Flash floods are caused by heavy rainfall that
has become more frequent. They are more prominent in low-storage karst
regions, although karst terrain often acts as a natural flood control particularly
when it is bare and dominated by conduits. A study using a
hydrogeochemical approach and assessing data from several springs in
different carbonate rock in western Turkey has made it possible to classify
karst aquifers based on their response to heavy rainfall events. According to
this aim, physico-chemical measurements in wet and dry seasons and discharge rates
in springs are compared in order to explain aquifer characteristics.
Groundwater samples have a pH ranging from 6.3 to 8.9, temperature (
Flash floods are deadly and costly natural disasters worldwide. A special kind of flash floods is the karst flash flood. Karst aquifers provide 20–25 % of high-quality groundwater for drinking water in the world. Therefore, they are particularly important for water management. Karst terrain are fragile and vulnerable to a variety of hazards (Parise, 2015), including flash floods due to heavy rainfall. They occur when intensive short-duration rainfall is recorded, such as 571 mm on 6 December 2004 in Sardinia, Italy (Cossu et al., 2007), or 200.5 mm on 12 September 2009 in Ayamama Creek, Istanbul, Turkey (Bacanli et al., 2011). Flash floods are intensified when heavy rainfall is combined with rapid snowmelt or man-made disasters (e.g. dam failure) on steep terrain with excessive antecedent precipitation and thin, bare and/or impermeable soil or covers (Bonacci et al., 2006).
Due to characteristics of groundwater flow in karst terrain, flash floods are strongly different from those in non-karst terrain. Karst terrain can often be considered as natural flood control, especially when it is bare and dominated by conduits. However, flash floods can become prominent in low-storage karst regions. Furthermore, karst groundwater watersheds often do not coincide with topographic divides, to indicate just one of the main differences between karst hydrology and other environments (Parise and Gunn, 2007; Parise et al., 2015). Delineation of drainage basins and drainage divides is complicated by the overlap of catchment areas where perched vadose flow crosses phreatic divides, and by the shifting of divides with time as the flow stage varies (Palmer, 2010). Divergent flow and disjunct contributory areas provide further complexity (Gunn, 2007).
Recharge of karst aquifers is described as infiltration – inflow to increase the groundwater level, and deep inflow to create a groundwater reservoir. Infiltration has two different forms: areal or spatial diffusive infiltration flows, passing directly into the groundwater reservoir. The spatial groundwater recharge process is diffusive, in which water reaches the groundwater table as a laminar flow through the intergranular unfractured bedrock and fractures (Shuster and White, 1971). The other form of recharge (point infiltration) passes via karst structures, such as caves and sinkholes. It depends on the pipe-like structures (conduits) in which flow is dominantly turbulent. The minimum diameter of karst conduits or fractures in which turbulent flow could exist should be greater than 5 to 15 mm (Bonacci et al., 2006).
Duration, intensity and type of precipitation are important factors which affect the recharge directly. The altitude of the region under consideration is also considerably important (Sutcliffe, 2004). In karst terrain, point infiltration and sudden rainfall are correspondingly effective. Studies indicate that less than 5 % of annual effective rainfall becomes groundwater recharge in non-karst regions, whereas recharge to karst aquifers is in excess of 80 % of effective rainfall. For instance, in karst regions in Saudi Arabia, 47 % of the average rainfall disappears at the intersections of dolines and cracks (Milanovic, 1981). According to De Vries and Simmers (2002), in Portugal, a dry and hot summer-climate, 150–300 mm (27–55 %) of the 550 mm annual precipitation infiltrates in karstic dolomite and marble regions. Similar values are reported for karstic aquifers in Israel, a Mediterranean climate (Issar and Passhier, 1990). In different regions of former Yugoslavia, infiltration is calculated to be between 70 and 90 % (Milanovic, 1981). In Turkey, in Tecer limestone (Sivas-Ulaş region in eastern central Anatolia), infiltration has a percentage as high as 55 % (Ekemen and Kacaroglu, 2001). The annual effective infiltration of the Gradole catchment in Croatia is between 36 and 76 % (Bonacci, 2001), while it is between 60 and 90 % for the mountainous karst regions in Switzerland (Malard et al., 2016). Similarly, in the Maiella aquifer in Italy, it is 900 mm (62 % of 1450 mm of total annual rainfall) (Fiorillo et al., 2015b).
