Journal of Arid Environments 183 (2020) 104299

Contents lists available at ScienceDirect

Journal of Arid Environments

journal homepage: www.elsevier.com/locate/jaridenv

Response of the Mogan and Eymir lakes (Ankara, Central Anatolia) to

global warming: Extreme events in the last 100 years
Ahmet Apaydın a, Faruk Ocakoğlu b, *
a
Giresun University, Department of Civil Engineering, 28200, Giresun, Turkey
b
Eskisehir Osmangazi University, Department of Geological Engineering, 26480, Eskişehir, Turkey

A R T I C L E I N F O A B S T R A C T

Keywords: Mogan and Eymir lakes are shallow wetlands which are potentially susceptible to global warming. Previous
Central Anatolia geological studies demonstrated several complete desiccation and shrinkage events in Lake Mogan since the
Drought 1800s. Archival records show at least three desiccation events in 1910, 1934 and 1938, and many severe
Flood
shrinkages later. Here, the significance of climatic and anthropogenic drivers is still a subject of curiosity. The
Global warming
Lake water budget
back-analysis of the hydrological data during the regulator intervention indicates that Lake Mogan would have
Mogan and Eymir lakes dried out in 1974, and would have shrunk considerably in the years 2007/2008 and 2017/2018 if human
intervention was not involved. Our lake budget analysis shows that, assuming no human intervention and stream
recharge, the water level of both lakes would have dropped below the threshold level in less than a week by
natural downstream discharge, and the remaining volume of the lakes would evaporate in less than two years.
This fact indicates the benefits of regulators for the protection of the lakes. The meteorological and lake budget
data suggest that the underlying climatic threat over the studied lakes is high frequency (sub-decadal) rainfall
scarcity coupled with long-term (several decades) insolation increase but not slow, steady deterioration of the
climatic parameters.

1. Introduction formed by natural dams created by alluvial deposits in the mid-Holocene

(Ocakoğlu et al., 2018). The following three characteristics give these
Since inland lakes are tightly coupled with the atmosphere, and lakes a unique standing in environmental and ecological concerns: 1-
many of them support an increasing number of human population in hosting a distinct habitat of both aquatic and bird species (Karabulut
nearby areas, they are physically and ecologically susceptible to any et al., 2009; Muluk and Beklioglu 2005; Yağbasan and Yazıcıgil, 2009),
changes in these two drivers. Global warming poses a serious threat to 2- serving as vital recreation areas for the citizens of Ankara, 3- posing a
lakes in semi-arid regions by increasing evapotranspiration, decreasing flood threat to the surrounding lands and the city of Ankara. Monitoring
precipitation and altering seasonal climate patterns. Direct anthropo­ and management of the water balance of lakes are especially crucial
genic impact, such as contamination and the use of inflowing streams for since heavy rainfall, and sudden snowmelt in the area have caused
irrigation, further aggravate lake-related problems. In particular, floods in the past. It is also essential to make sure that the level of lakes in
shallow and small lakes react to even small changes in precipitation and dry seasons should not fall below critical values that would harm the
evaporation and are highly susceptible to anthropogenic interventions. ecology of lakes. Two control structures were built to regulate the dis­
The most common response of such lakes is temporary or permanent charges of the Eymir and Mogan lakes in 1957 and 1974, respectively.
shrinkage or even complete drying out. The Mogan and Eymir lakes, The present study analyses the budget and dynamics of the Mogan
which are the subject of this study, are sensitive to atmospheric and and Eymir lakes with a particular emphasis on the extreme meteoro­
anthropogenic impacts. To exemplify, these lakes have not completely logical events. In this regard, lake responses to severe droughts in the
dried out in the past several decades, but they have, from time to time, last 45 years of controlled regulation are investigated. The study also
temporarily shrunk. Two flooding events in the last decade seriously provides some estimations for what would have happened to these lake
incapacitated the lakeside facilities, too (Ocakoğlu et al., 2018). systems if they have not been manipulated artificially. The outcomes of
Mogan and Eymir lakes, located in close vicinity (20 km) to Ankara, the present study would provide valuable insights for the sustainable

