The consequences of climate change on water resources is due to the alteration of the seasonal distribution of precipitation, temperature increase and the change in evapotranspiration, as a result of the change on vegetation cover. All these changes affect the recharge regime of groundwater systems. Long-term droughts may result in storage reduction and in groundwater discharge to streams and other surface water bodies. Therefore, if the effects of the droughts are to be mitigated by the use of groundwater resources, it is essential to establish in advance an effective management plan. Karst aquifers are developed primarily by the use of springs and galleries, because in most cases the system is generally too complex to locate a proper borehole on the right place. There are several methods of spring development in karst aquifers. Springs are either captured or regulated for development. Practices of karst spring capture include installing pumps in dissolution channels below spring elevation. Horizontal galleries or large deep wells that reach a phreatic conduit are among the best methods. Numerous examples are documented in the literature [11, 12].
In contrast to capture water at the possible lowest elevation, management of karst aquifers under climate change should consider raising the discharge elevation by capturing the flush waters of floods and thus store more water for later use during the long-term droughts. However, the shorter the residence times the lower the possibility of availability of the flood water. This hydraulic behavior is reflected in a high recession coefficient and low regulation power. According to the conceptual scheme suggested by Hobbs and Smart , karst aquifers with point recharge, low storage and conduit type flow are considered to be hypersensitive to external factors. They behave like surface waters rather than groundwater systems. This is also true for upper sections of large karst rock masses where karst is better developed than the deeper parts. Shallow karst may either be the subcutaneous (epikarst) zone or the vadose and immediately underlying epiphreatic zone.
Under normal climatic conditions, the relatively small size, well karstified carbonate rock masses with shallow karst base are not considered as reliable aquifers in the sense of sustainable use. This is mainly because of the rapid transit time (short residence time) and the lack of a permanent and large volume of phreatic (saturated) zone. The carbonate rocks where karst morphology is well developed, usually having high recession coefficient, low regulation power, and where the flow is dominantly of conduit type, and the recharge is overwhelmingly point input type, are found in specific karst types. In northeastern Central Anatolian region, where evolutionary karst has been developed, dissected/relict karst is common. Rock masses of various sizes, from less than a kilometer to several tens of kilometers, with shallow karst base, occur in this region where people generally suffer of lack of adequate water. Similarly, karst in rock blocks of allocthonous carbonate units of various sizes is very well developed and the karstification base is generally marked by the underlying non-karstic units at shallow depths.
Because of the characteristics outlined above, these blocks are considered to be less reliable as a water supply source. Allocthonous carbonate rock masses are scattered at various regions of Turkey, but mostly at the southern and western parts. On the contrary, authocthonous carbonate units form high yield karst aquifers owing to their large size, large thickness, and deep karstification base, generally lower than the sea level. Thick phreatic zone provides a significantly large storage capacity and long residence times, making the aquifer reliable for water supply. Thick to very thick vadose zones are common in the recharge areas of these aquifers in the upland regions. Significant portion of the water stored in this type of aquifers are discharged through karst springs close or beneath the sea level. Therefore, it is hard to get groundwater at the highlands of the authocthonous carbonate rocks, except at mostly seasonal small springs of shallow circulation which are normally related to a local karstification base. These intermittent springs with high recession coefficient and low regulation power are regarded as unreliable for water supply.
Based on their hydraulic characteristics, these carbonate rock masses are extremely sensitive according to the conceptual scheme of Hobbs and Smart  when they are regarded as aquifers. With the expected adverse affects of the climate changes, their sustainability even worsen. However, it is possible to utilize these well developed karst systems to mitigate the adverse effect of seasonal precipitation distribution due to climate change. This can be achieved by elevating the discharge level of the aquifer, which will help to capture and store the flood waters for later use.