In general, the frequency or intensity of heavy rainfall events have increased over the midlatitude land areas of the Northern Hemisphere since 1901 (with a medium confidence before and a high confidence after 1951) (IPCC, 2013). According to the country report presented by the Turkish government and the United Nations Development Programme (UNDP), precipitation decreases along the Aegean and Mediterranean coasts and increases along the Black Sea coast of Turkey. Central Anatolia shows little or no change in precipitation (OECD, 2013). The most prominent result of the climate change data is that the number of days with excessive precipitation has been increasing in Turkey, which usually causes extreme floods. Even in regions where annual rainfall is decreasing, there is a trend for a higher amount of rain that reaches the ground in heavy downpours (CCSP, 2008; Karl et al., 2009).
Even if cumulative precipitation has increased, the groundwater recharge may have been less in diffuse flow-dominated aquifers (Williams and Lee, 2008). In other words, the same amount of rainfall does not recharge aquifers as in the past because of change in rainfall intensity and its type as stated by Fiorillo et al. (2015b) who explained this response by an increase in evapotranspiration due to an increase in temperature. Unlike other aquifers, vertically oriented karst features collect and quickly transport water from the surface to the underground fluvial system. Because high-intensity rainfall is able to infiltrate fast enough before evaporating (Williams and Lee, 2008), practically all rainfall quickly penetrates into the karst underground system, where it fills karst voids of different dimensions, and flows under the effect of gravity (Bonacci et al., 2006). In these systems, flow primarily occurs rapidly in large fissures through irregular conduits known as quick flow or fast flow systems. Quick flow comprises stream sink water and percolation from closed depressions, while baseflow includes slow percolation from areas not drained by depressions (Fiorillo, 2009). Therefore, it can be noticed and expected that higher recharge in limestone dominated by conduits is due to heavy rainfall. However, it should be mentioned that an accurate and reliable estimation of recharge into karst aquifers is difficult because of their heterogeneous local structures (Bonacci, 2001; De Waele and Parise, 2013; Fiorillo et al., 2015a; Bakalowicz, 2015).
Developed karst sinkholes allow fast percolation of up to 80 % of heavy rainfall into the aquifer. However, the very low storage combined with the high transmissivity means that most of the recharge will not be retained by the karst system, but will rapidly flow out (to springs, rivers, lakes, sea). The fast response of water level to the rainfall combined with the capability of karst systems to transmit groundwater increases flooding quickly (Williams and Lee, 2008).
Location map of the study area and springs.
Numerous studies have been conducted to determine aquifer characteristics of the discharging springs using their physico-chemical parameters, since each reservoir will have a characteristic chemical signature (Aquilina et al., 2005). Chemical characteristics of an aquifer are functions of residence time and flow conditions in the aquifer (Freeze and Cherry, 1979). Changes in the quantity and quality of the discharge of karst springs are largely caused by recharge parameters in the recharge area, such as rainfall and snowmelt (i.e. autogenic recharge), and point infiltration of streams (i.e. allogenic recharge) as well as lithology, hydraulic conductivity and residence time of aquifers that also affect changes. As a matter of fact, Shuster and White (1971), Atkinson (1977), Aydin (2005) and Demiroglu (2008) argued that the reason for the change observed in the spring discharges was the recharge type or flow type (diffuse or conduit) with high or low storage. A high number of studies have been done to define aquifer characteristics using physico-chemical measurements (Shuster and White, 1971; Raeisi and Karami, 1997; Massei et al., 2007; Raynaud et al., 2015) by considering only one or a few physico-chemical parameters that characterize groundwater, which has the potential to lead to an erroneous interpretation of the hydrodynamics of the karstic catchment draining into the spring (Chicanoa et al., 2001; Massei et al., 2007).
This study aims to classify karst aquifers based on their response to heavy rainfall events, by using a hydrogeochemical approach, assessing data from several springs in different carbonate rock in western Turkey. With this approach, it is expected that flood-prone areas can be defined on the basis of lithological and structural features of karst terrain.
Karst aquifers with Paleozoic marble, Mesozoic limestone and Neocene limestone were chosen from western Turkey to understand aquifer characteristics for a possible classification in terms of their responses to heavy rainfall. All springs are shown in Fig. 1, with coordinates and elevations given in Table 1. Karst features of western Turkey demonstrate the tectonic, lithologic and climatic specificity underground flow movement, and chemical characteristics of groundwater (Gunay et al., 2015).
Location of springs.