* Corresponding author.
E-mail addresses: [email protected], [email protected] (F. Ocakoğlu).

https://doi.org/10.1016/j.jaridenv.2020.104299
Received 18 June 2020; Received in revised form 20 August 2020; Accepted 21 August 2020
Available online 2 September 2020
0140-1963/© 2020 Elsevier Ltd. All rights reserved.
A. Apaydın and F. Ocakoğlu Journal of Arid Environments 183 (2020) 104299

management of the Mogan and Eymir Lakes and understanding the to calculate the lake volumes for various water levels. Data of stream
susceptibility of shallow inland lakes in other semi-arid regions to the discharge, lake level change and evaporation loss between 1998 and
climate change. 2004 were provided by EİEİ (2005). Water level measurement in the
lakes started in 1960 by DSİ. Following a large gap between 1974 and
2. Global warming and inland lakes 1996, lake level measurements continued until recently in Lake Mogan.
But, measurements in Lake Eymir paused many times between 1963 and
Panel report of the last widespread initiative on the global warming, 1977 and stopped entirely in 2008. All recorded data from different
IPCC-2018, used a new series of data to clearly demonstrate dominant research centres were combined and used in the budget analysis for: 1-
anthropogenic causes of climate change (i.e. greenhouse gases) and its determination of the size of lakes in response to extreme dry years
impacts upon the natural and human systems (Hoegh-Guldberg et al., (1974, 2006–2008, 2015–2017) when the control structures were closed
2018). It is now generally accepted that global warming, though varying to prevent downstream discharge, 2- back-analysis of the lake area
in intensity and leading to different climatic trends from region to re­ changes using the relationship between the inputs (riverine and
gion, is altering the physical system and interrelated ecosystem, groundwater contribution to the lake) and outputs (evaporation from
including human societies, in a measurable level (Hoeg-Guldberg et al., the lake surface and downstream stream discharge) of the lake budget
2018; Reyer et al., 2017; Dokulil, 2014). Although an overestimation in components under natural/unaltered conditions, 3- Analysis of the
climate model simulations was suggested (Fyfe et al., 2013), there are a meteorological conditions causing extreme flood and drought events by
wealth of data collected in the last 20 years demonstrating that a com­ using the data recorded at the Ankara Meteorological Gauging Station,
bination of anthropogenic and natural climatic factors would cause 20 km away.
global warming of 2 ◦ C compared to the pre-industrial level, which is
well above the limit of 1.5 ◦ C adopted by IPCC (Xu et al., 2018). The 4. The Mogan and Eymir lakes
long-term climate change estimations for the Mediterranean region that
comprises Anatolian peninsula are equally worrying. Using a climatic 4.1. General characteristics
simulation for the years 2070–2100, Sanchez et al. (2004), suggests that
for Central Anatolia, mean maximum temperatures will be 6–7 ◦ C and Mogan and Eymir lakes are located within the boundaries of the
3–4 ◦ C above the present values in summer and winter months, Gölbaşı district, 20 km south of the capital city Ankara (Fig. 1). The area
respectively. They similarly estimated a precipitation decrease of is between 900 and 1800 m above sea level and constitutes the upstream
25–50% and 10% in summer and winter, respectively. Recent regional of the Ankara River Sub Basin belonging to the Sakarya River Basin, one
estimations confirm a very high vulnerability of the Mediterranean basin of the largest basins in Turkey. A district centre (Gölbaşı) and many
to aridity (Dalfes et al., 2017; Hoeg-Guldberg et al., 2018). villages are present in the drainage basin of the lakes. A semi-arid
Inland lakes are among the hydrologic systems (both physically and climate prevails in the study area. Accordingly, summers are hot and
ecologically) most likely to be adversely affected by global warming dry, and winters are cold, rainy and snowy. Annual precipitation is
since their hydrology is primarily related to atmospheric conditions. about 360 mm (DMİ, 2018). Snowfall usually occurs from December to
