layers during evaporation of a closed or restricted marine basin several million years ago (El Naka and Al Kuisi 2004) (Fig. 4(a)). The molar ratios of Na⁺/Cl⁻ for groundwater samples of the study area generally range from 0.65 to 4.14, with a median value of 1.09 (Fig. 4(b)). Most of the Na/Cl molar ratios of groundwater samples with Cl⁻ <5 mmol/L exhibit higher values (0.9–4.1) than water in the hydrological cycle (0.86–1) suggesting an additional source of Na⁺, perhaps released from water–rock interaction, such as plagioclase dissolution of the volcanic basalt abundant in the study area (Magaritz et al. 1981, Sami 1992). Many of the groundwater samples of the unconfined aquifer of the Hama Uplift have lower Na/Cl ratios that are more similar to those of modern rainfall, reflecting recharge directly to the aquifer. The major part of the deep groundwater samples of the confined aquifer in the Homs Depression has Na⁺/Cl⁻ molar ratios of approximately 1, suggesting halite dissolution (Jankowski and Acworth 1997). The trend of Ca²⁺/Cl⁻ and Mg²⁺/Cl⁻ ratios versus Cl⁻ for the groundwater samples is similar to Na⁺/Cl⁻ versus Cl⁻ concentration (Fig. 4(c,d)). These data support a model whereby the chemistry of the low-salinity groundwater with Cl⁻ <5 mmol/L is controlled by different processes (mixing, evaporation and water–rock interaction). However, halite dissolution in the more confined aquifer, with Cl⁻ concentrations of >5 mmol/L, produces groundwater with lower, less variable cation/Cl ratios, consistent with homogenization and dissolution of evaporite sediment that produces the high groundwater salinity.

Fig. 4 Relationship between different cations and anions in the groundwater of the studied area.

The groundwater samples have concentrations of SO₄ ^2- of between 2 and 1200 mg/L, and of Ca²⁺ between 10 and 300 mg/L. The plot of Ca²⁺ and SO₄ ^2- (Fig. 4(e) shows a clustering of data at low SO₄ ^2- values, which may indicate changes in the Ca²⁺ concentrations that are associated with relatively minor change in SO₄ ^2- (Table 1). However, for some groundwater samples, there is a linear increase between Ca²⁺ and SO₄ ^2- that is close to the theoretical regression line for gypsum dissolution in equilibrium (slope = 1) suggesting that both ions are controlled by gypsum dissolution. The computation of saturation indices shows a progressive saturation in gypsum with an increase in SO₄ ^2- concentration (Fig. 5). Also, the strong observed relationship (R² = 0.87) between SO₄ ^2- and Mg²⁺ (Fig. 4(f)) indicates that these are most likely derived from the same evaporite dissolution. Water samples that are placed below the 1:1 line indicate a deficit of Ca²⁺ and Mg²⁺ relative to SO₄ ^2-. This depletion of Ca²⁺ and Mg²⁺ indicates the contribution of another geochemical process that effects cation contents and may be related to the mechanism of base exchange, by which Ca²⁺ and Mg²⁺ are adsorbed on the surface of clay minerals with the release of Na⁺, causing more saline water (Chapelle and Knobel 1983, Plummer et al. 1990). In addition, some sulphate reduction is detectable in several wells (23, 24, 33 and 39). The deepest water (from well 39) has isotopically light δ¹³C ratios, which are possibly evidence of bacterially-mediated sulphate reduction (Edmunds et al. 1995).

Fig. 5 Relationship between mineral and water saturation indexes (SI) of calcite, dolomite, gypsum and halite.