Springs are organized in three groups. The first group of springs
(S1 Döskaya, S2 Nardin), located in the central Sakarya basin (Fig. 2), recharge
and discharge from Jurassic Bilecik limestone. In this area, the Harmankoy–Beyyayla
Karst System (HBKS), studied by Aydin (2005), forms the highlands
in the Central Sakarya Basin. The HBKS is located in Bilecik and Eskişehir
and extends over a surface area of 49.5 km
The second group of springs is located in Çifteler and Günyüzü subbasins
within Sakarya basin, Sivrihisar-Eskişehir (Fig. 3). Sakarbasi and Kaymaz
springs (S3, S4) were studied by Guner and Guner (2002), who determined
little or no change in the in situ measurements during three seasons of
sampling. The reservoir rock of Sakarbasi springs (S3; Sadıroğlu, Eminekin,
Başkurt, Ilicabasi and Pınarbaşı) are composed of marble. This is named the
Gökçeyayla formation, which is primarily shelf-type carbonate that is
deposited during the Triassic–Upper Cretaceous. Dolomitic limestone is
dominant in the lower section of the unit, while the upper section is mostly
chert limestone (Gunay, 2006). The total groundwater discharge
(233
Günyüzü springs (S5–S10), located in Sakarya basin, were studied by Demiroglu (2008)
(Fig. 3). Paleozoic marble, which is the main reservoir rock for hot
and cold water, is bordered by impermeable diabase dykes at the sides and
by impermeable granites and schists. Marble, at the top of the metamorphic
series, at higher elevations of the basin, represents the upper part of the
aquifer system. The S9 (Çukurçeşme) spring recharges, circulates and discharges
from this system. This shallow circulation of water has laminar flow
conditions (Demiroglu, 2008). Other shallow circulating waters (S5, Babadat
and S6, Nasrettin Hoca) mostly recharge from the marble but discharge from
Neogene units. Marble at the bottom of the basin with faults recharges and
stores deep circulating water, where fracture permeability and diffuse
infiltration (laminar flow conditions) control groundwater flow (S7, Çardak
Hamami, S8, Subaşı and S10, Yeniçıkrı). This deep circulating water discharges
from Neogene units as well. However, partly developed conduit permeability
and point infiltration from old karstic structures (sinkholes), fractures
and joints in the marble reveal a turbulent regime in the vadose zone
(Demiroglu, 2008). The groundwater total recharge and discharge were
calculated as 30
Measurements, observations and calculations given for the above examples show that discharge from aquifers is always higher than the recharge into the aquifer. The difference is usually assumed to be balanced by contribution of neighbouring subbasins. However, this is not the only way to explain the higher recharge. A more important reason, particularly in karst aquifers dominated by point infiltration, is the fast response of the aquifer to heavy rainfall that is expected to become more frequent in the future. This can be linked to the climate change, for which spring discharge is found to be a robust indicator (Fiorillo et al., 2015b).
The third group of springs is located in the Lake District (S10–S20 and Sütçüler springs) (Fig. 4). Mesozoic limestone is the most common unit around the lake, both autochthonous and allochthonous in origin. The allochthonous limestone makes up the major part of the Lycian Nappes. Middle–Upper Triassic dolomitic limestone, containing different lithofacies from thin layers to medium thick bedded levels, is the oldest part of the Mesozoic series. Jurassic-Cretaceous limestone contains marine sediments from a deep to shallow marine environment. Thinly bedded chert micritic levels are also observed. Allochthonous nappes thrust over their autochthonous units in the form of tectonic slices. The autochthonous limestone is situated to the southeast of the Lycian overthrusted front. Slices show the ophiolitic melange feature at the bottom of this structure (Ekmekçi, 2005; Davraz et al., 2008). Important springs are located mostly along the tectonic contact between overthrusted limestone formations and impervious hydrogeological barriers (Gunay et al., 2015).
Kirkgozler springs (S19, S20) are very important for the Antalya region in the southwest of Turkey in supplying drinking water and for irrigation. They discharge from the boundary between the autochthonous limestone of the Beydağlari mountains and the impermeable ophiolite rock (Fig. 4) located at 300 m altitude. They are characterized by a highly regulated flow regime (Ozyurt, 2008).
Also a historical spring, Sagalassos (S18) exists in the Lake District and discharges from the deeply fractured allochthonous Cretaceous limestone. Sagalassos settlements from Roman times, with different types of infrastructures for the collection, transport and distribution of water, are worthy of being reevaluated.