Global warming, in combination with direct anthropogenic impacts such March. In the extremely cold winters, the Mogan and Eymir lakes
as irrigation, often leads to scarcity of freshwater resources at mid- frequently freeze.
latitudes (Rodell et al., 2018). The apparent effects of global warming Mogan and Eymir lakes are hydraulically connected but separated by
in many European lakes are the alteration of patterns of stratification, a 2.5 km-wide swamp area. The discharge of Lake Mogan is through
mixing, ice duration and biological processes (Mooij et al., 2005; Lake Eymir via a channel. Lake Eymir outflows into the İncesu Stream
Dokulil, 2014). Shallow lakes in semi-arid regions also experience sig­ flowing northward along the İmrahor Valley (Fig. 2). Volume versus
nificant changes in hydrologic conditions (lake level, residence time), areal extent of the lakes is shown in Table 1. There was an average of 3 m
major ions, nutrient dynamics and process and aquatic ecology of elevation difference between Mogan and Eymir lakes in the natural
(Beklioğlu et al., 2007) due to global warming. Even large water bodies state, but the water level in Lake Mogan increased due to the embank­
in a semi-arid setting such as the Aral Sea in central Asia (Lioubimtseva, ment (also called a regulator) built in 1974. In 1956, a control structure
2014; Scott et al., 2011; Micklin, 2007) and Urmiye Lake in Iran was built downstream of Lake Eymir. Today, the lake level difference
(Choobari et al., 2016; Mahsafar et al., 2017) have been adversely between the two lakes is about 4 m. In dry years, both control structures
influenced by a combination of climate change and local anthropogenic are closed in the summer (mostly from May 1 to October 1) to prevent
forcing. Parallel to global trends, some of the largest lakes in central excessive water loss, an undesirable situation for the ecosystem of the
Turkey (e.g. Burdur, Acıgöl, Seyfe) have shrunk (Ataol, 2010; Temiz and lakes. On the other hand, the lake level should not exceed 973.5 m asl in
Durduran, 2016; Kıymaz, 2009), and some others (Akgöl, Akşehir) even order not to flood the roads and facilities (restaurants, parks, cafes,
wholly dried out in the last 30 years (Dursun, 2010; Şener et al., 2010; sporting areas etc.) around Lake Mogan. A similar situation occurs in
Çatal, 2015; Temiz and Erduran, 2016; Bucak et al., 2017; Dervişoğlu Lake Eymir where the maximum safe level is 969.5 m asl (METU, 1995;
et al., 2017). Debates are still ongoing as to whether climate change or Yağbasan, 2007; Yağbasan and Yazıcıgil, 2009; Ocakoğlu et al., 2018).
anthropogenic impacts are more dominant in this dangerous trend
(Jeppesen et al., 2015). However, studies dedicated to understanding 4.2. Previous paleoclimatological studies
the relative effects of climate change and human activities (i.e. irriga­
tion) on lake budget are scarce (Yağbasan and Yazıcıgil, 2012). The Early drilling studies demonstrated a maximum sediment thickness
main reason for this is the insufficient long-term instrumental data of 7 m in Lake Mogan (UKAM, 1998). But in the swamp area between
which is a prerequisite for a reliable estimation of lake budget. two lakes, drill holes continued in the incohesive sediments until a depth
of 13.5 m without reaching the basement (Dişli, 2007; Yurtseven, 2006).
3. Materials and methods A notable study by Çakıroğlu et al. (2016) investigated past salinity
estimates in a 55 cm thick core from Lake Mogan using Cladoceran
Budget analysis of the Mogan and Eymir lakes involves stream biostratigraphy. They differentiated three salinity zones along with the
inflow, groundwater recharge, evaporation losses from lake surface, core: Zone 1 (1860–1930 AD) is characterised by high salinity and is
water level changes and downstream discharges. The split of the budget followed by slightly increased salinity along Zone 2 (1930–1990 AD).
components between lake water level and downstream discharge de­ Salinity is most elevated in Zone 3 (1990 to recent) which is also
pends on lake bathymetry and elevation of the outlet. Bathymetry maps instrumentally confirmed. Later, Ocakoğlu et al. (2018) investigated a
(NFB, 2015) by The Turkish State Hydraulic Works (DSİ) have been used 125 cm thick core collected from the centre of Lake Mogan using