The mineralogical constraints of the water–rock interactions are confirmed by examination of the saturation indexes (Fig. 5). Saturation indexes and carbon partial pressures (P _CO2) of all groundwater samples were calculated using the geochemical code NETPATH 2.0 (Plummer et al. 1994). Taking into account the accuracy of the field measurements of pH and alkalinity, and the uncertainties of the other analytical data, the waters are in equilibrium with respect to calcite if the saturation index ranges from –0.1 to +0.1, and from –0.2 to +0.2 for dolomite (Nordstrom and Ball 1989). Accordingly, our results show that the groundwater samples in the recharge area are in equilibrium with calcite and dolomite at low P _CO2 conditions (10⁻⁴ to 10⁻³), and are strongly under-saturated with respect to gypsum and chloride. Most groundwater samples from the unconfined aquifer of the Hama Uplift and confined aquifer of the Homs Depression are in equilibrium or over-saturated with respect to calcite and dolomite, and some approach saturation with respect to gypsum. The P _CO2 values of these samples are nearly an order of magnitude higher than those of the recharge area (10⁻²). All the groundwater samples are under-saturated with respect to halite, but for several groundwater samples from the Homs Depression, the saturation state for halite is about four orders of magnitude higher than in the groundwater from the Hama Uplift and recharge area.

Table 2 Isotopic composition of groundwater samples collected from the Upper Cretaceous aquifer system in the Orontes basin. See Fig. 2 for location of sites

Site no.	Depth	Lat. N	Long. E	δ^18O	δ^2H	d-Excess	Tritium
	(m)			(‰)	(‰)	(‰)	(TU)
Coastal Mountains
1	500	35.06	36.34	−6.57	−32.70	19.86	4.40±0.2
2	560	35.17	36.31	−6.80	−33.81	20.59	2.59±0.2
3	560	34.89	36.37	−6.65	−30.01	23.19	1.77±0.2
4	430	34.80	36.38	−6.53	−28.68	23.56	4.37±0.2
5	537	34.78	36.46	−6.71	−29.70	23.98	n.m
Hama Uplift
6	250	35.15	36.57	−5.95	−29.47	18.13	2.53±0.2
7	700	35.31	37.07	−4.01	−26.64	5.44	0.84±0.2
8	500	35.35	36.97	−6.05	−39.10	9.30	0.60±0.2
9	336	35.31	36.80	−5.24	−33.10	8.82	0.60±0.2
10	200	35.26	36.67	−5.67	−30.89	14.47	1.97±0.2
11	350	35.21	36.72	−5.60	−32.13	12.67	0.98±0.2
12	258	35.44	36.62	−5.57	−35.46	9.10	0.60±0.2
13	225	35.43	36.50	−6.10	−35.11	13.69	0.60±0.2
14	130	35.44	36.40	−5.84	−33.91	12.81	0.91±0.2
15	350	34.99	36.76	−5.13	−28.68	12.36	0.60±0.2
16	280	35.12	36.87	−5.13	−32.63	8.41	0.0±0.2
17	250	35.12	36.94	−5.46	−34.87	8.81	0.60±0.2
18	200	34.96	36.68	−6.20	−34.34	15.26	0.60±0.2
19	200	34.55	36.83	−5.34	−34.02	8.70	0.60±0.2
20	225	35.06	36.72	−5.93	−28.96	18.48	0.60±0.2
21	380	35.00	36.50	−5.90	−29.24	17.96	1.09±0.2
22	190	35.25	36.47	−5.02	−26.45	13.71	0.83±0.2
23	250	35.25	36.47	−4.76	−26.47	11.61	n.m
24	300	34.65	36.65	−5.98	−34.45	13.39	n.m
Homs Depression
25	825	34.87	36.74	−7.86	−46.71	16.17	0.00±0.2
26	1130	34.82	36.8	−8.08	−44.14	20.50	0.00±0.2
27	1100	34.84	36.83	−8.24	−45.63	20.29	0.00±0.2
28	1027	34.88	36.87	−7.96	−49.39	14.29	0.00±0.2
29	1045	34.69	36.99	−7.71	−49.72	11.96	0.60±0.2
30	930	34.64	36.98	−8.67	−55.05	14.31	0.00±0.2
31	488	34.72	37.05	−6.86	−43.66	11.22	0.00±0.2
32	200	34.56	36.96	−6.88	−43.09	11.95	0.60±0.2
33	1380	34.81	36.94	−7.76	−47.53	14.55	0.00±0.2
34	910	35.00	36.97	−7.68	−44.10	17.34	0.00±0.2
35	400	35.01	36.75	−7.47	−42.87	16.89	0.60±0.2
36	1360	34.70	36.86	−8.39	−51.24	15.88	0.00±0.2
37	950	34.65	36.98	−8.74	−52.94	16.98	0.60±0.2
38	250	34.51	36.68	−7.15	−47.52	9.68	0.00±0.2
39	700	34.34	36.76	−8.49	−54.92	13.00	0.00±0.2
40	1659	34.72	36.79	−7.81	−51.29	11.19	0.00±0.2
Note: n.m: not measured.