Other examples in the Lake District are Sütçüler small springs. There are no regular yield and water chemistry measurements for these springs (Fig. 5). Sütçüler, a town located on the allochthonous limestone, faced flood disasters in November 1995 and October 2011. The Sütçüler small springs are activated by a sudden rise in groundwater. Here, heavy rainfall is combined with the capability of the karst system. During karst flash floods, a sudden rise in the groundwater level occurs, which causes the appearance of numerous, unexpected, abundant and temporary karst springs (Bonacci et al., 2006). The area was studied by Karaguzel and Akinci (1998) after flooding. The ophiolite complex consists of ultrabasic and sedimentary rock at the base, with the Jurassic–Cretaceous carbonate mass spread over a wide area and overlapping the ophiolitic rock. Erenler Mountain, located in the north of Sütçüler, is composed of limestone and has developed karstic structures (sinkholes, dolines) on the ophiolite complex. Plio-Quaternary travertine is exposed in the town; sand, gravel and block-size debris pile up at the base of steep slopes (Fig. 5). The town centre is mainly founded on the travertine and ophiolite complex. There is no permanent surface water flowing in the region. Springs discharge from the limestone-ophiolite contact.
Representative and organized sampling of springs in Günyüzü basin, Eskişehir, Turkey, indicated that karst groundwater characteristics could be defined and used for the classification of karst aquifers regarding their contribution in generating flash floods (Demiroglu, 2008). In this manner, data related to the study area as listed in Table 1 were compiled from previous studies (Atilla, 1996; Aydin, 2005; Demiroglu, 2008) and from State Hydraulic Works, DSI of Turkey. It can be followed from Table 2 that some springs (S3, S9, S10) recharge, circulate and discharge from Paleozoic marble, some (S1, S2, S4, S15, S16, S17, S18, S19, S20) from Mesozoic limestone, while others (S4, S5, S6, S7, S8) recharge and circulate in marble, and circulate and discharge from Neogene limestone and clastics.
In situ groundwater hydrochemical measurements are important to identify the origin of groundwater and flow regimes, especially in karst regions. Shuster and White (1971) recorded that some springs have no seasonal fluctuations in their temperature and total hardness, while others show continuous seasonal fluctuation in spite of the fact that all springs are in similar geological environments. The fluctuations have been reduced to a single number for each spring by calculating the coefficient of variation (CV) of the total hardness. Further, the measure of electrical conductivity (EC) has widely been used instead of hardness to classify karst springs as a useful and practical tool (White, 2002). However, aquifers are combinations of diffuse and conduit systems (or slow and fast flows) due to the triple permeability (matrix, fracture and conduit). Lower variability in hydrochemical characteristics can be observed not only in diffuse flow but also in conduit flow (Atkinson, 1977). This means that the aquifer either has a diffuse-infiltration-dominated recharge or it is a conduit flow aquifer with large storage, showing almost no response to allogenic recharge. Massei et al. (2007) suggested that the different modalities of the specific conductance frequency distributions of karst spring discharge reflect the movement of geochemically distinct masses of water through the aquifer and that the mean specific conductance of an individual water mass, or type, depends on its origin and residence time.
Hydrogeological data of springs.
Physico-chemical data of springs.
In this study, physico-chemical analysis of ground water is used for the
classification of aquifers. The relationship between electrical conductivity (EC),
temperature (
Data compiled for the study area were analysed based on the proposed approach for the three groups of springs as follows.
As the signature of karst aquifers, in situ measurements in wet and dry
seasons are first separated in order to evaluate the seasonal variability in
water geochemistry and dilution effect. As demonstrated in Table 3, the
groundwater has pH values ranging from 6.38 to 8.9, temperature from 5 to
34.8
For the first group of springs (S1 and S2 in Bilecik area), 2 years of
in situ measurements in dry and wet seasons are given in Table 3, from which
fluctuations in the physico-chemical data can be noticed. Discharge rates and
the CV of physico-chemical data are given in Table 4. Change in discharge is
within 2 orders of magnitudes ranging from 0 to almost 400 L s
EC measurements show that variations in physico-chemical data depend not only
on circulation depth and residence time but also on lithology. For example,
springs S3 and S8 (Table 3) nearly have the same temperature and DO
(26.6–30
Average Ca and EC were measured to be 3 meq L
Hyetograph, hydrograph and cumulative departure of the Subaşı spring (S8) (vertical axis on the left is used for both rainfall and discharge, vertical axis on the right for cumulative departure only).
Discharge characteristics of springs and variations in physico-chemical data.
Table 3 shows that the second group of springs (S3–S10) displays nearly
constant temperature, low variations for the measurements both in dry and
wet seasons and low discharge rates (
As in the example of Subaşı spring (S8), aquifers dominantly controlled by diffused groundwater flow can also cause inter-catchment overflow and redistribution of catchment areas at very high groundwater levels because of fossil and inactive conduits and spring activation in vadose zone (Bonacci et al., 2006). It is directly linked to the structure and hydraulic properties of karst aquifers (Fleury et al., 2013). Therefore, karst structures should be taken into account as a component of the hydrological budget of the watershed to avoid the unexpected, uncalculated additional water coming from neighbouring watersheds. For example, Eris and Wittenberg (2015) showed that water transfer between neighbouring karstic watersheds in Mediterranean Turkey was considerable. Furthermore, Parise (2003) and Gutierrez et al. (2014) emphasized the influence to karst floods by clogged swallow holes, new sinkholes consequential to high rainfall rates and alterations to the natural system of caves.