2
A. Apaydın and F. Ocakoğlu Journal of Arid Environments 183 (2020) 104299

Fig. 1. Location of the Mogan and Eymir lakes (source of shaded relief map is USGS-NGA-NASA (2015)).

different paleoclimatological proxies. They discovered interbedded soil elevation-volume graph is rather linear (Fig. 5). Lake area increases
structures (columnar joints and breccia) at four different levels within rapidly up to 966 m asl, then is negligible between 966 and 968.5 m asl.
the lake sediments after 1800 AD, and significant fluctuation in the δ18O From 968.5 to 969 m asl, the lake area rapidly increases again followed
curve (up to 0.5 per mil) after 1900 AD. They concluded that the limnic by a final decrease.
signature of Lake Mogan oscillated between dry and wet conditions, and
in some cases complete desiccation, over the last 200 years. 5. Water balance of the Mogan and Eymir Lakes

4.3. Bathymetry of the lakes 5.1. Stream flow and groundwater flow

The shape of the fluvial valley occupied by the Mogan and Eymir Lake Mogan is mainly recharged by surface flow, while the
lakes is controlled by the underlying lithology and principally de­ groundwater recharge is negligible. The Çölovası, Yavrucak and Sukesen
termines the lake bathymetry. Since Lake Mogan occupies a broad valley streams are the main contributors to the lake budget (Fig. 2). According
incised in the loose Plio-Quaternary sediments, its lateral margins are to the measurements between 1999 and 2004, the total share of the
relatively smooth, and the longitudinal terminations dip towards the Çölova, Yavrucak and Suksesen streams’ input to Lake Mogan varies
centre due to the sediment influx from both ends (Fig. 3). Lake Eymir, between 73 and 93% (EİEİ, 2005). Lake Mogan is the primary source for
bottomed by stiff schists in a meandering valley, has steep margins and Lake Eymir; it discharges from the surface and underground into Lake
almost flat bottom topography, and is similarly filled with subaqueous Eymir (METU, 1995; Yağbasan and Yazıcıgil, 2009). Lake Eymir out­
sediment flows from both ends (Fig. 3). flows to İmrahor Stream to the northeast through a collector canal
According to the most recent bathymetry maps (Figs. 3 and 4) of the (Fig. 2). The lake also discharges from underground into the alluvium of
studied lakes (NFB, 2015), the elevation-volume graph of Lake Mogan İmrahor Valley (Özaydın, 1997; DSİ, 2007).
consists of three segments with different slopes (Fig. 5). The volumetric Lake Mogan is mostly surrounded by the Pliocene and Miocene
increase is low up to 971 m asl, moderate between 971 and 973 m asl mudstone, marl and sandstone. At the same time, the loosely-packed
and high above 973 m asl. The lake area increases linearly up to 973 m Pliocene mudstone-sandstone and Triassic schist-greywacke forma­
and then decreases. Since the margins of Lake Eymir are steep, its tions crop out in the vicinity of Lake Eymir. Between the lakes, alluvial

Table 1
Volumetric data for the minimum (spillway), optimum (desired) and maximum (upper limit) levels of the Mogan and Eymir lakes.
Parameter Mogan Eymir

Water level (m) Minimum (971) Normal (972) Maximum (973.25) Minimum (967) Normal (968.5) Maximum (969)
Volume (106 m3) 5.0 9.71 18.9 2.15 3.89 5.44
Area (km2) 4.7 5.92 7.68 1.06 1.27 1.66
Max. depth (m) ≥2 ≥3 ≥4.25 ≥3.5 ≥5 ≥5.5

3
A. Apaydın and F. Ocakoğlu Journal of Arid Environments 183 (2020) 104299

Fig. 2. Stream network in the Mogan and Eymir lake basins.