Table 3 Mean isotopic composition values of the three groundwater clusters of the Upper Cretaceous aquifer in the Orontes basin

Name	δ^18O (‰)	δ^2H (‰)	d (‰)	n
Recharge area	−6.65 ± 0.10	−30.98 ± 2.17	22.2 ± 1.90	5
Hama Uplift	−5.52 ± 0.55	−31.89 ± 3.55	12.3 ± 3.70	19
Homs Depression	−8.01 ± 0.46	−48.78 ± 3.95	15.2 ± 3.16	16
Note: ± is the standard deviation; n: number of samples.

4.2 Stable isotope composition in the groundwater

The stable isotope composition of oxygen and hydrogen is used to determine the origin of the groundwater with respect to the source of precipitation, the altitude of the recharge area and the approximate groundwater age.

The stable isotope composition of modern precipitation was monitored at Bloudan station (1550 m) during the period 2000–2008 and at Kadmous station (750 m) during the period 2005–2008. These two highland stations represent the regional groundwater recharge area of the Anti-Lebanon and Coastal Mountains, respectively (Fig. 1). The δ²H versus δ¹⁸O relationship for all monthly rainwater samples is defined by the following equation (Fig. 6): δ²H = 7.9δ¹⁸O + 20‰ (Al-Charideh 2011a). The stable isotope composition of precipitation in this region is similar to that which defines the Eastern Mediterranean meteoric water line (EMWL: δ²H = 8δ¹⁸O + 22‰, Gat and Carmi 1970), which differs from the global meteoric water line (GMWL: δ²H = 8δ¹⁸O + 10‰, Craig 1961) by a strong deuterium excess (+22‰) due to strong primary evaporation.

Fig. 6 δ¹⁸O–δ²H diagram of groundwater samples from the Upper Cretaceous aquifer in the Orontes basin. The mean and standard deviations of the δ values of the rainwater collected at the monitoring stations Bloudan and Kadmous are also shown.

The results of the isotopic composition determinations of the groundwater samples collected from the study area, together with the deuterium excess (d_ex) and tritium values, are reported in Tables 2 and 3. The plot of the δ²H values versus the δ¹⁸O values of the groundwater samples shows three well distinguished clusters (Fig. 6 and Table 3). The delta values of the groundwater samples from the Coastal Mountains form a cluster on the Local Meteoric Water Line (LMWL), indicating that only westerly storms approaching via the Mediterranean Sea have recharged this groundwater. However, the delta values of groundwater samples from the unconfined aquifer of the Hama Uplift and the deep confined aquifer of the Homs Depression areas form separate clusters between the LMWL and the GMWL (Fig. 6). This means that both groundwaters are recharged by westerly Mediterranean and by northern continental storms, which fit to the GMWL. There is no evidence for isotopic evaporation effects as both clusters fit to regression lines with a slope near 8. There is one exception (well no. 7), which is situated in the unconfined part and contains some tritium. Irrigation return flow might have been admixed because intensive agriculture is practised in this area as shown by the high NO₃ (40 mg/L) concentration.

The groundwater of the confined aquifer section is more depleted in heavy isotopes than that of the unconfined aquifer (Table 3). Two factors may be responsible: (i) the altitude effect: the δ¹⁸O value of rainwater decreases with increasing altitude, and (ii) the palaeoclimatic effect: the δ¹⁸O values of Pleistocene groundwater are by 1.5 to 1.8‰ more negative than those of the Holocene groundwater (Bath et al. 1979). As the difference of 2.5‰ between the means of the two clusters (–5.5‰, Hama; and –8.01‰, Homs) is much larger than 1.8‰, the palaeoclimatic effect alone cannot explain these results. The plot of δ¹⁸O and d_ex versus groundwater ¹⁴C age (Fig. 7) reveals the required information. The calibrated groundwater ¹⁴C ages were calculated with a constant initial ¹⁴C value of 60 pmC (Geyh 1992).