Construction of cutoff wall in front of the lowest discharge point (or zone) to raise the water level in the aquifer is one of the techniques that proved to be effective (Fig. 9.1). This technique can be applied for capturing flood waters coming rapidly from infiltration in dolines and sinking streams trough a well developed karst system inside the allocthonous carbonate rock mass. It is also efficient to tap the water of highland springs discharging at a local karstification base of a thick authocthonous carbonate rock aquifer.
Another technique is to construct a cutoff wall all around a relict karst block by for instance, shotcrete application (Fig. 9.2). Due to well developed karst, the specific yield of these aquifers is very high. So, a cutoff wall to raise the groundwater level only about 2-3 m is sufficient to capture a significant amount of flood water that otherwise would be discharged rapidly from the groundwater system.
This method of storing flood water in a karst aquifer is more advantageous than surface reservoirs, at least in decreasing the evaporation loss during long-term droughts.
The method has proved to be effective in karst aquifers of allocthonous carbonate rocks, as shown by a case study in Elmali, Antalya, Turkey. The Kazanpinari karst springs discharge along about 500 m with a total flow rate of about 2.3 m3/s. The recession coefficient is calculated as 0.004 day-1. These springs function as spills of the aquifer and full flow type of springs. Pumping tests in wells drilled in this aquifer revealed a transmissivity coefficient ranging between 500 and 20,300 m2/day, a large interval indicating a very high heterogeneity.
The flow rate of the springs was measured and ranges between 0.2 and 5.7 m3/s. The variation of flow is about 238% and highly variable according to Meinzer's variability index. The discharge level of the springs was first lowered and the increase in flow rate with time was recorded together with the decline in ground-water level in the aquifer. The storage coefficient (which is the specific yield in this case) was calculated for the aquifer using the relationship between the water level and the amount released from the aquifer. This was later used to compute the amount of water that can be stored by raising the water level in the aquifer by elevating the discharge level of the springs . Construction of a cutoff wall in front of the spring zone caused a rise of about 2.5 m in the aquifer. This average rise in groundwater corresponds to an increase of about 5x106 m3. This means that an amount of flood water can be stored in the aquifer by raising the discharge level of the springs. Moreover, a gate at lower elevation than the actual level of the springs can be constructed to regulate the flow rate.
Different from the Kazanpinari karst springs which are a full flow type, the Kirkgoz karst springs are sort of overflow of the aquifer. They discharge at an elevation of 300 m above sea level at the edge of a large travertine plateau. The mean discharge is about 13 m3/s, and the minimum and maximum rates are recorded as about 5 m3/s and above 25 m3/s respectively. The springs have a discharge variability of more than 200% according to Meinzer's index. The recession coefficient is around 0.0035 day-1. The transmissivity of the aquifer is calculated to vary between 500 and 5,000 m2/day. The specific yield of the karst aquifer is calculated from the observations of groundwater levels and the amount of water abstracted from the
Fig. 9.3 Capture and storage of flood waters in the Kirkgoz karst aquifer non-karst
Fig. 9.3 Capture and storage of flood waters in the Kirkgoz karst aquifer reservoir . The spring water forms a large wetland in front of the orifices along a 1 km long zone. Naturally, the wetland water is discharged into a sinkhole through a small stream. A project to raise the water level in the wetland has been planned but not implemented yet. Hydraulic calculations have shown that about 2.5 m raise in groundwater level in the aquifer would allow extra storage of about 65x106 m3 of water in the aquifer. However, because the karst base of the aquifer is far below the elevation of the springs, an increase of about 2.5 m head will also increase the discharge of springs at lower elevations in the travertine plateau, including the submarine springs (Fig. 9.3). It is most likely that further storage will be possible at the travertine aquifer which supplies drinking water to about one million of people in the Antalya city. Flush water during flood periods will be captured and stored in the aquifer for later use during long-term droughts.
In spite of its efficiency in terms of sustainable use of an un-sustained ground-water resource, the impact of this capturing flood water technique on the existing ecosystem should be studied prior to this practice becomes common.
Was this article helpful?