Table 2
Monthly discharge in Lake Mogan at selected lake levels.
Lake level (m.asl) 971 971.25 971.5 971.75 972 972.25 972.5 972.75 973 973.25 973.5

Monthly discharges (1000 m3), 1999-2000-2001 hydrologic year 0 70 75 110 200 225 400 500 1780 3470 5000

4
A. Apaydın and F. Ocakoğlu Journal of Arid Environments 183 (2020) 104299

Fig. 3. Bathymetry map and cross-sections of Lake Mogan.

sediments extend for 2.5 km down valley. The lakes are hydraulically the alluvium was 9 l/s. In the same period, the total recharge into Lake
connected via surface and groundwater flow. Eymir is 17 l/s and the discharge to İncesu alluvium is only 2 l/s. The
The drilling studies performed by UKAM (1998) in Lake Mogan show majority of the inflow to Lake Eymir is from Lake Mogan (9 l/s). The
up to 6 m of fine-grained material beneath the lake bottom. Two layers laboratory experiments on cores taken from the alluvium between the
of hard plastic clay were distinguished in the upper core. Ocakoğlu et al. Mogan and Eymir lakes display very low (between 10− 2 -10− 6 m/s)
(2018) found up to 30 m of dominantly fine-grained Holocene sediments hydraulic conductivity (Dişli, 2007). Furthermore, the hydraulic
beneath Lake Mogan. Dişli (2007) experimentally demonstrated that the gradient between the lakes is low (0.12–0.13). These data confirm the
hydraulic conductivity is very low (less than 10− 10 m/s) in the Pliocene previous suggestions that the groundwater recharge from Lake Mogan to
sediments surrounding the alluvium between the Mogan and Eymir Lake Eymir is negligible (METU, 1995; Yağbasan, 2007; Yağbasan and
lakes. Based on the lithology, a similar low conductivity can be attrib­ Yazıcıgil, 2009).
uted to the Triassic schists and meta-sediments, and underlying Holo­
cene lake muds. This fact implies a very low possibility of water 5.2. Lake level fluctuations and climate
circulation between Lake Eymir and the subsurface. Accordingly, it was
concluded that the groundwater transfer between Mogan and Eymir In the monitored period, the lake levels fluctuated within a relatively
lakes would only be possible through modern coarse alluvium at the narrow interval of 972–974 m in Lake Mogan, and 967–969 m in Lake
mouth of major streams. Previous water budget estimations confirm a Eymir (Fig. 6). Since they are hydrologically coupled, their level changes
small contribution through groundwater: METU (1995) suggests that are also compatible. In Lake Eymir, the seasonal fluctuations follow
95% of the input to Lake Mogan occurs from surface water, while different long-term trends with time: a horizontal pattern between 1960
groundwater contribution is negligible. METU (1995) and Yağbasan and 1967 years (967.5–698 m asl), a decreasing trend between 1968 and
(2007) estimated a total groundwater inflow rate of 20 l/s into Lake 1974 (from 973.5 to 971.5 m asl), a flat trend between 1996 and 2001
Mogan between 1998 and 2004, while the outgoing groundwater from period (around 973 m asl), a decreasing trend in 2001–2002 period over

5
A. Apaydın and F. Ocakoğlu Journal of Arid Environments 183 (2020) 104299

Fig. 4. Bathymetry map and cross-sections of Lake Eymir.