Fig. 7 The δ¹⁸O (blue) and d_ex (green) values plotted vs groundwater ¹⁴C age (A₀ = 60 pmC) form six distinct clusters which are discussed in the text.

Groundwater of the Coastal Mountains is continuously recharged and very young. Its d_ex values of +20 to +24‰ indicate that Mediterranean storms are the source of this groundwater, similar to results for spring waters from both the Anti-Lebanon and Coastal Mountains (Al-Charideh 2011a).

Well no 12 is an exception; the groundwater in the Cretaceous Hama Uplift aquifer is aged between 5000 and 9000 years Before Present (BP). Therefore, the palaeoclimatic effect is excluded. The mean of the δ¹⁸O values of the groundwaters from the Coastal Mountains and the Hama Uplift differ by 1.2‰ (Table 3). This can be attributed to the altitude effect as the Hama Uplift recharge area is about 400 m lower than the Coastal Mountains (Fig. 1), assuming an altitude effect of –0.23‰/100 m as for the Anti-Lebanon Mountains (Al-Charideh and Abou Zakhem 2010). The wide range of the d_ex values (8–18‰) provides evidence that both Mediterranean and continental storms feed the groundwater recharge. It is noteworthy that groundwater well nos. 6, 20 and 21, with d_ex values >18‰ are located in the most western part of the study area. There is a topographic window between the Coastal Mountains and the northern part of the Anti-Lebanon Mountains allowing Mediterranean storms to penetrate directly. Northern continental precipitation has been responsible for the recharge of the groundwater pumped from well nos. 8, 9, 12, 16 and 17 in the Hama Uplift aquifer.

The groundwater in the confined aquifer of the Homs Depression could be hydraulically connected with the open aquifer of the Hama Uplift area although the visible faults may act as barriers (Fig. 2(a)). The groundwater is aged between 17 000 and 24 000 years BP, with exception of that from well no. 39 located close to the Anti-Lebanon Mountains (Fig. 2(a)). The groundwater ¹⁴C ages might be too young if the well catchment area is covered by sediments and an initial ¹⁴C value of around 84 pmC, rather than 60 pmC, is valid. In any case, the palaeoclimatic effect has to be taken into account for the Pleistocene groundwater. However, the difference of 2.5‰ between the means of the corresponding δ¹⁸O values to that of the groundwater from the Hama Uplift aquifer is too large and a complementary isotopic altitude effect must have modified the isotopic composition.

The spatial distributions of the d_ex values suggest that the catchment area of this groundwater might be situated in the northern piedmont of the Anti-Lebanon Mountains. The confined Homs Depression aquifer is close to this piedmont. Both Mediterranean (detected as predominant in samples nos. 26 and 27) and continental storms (predominant in sample nos. 31, 38 and 40) participated in the groundwater recharge. Well no. 39, from which the outlier sample was collected, is situated within the piedmont of the Anti-Lebanon Mountains. The δ¹⁸O value is more negative and the ¹⁴C age is lower than those values of all other samples possibly due to a somewhat higher and locally restricted catchment area and a shorter flow distance.

Table 4 Application of ¹⁴C age correction models for groundwater in the Orontes basin