Data of the first and second groups of springs in Bilecik and Eskişehir can
be considered enough for a reliable prediction of aquifer properties.
However, this is not the case for the Lake District springs, except for the
well-studied Kirkgozler springs. Therefore, the Lake District needs further
organized sampling. The Kirkgozler springs (S19, S20), the historical
Sagalassos spring (S18) and Sütçüler springs in the Lake District are
discussed below. It can be said that Lake District aquifers are dominated by
conduit permeability with low storage, except for S17 and Kirkgozler
springs (S19, S20) when the dry and wet season in situ measurements of the springs
are considered (Tables 3 and 4). Temperature changes in S13 slightly.
However, EC values are higher in the wet season than in the dry season
because the aquifer is recharged by Miocene sediments. For this reason,
additional data are required for spring S13 to classify it as either conduit-
or diffuse-flow-dominated. Another aquifer, S17, with developed karst
structures, shows slight differences in
Hydrograph for fast flow springs:
Another exceptional example is the Pınargözü spring (S16). Its discharge has
a high variation that can be tripled or sharply lowered by one-third. For
example, the discharge 887 L s
The Sagalassos spring (S18) also shows rapid changes in measurements in dry and wet seasons (Table 3), which is a good example of a high response capability to heavy rainfall taking place in the ancient city. Surprisingly, natural flood risks were taken into account in Roman times in Sagalassos. Excavations indicate that large open areas were carefully designed into the urban infrastructure to collect and drain the natural floods, flushing down from the mountains, hence protecting the buildings from damage (UNESCO, 2009). Karst lands cannot be disregarded based on lessons learnt from ancient populations (Parise and Sammarco, 2015).
The most prominent effect of the climate change is that the number of days with excessive rainfall has been increasing in Turkey, and the amount of rain falling in heavy events is likely to increase and be more frequent (Aksoy et al., 2008). Despite this, there is a general lack of awareness of the impact of karstic springs on flooding although there have been several heavy rainfall events in the karstic southwest area of Turkey, such as flood events in Sütçüler (Karaguzel and Akinci, 1998). Being aware of the importance of heavy rainfall and flooding afterwards in the karst areas, the General Directorate of Combating Desertification and Erosion, established under the Ministry of Forestry and Water Affairs of the Republic of Turkey, conducted a project financed by the World Bank in 1999 to combat desertification and land degradation in the Sütçüler region. In this project, 2602 ha of erosion control and 490 ha of pasture improvement works were realized. Despite these measures, a flood occurred again on 25 December 2011 in the Sütçüler region. As considered in the ancient time similar to the Sagalassos case, for the safety of urban areas such as Sütçüler, urban drainage systems must be designed, taking karst springs into account. Additionally, hydrology and hydrogeology, two well-connected disciplines on flood events in karst regions of Turkey, must be included in flood control projects.
Based on analyses and discussions of the data compiled for the karst springs
in the study area, the following concluding remarks can be derived.
An approach based on groundwater temperature, physico-chemical properties
and discharge rates of springs, in addition to lithological and structural
features, is used to determine the storage and flow conditions of the
aquifer. This approach can also be employed for the identification of
flood-prone areas in large regions, provided that representative and
organized data sampling is available at least twice, in wet and dry seasons. Recharge of karst aquifers dominated by conduit permeability increases
with heavy rain, which is becoming more frequent with climate change. Deeply fractured karst aquifers located on steeper parts of the land and
bordered by impervious rock can transfer a considerable amount of water from
different hydrological drainage basins to flood areas. It should be kept in
mind that each karst aquifer is site-specific and has its own signature. In
situ measurements and discharge rates ratio can give quick assessments and
insights before a detailed analysis-based validation is performed. For the safety of urban areas, drainage system designs should consider the
impact of karst springs, as in ancient times. Finally, to achieve more reliable estimates, the cooperation between
hydrologists and hydrogeologists is important and needs to be emphasized.
The author would like to thank the reviewers and the editor for their constructive and courageous comments that improved the paper substantially. The manuscript has been edited in terms of its language by Darwin E. Fox, whom the author deeply thanks. Edited by: M. Parise Reviewed by: Y. Örgün and one anonymous referee