973 m asl), a short-term rising trend in 2002–2003 period (again on the Mogan Lake. Additionally, Alicin Stream, which traverses a relatively
973 m asl), a decreasing trend between 2003 and 2010 years large watershed around Yaylabağ village in the east, only feeds Lake
(973.5–972 m asl), and an increasing trend between 2014 and 2015 Eymir during rainy seasons (Fig. 2). Alicin Stream is occasionally
years (from 973.5 to 973.5 m asl) (Fig. 6). Although data is missing in diverted to İncesu Stream without joining Lake Eymir. The discharge
the earliest part of the Eymir record, a sharp decline (from 969 to 966 m from Lake Eymir occurs via İncesu Stream on the surface, infiltration
asl) after 1974 can still be differentiated. Following a rise between 1977 into the alluvium downstream, and by evaporation from the lake
and 1982 and a fall between 1987 and 1994, a three-year flat trend surface.
(around 968 m asl) occurs in the 1998–2001 period. After a short-term The water balance components of Lake Mogan include: 1) direct
decline between 2001 and 2002 (from 969 to 967.5 m asl), an increase rainfall to the lake surface (P), 2) inflow from the feeder streams (Qis), 3)
occurr in 2002–2003. After a long term decrease (from 699 to 697 m asl) inflow from the groundwater (Qig), 4) stream discharge from the lake
in the 2003–2008 period, a data break occurs (Fig. 6). Lake level con­ (Qds), 5) flow into groundwater (Qdg), 6) lake evapotranspiration (Etp)
tained fluctuating after 2014. and 7) change in the storage (ΔR). So, we can assume the following
Fig. 6 also reveals the close relation of lake levels with precipitation balance equation for Lake Mogan:
and evaporation. The monthly evaporation in the monitoring period
displays a sinusoidal variation with two highs in 1966 and 1998 and two Pm + Qism + Qigm = Etpm + Qdsm + Qdgm ± ΔRm
troughs in 1984 and 2014. The potential desiccation/shrinkage as well Similarly, the water balance components of Lake Eymir are: 1) the
as flooding events revealed by our back-analysis matches with the rising direct rainfall to the lake surface (P), 2) surface flow from the Mogan
segment of the evaporation curve. Furthermore, lake level drops also (Qism) 3) flow from the own basin streams (Qis), 4) inflow from the
align well with the prolonged scarcity of rainfall below the long-term groundwater (Qig), 5) outflow into the İncesu Stream (Qds), 6) flow into
average. So we suggest that the long-term (several decades) insolation the groundwater of the İncesu Valley (Qdg), 7) Lake evapotranspiration
increase and the high frequency (several years-long) precipitation (Etp) and 8) change in the storage (ΔR). Then, the following water
deficit are the basic climatic influences for the lake desiccation/ balance equation can be assumed for Lake Eymir:
shrinkage. On the other hand, the main driver of the flooding seems to
be flash summer rains and the spring (mainly March) snowmelt due to Pe + Qisme + Qise + Qige = Etpe + Qdse + Qgge ± ΔRe
rapidly rising air temperatures. The data from DMI shows a discernible
Fig. 7 depicts the average budget data calculated from the recharge/
increase in mid-March air temperature starting in 1997. Rising air
discharge model of the Mogan and Eymir lakes between 1999 and 2004.
temperature, along with the average increase in annual temperature, is
Here, data for the surface flow, direct precipitation, evapotranspiration
regarded as a direct manifestation of the global warming in central
and discharge were taken from EİEİ (2005) whereas the data for
Anatolia (Altın et al., 2012; Şen, 2018).
groundwater recharge and discharge were obtained from METU (1995).
The following remarks can be made in terms of the budget of Mogan and
5.3. Budget analysis of the lakes
Eymir lakes.

As mentioned earlier, surface flow dominates over groundwater

inflow into Lake Mogan. The main output occurs as evapotranspiration
from the free lake surface, and surface and underground flow towards
Lake Eymir. Lake Eymir mainly supplies its water from the upstream

6
A. Apaydın and F. Ocakoğlu Journal of Arid Environments 183 (2020) 104299

Fig. 5. Elevation versus volume and area graphs of the Mogan and Eymir lakes.

Fig. 6. Change of lake levels and meteorological parameters (data from Ankara Meteorological Station) between the years 1926 and 2019.

7
A. Apaydın and F. Ocakoğlu Journal of Arid Environments 183 (2020) 104299

Fig. 7. A model of the budget components belonging to the Mogan and Eymir lakes (calculated from the data in EIEI (2005) and METU (1995)).

Fig. 8. Lake level versus discharge graph of Lake Mogan.