Site no.	δ^13C (‰)	^14C (%)	Ages	Age correction models:
			A0 = 100%	Tamers	I&P	F&G	IAEA	Evans	Eichinger
Recharge area (Coastal Mts)
1	−11.6	53.4	5200	0.550	−00 100	3000	3950	−400	−700
2	−12.9	80.6	1800	−03 950	−02 600	1800	1700	−2800	−3000
3	−19.9	69.4	3000	−02 700	2200	6500	6400	2100	1900
Hama Uplift (unconfined aquifer)
12	−10.4	9.0	19 900	14 200	13 750	17 500	17 400	13 300	13 300
14	−10.4	23.8	11 850	6100	5700	9500	9500	5300	5200
17	−8.5	22.1	12 500	6750	4600	8300	8200	3950	3900
20	−12.0	17.2	14 600	8850	9500	13 200	13 100	9200	9200
Homs Depression (confined aquifer)
25	−10.3	3.9	26 800	21 100	20 500	23 700	23 600	19 900	20 000
26	−10.5	7.2	21 800	16 000	15 700	18 600	18 500	14 900	15 200
28	−10.4	5.6	23 800	18 100	17 600	20 200	20 100	16 700	17 000
29	−9.8	5.0	24 700	19 000	18 100	21 600	21 500	17 500	17 500
31	−10.6	3.6	27 600	21 900	21 500	25 200	25 100	21 100	21 000
33	−7.2	6.5	22 600	16 900	13 400	16 300	16 000	11 800	12 100
34	−6.9	4.6	25 500	19 800	15 800	18 800	18 400	14 000	14 400
39	−10.5	24.6	11 600	5900	5500	9300	9200	5100	5000
40	−8.7	5.9	23 400	17 700	15 800	18 500	18 400	14 600	14 900
Note: Tamers: Tamers (1975); I&P: Ingerson and Pearson (1964); F&G: Fontes and Garnier (1979); IAEA: Salem et al. (1980); Evans: Evans et al. (1979); Eichinger: Eichinger (1983).

4.3 Tritium and ¹⁴C residence time

Tritium values help to identify the presence of modern groundwater in samples which might have been recharged or admixed after about 1960. Admixture may be due to leakage through the confining bed, hydraulic shortcuts in the aquitards or defects in the well casings.

Tritium values of between 2 to 4 TU for the groundwater samples from the Coastal Mountains correspond to mean residence times of several decades. This is in agreement with those of the spring water discharging from the high Anti-Lebanon and Coastal Mountains (Al-Charideh 2011a). The majority of groundwater samples from the unconfined aquifer of the Hama Uplift contain tritium at between 2 TU and the detection limit of <0.4 TU, evidence that some groundwater recharge is occurring, and have higher mean residence times of 100 years, or even more. Most groundwater samples from the confined aquifer of Homs Depression had tritium values below the detection limit, as expected for Pleistocene groundwater in a confined aquifer. However, well nos. 29, 32, 35 and 37 (all 0.6 TU) might contain some modern groundwater.

Radiocarbon activities and δ¹³C values are commonly used to identify carbon origin in groundwater. In fact, atmospheric waters are “marked” by dissolution/precipitation processes taking place during water–rock interactions (Clark and Fritz 1997).

The δ¹³C contents of the TDIC ranges from –19.9 to –11.6‰ versus PDB for the recharge area, and from –12 to –6.9‰ for the unconfined and confined aquifer system of Hama Uplift and Homs Depression (Table 4). The ¹⁴C activities vary widely from 80.6 to 4% pmC. The highest values were measured in the recharge area which reveals the existence of modern recharge. The ¹⁴C activities decrease rapidly moving from the unconfined to the confined aquifer, ranging between 23 to 9 pmC in samples from the unconfined aquifer of Hama Uplift, to very low, between 7 and 4% pmC, in the confined aquifer which implies long water residence time (Table 4). Data indicate that groundwater is evolving from recent water with high ¹⁴C activities and low δ¹³C values towards older groundwater relatively enriched in ¹³C. This argument relates to different water–rock interaction conditions enhanced by different groundwater residence times (Fig. 8). Longer water residence times enhance equilibration with aquifer carbonate minerals resulting in lower ¹⁴C activity and more enriched δ¹³C compositions (Bath et al. 1979, Clark and Fritz 1997, Mook and De Vries 2000). As mentioned above, the several depleted values of δ¹³C observed in the confined aquifer could be explained by a larger contribution of organic matter resulting in a δ¹³C depleted source of carbon (Edmunds et al. 1995), or by different recharge regimes such as palaeo-recharge (Zourai et al. 2011).

Fig. 8 Evolution of ¹⁴C and δ¹³C values in the groundwater samples from the Upper Cretaceous aquifer in the study area.

In order to assess the groundwater residence time, different classic correction models of the initial activity within a carbonate aquifer have been used. These models are based on different mixing processes: models of pure chemical and isotopic mixing (Ingerson and Pearson 1964, Tamers 1975), and models of chemical mixing and isotopic exchange (Evans et al. 1979, Fontes and Garnier 1979, Salem et al. 1980, Eichinger 1983). The signatures for different carbon sources introduced in these calculations for our study have the following isotopic characteristics:

a.	for biogenic carbon δ13Csoil = –21‰ vs V-PDB and 14C = 100% pmC
b.	for carbonate matrix δ13Ccarb = 0‰ vs V-PDB and 14C = 0% pmC

Excluding the Tamers model, the ¹⁴C modelling indicates that the results of the five C¹⁴ correction models can be divided into two groups according to groundwater age range (Table 4). The first group includes the Pearson (Ingerson and Pearson 1964), Evans (Evans et al. 1979) and Eichinger (1983) models which are based on isotope and chemical exchanges. The second group, from the IAEA (Salem et al. 1980) and Fontes and Garnier (1979) models based on isotopic exchanges, lead to groundwater ages much greater than the first group; a maximum difference of 4000 years is observed. Consequently, adoption of the maximum range of the actual uncertainty of the different models into the error propagation leads to standard deviations of the calibrated groundwater ¹⁴C ages of ±5000. Therefore the spread of the calibrated ages is still within the actual confidence interval (Fig. 9). The Tamers model gives results similar to the first group for ages less than 15 000 years BP with a maximum difference of 3000 years less than the second group. However, this chemical model cannot be considered in this case study since the δ¹³C contents vary with the ¹⁴C activities (Fig. 8), which suggests an exchange between TDIC and the matrix of the carbonate aquifer through time.

Fig. 9 Relationship between δ¹⁸O and corrected age values for groundwater in the study area. Standard deviation shown by ± “whiskers” (5000 years for ¹⁴C age and 0.1‰ for δ¹⁸O).

The IAEA and Fontes and Garnier models suggest exaggerated calibrated ¹⁴C ages for groundwater in the recharge area and the unconfined aquifer of Hama Uplift relative to those indicated by their stable isotope composition and tritium contents. Thus, in this case study, the first group of models (Pearson, Evans and Eichinger) seems to be more appropriate to adopt. Pearson ages are present day for the group in the recharge area, between 5000 and 13 750 years BP for the unconfined Cretaceous aquifer of Hama Uplift, and reach 21 500 years BP for the confined part of this aquifer in the Homs Depression.

Based on this correction, the calculated age indicates three groundwater recharge periods (Fig. 9):

a.	Present-day recharge with δ18O similar to the modern rainwater, between –6.5 and –6.8‰, and 14C activities of 50–80 pmC. This water represents the recharge area of Coastal Mountains. Based on the stable and radioactive isotopic data, it is suggested that the present recharge continues to function, notably in the west, where the aquifer outcrops.
b.	Holocene groundwater recharge, where the corrected age range is 5000 to 13 000 years BP. The groundwater has enriched δ18O values (–5.46 to –5.93‰) relative to modern groundwater. The shift in δ18O to higher values signifies warmer climatic conditions at the beginning of the Holocene (Bath et al. 1979). The groundwater of this group is found in the unconfined part of the Upper Cretaceous aquifer of the Hama Uplift.
c.	Late Pleistocene groundwater recharge, where the corrected age ranges between 10 000 and 22 000 years BP. The groundwater has depleted δ18O values (–7.96 to –8.67‰) and low radiocarbon activities (3.9–7.2 pmC), which represent the groundwater of the confined aquifer of Homs Depression. It is suggest that the residence time of the water in the confined Cretaceous aquifer is considerably longer than that in the unconfined area.

The spatial distribution of the calibrated groundwater ¹⁴C ages calculated with any of the models used (Table 4) does not allow determination of the groundwater flow direction or estimation of the groundwater velocity. This may be due to the locally variable mixture of different old groundwaters flowing in various non-homogeneously distributed wide flow paths in the hard-rock aquifers. Another reason may be that groundwater is usually pumped only from a restricted section of the total water column of the aquifer, which differs from well to well.

In fact, these results of low stable isotope composition and ¹⁴C activities in the deep groundwater of the Upper Cretaceous aquifer in the Homs Depression are similar to those for the Upper Cretaceous aquifer in the Syrian Desert (Ferronsky et al. 1991), the Upper Jezireh in northeast Syria (Al-Charideh and Abou Zakhem 2009) and the Aleppo basin in northern Syria (Al-Charideh 2011b). According to the palaeoclimatic model of the Syrian Desert (Ferronsky et al. 1991), the period from 20 000 to 10 000 years BP can be characterized as being the climatic temperature minimum, when annual average temperatures were lower than modern ones by about 6°C. The main recharge events could have occurred and formed the major reservoir of groundwater in the Upper Cretaceous aquifer in Syria, in this period. Based on the similarity of the δ¹⁸O values and ¹⁴C activities in the deep groundwater from the Homs Depression, Aleppo basin, Syrian Desert and Upper Jezireh, one might suggest simultaneous recharge events. However, given the uncertainties regarding the contribution of ¹⁴C-free carbon by different factors, this inference should be considered with caution.

5 CONCLUSION

The chemical and isotopic characteristics of the groundwater in the Upper Cretaceous aquifers of the Orontes basin have revealed many important aspects of groundwater evolution in aquifers that have been created from carbonate rocks.

The hydrochemistry of groundwater from the Upper Cretaceous aquifer system varies in relation to different water–rock interaction processes. The dominant anions are HCO₃ ⁻ for fresh water and SO₄ ^2- and Cl⁻ for the more saline waters. The geochemical processes generating the chemical composition of groundwater in the confined aquifer of the Homs Depression indicate a long-lasting water–rock interaction probably dominated by dissolution of halite and gypsum, and cation exchange of varying intensity could explain most of the observed spatial variability of the chemical groundwater composition.

The δ²H and δ¹⁸O results confirm this hydrogeological concept in general and improve the details. The groundwater in the Coastal Mountains is solely recharged by Mediterranean precipitation. The groundwater in the unconfined Hama Uplift aquifer and the confined Homs Depression aquifer is fed by the same precipitation but also recharged by continental rainfall further to the east. The groundwater of the unconfined Hama Uplift aquifer is recharged locally. Most groundwater in the confined Homs Depression aquifer might be recharged in the northern piedmont of the Anti-Lebanon Mountains at an altitude around 1900 m, though an inflow from the Hama Uplift aquifer cannot be excluded in the northeast. The lower recharge rate of the groundwater in the Hama Uplift aquifer, and even slower recharge of the Homs Depression aquifer is reflected by groundwater ¹⁴C ages of more than 5000 and over 22 000 years, respectively. The dispersed distribution of flow paths in the hard rock aquifers prevents determination of the flow velocity and estimation of the groundwater model ages. Another reason may be pumping from sections that deviate from the total aquifer. The calibrated groundwater ¹⁴C ages obtained using one of the models applied might be underestimated due to an incorrect choice of any of the parameters used, especially the initial δ¹³C and ¹⁴C values of the soil CO₂ and soil carbonate.

Finally, in semi-arid regions such as the present area, where there are prolonged dry periods associated with little recharge and a high demand for groundwater for irrigation, environmental impact arises due to groundwater use. In this case, the extraction of groundwater appears to be exploitation of a non-renewable resource. Even if some recharge is taking place, it is not sufficient in volumetric terms to resolve the environmental problems. It is strongly recommended to improve and develop sustainable water resources management by using modern irrigation methods.

Acknowledgements

The author would like to thank Prof. I. Othman, Director General of AECS for use of their facilities during this study. We are also grateful to M. Geyh, W.R. Agha and S. Rammah for their helpful comments and useful discussion. The author thanks Z. Kattan, Head of the Geology Department at the AECS, and the staff of the laboratories for their cooperation in performing the isotopic and chemical analyses.