5.3.1. Lake Mogan lake level will reach to Ln (972 m) in 73 days and Lm (975.3 m) in
284 days.
1 The regulator threshold in front of Lake Mogan is at 971 m while the 3 In the Lm condition (973.25 m), the volume of the lake is 18.9 hm3,
lake bottom stands at 969.10 m altitude. In this lake, the normal (Ln- of which 12.7 hm3 (active fraction) stands over Ln case (971 m). If
desirable minimum) water level is 972 m and tolerable maximum the water level drops from Lm to Ln (from 973.25 m to 971 m) under
water level (Lm-edge of the coastal facilities and roads) is 973.25 m. imaginary “no inflow - no evaporation” conditions, the removal of
The amount of lake water in Lm and Ln conditions is 18.9 hm3 and 9 the active volume through the discharge channel at maximum ca­
hm3 respectively (1 hm3 = 106 m3). When the water level recedes to pacity (29 m3/s) occurs in only 5.06 days. If the weak groundwater
971 m, the depth at the lake centre decreases to 190 cm. In this case, inflow is considered, the lake level will drop to regulator threshold
the remaining water volume is 6.2 hm3, and the lake area diminishes on the sixth day.
to 4.7 km2. When the water level rises to Ln and Lm, the maximum 4 At Ln conditions (971 m), if there is no streamflow into the lake and
water depth reaches 290 cm and 415 cm, respectively. the downstream discharge is wholly halted through the regulator, all
2 According to the 1999–2004 data, the average contribution of sur­ lake water (6.2 hm3) will evaporate almost in a year (354 days)
face flows to Lake Mogan is 10 hm3/year. If the lake completely dries under average evaporation conditions (6.4 hm3). In other words,
out, the ensuing average flows will raise the lake level to the regu­ Lake Mogan at Ln condition is drought resistant for a maximum of
lator threshold (971 m) in 226 days. Almost a year (353 days) is one year. An extreme summer season would even be enough for
required to reach the Ln (972 m) level. However, in wet years, such complete desiccation of the lake. Using the relationship between
as the year 2000, a shorter duration (178 days) following the hy­ water level and discharge amount of Lake Mogan, the discharge rate
drologic year (October) will suffice to reach Ln level. If the water of the lake was estimated. For this purpose, we used monthly data
level stands at 971 m asl at the early hydrologic year of 2000, the recorded by EİEİ between 1999 and 2004 and by DSİ between 2014
and 2017 while the regulator was open and the monitoring was

8
A. Apaydın and F. Ocakoğlu Journal of Arid Environments 183 (2020) 104299

ongoing. Fig. 8 displays the monthly discharge as the function of the level of Lake Mogan dropped from 972.72 m to 971 m. Accordingly, the
lake level. The discharge is higher over the 972.75 m water level and lake level at the end of September would have dropped to 971.85 m.
logarithmically decreases to zero at 971 m. Table 2 lists the monthly Nevertheless, the actual measured level at this time was 21 cm higher
discharge values for selected elevations derived from Fig. 8. The (972.06 m).
highest natural discharge mostly occurs between April and June In the Lake Mogan basin, the years 2006, 2007 and 2008 were dry,
when the lake level is highest. and the average annual precipitation was 371, 305 and 323 mm
respectively at Ankara DMİ station. The rainfall in spring and summer
5.3.2. Lake Eymir was unexpectedly reduced in 2007, but in November significant rainfall
resumed. The winter of 2007–2008 was quite arid. Following a wet
1 Here, the regulator threshold is at 967 m, while the lowest lake March and May period, summer was very dry again. Due to this drought,
bottom is at 964 m. The normal (Ln-desirable minimum) is 968.5 m, the regulator was kept closed between June and December 2006 and
and tolerable maximum water level (Lm) is 969.5 m. When the lake June 2007 and end of 2008. The water level in the lakes could be
level drops to 967 m (a shoaling of 300 cm), the volume and extent of maintained above a certain level as the discharge was controlled by
the lake reduce to 2.15 hm3 and 1 km2, respectively. If the lake level closing the regulator in the summer before these three dry years. If the
raises to Ln and Lm elevations, the maximum depth of the lake will lake had been kept in a natural state for many years before 2007, the
be 450 cm and 550 cm, respectively. lake level would have decreased significantly compared to the measured
2 In the Lm case, the lake volume is 5.4 hm3, of which 3.25 hm3 seems level. For example, if the Mogan regulator had been kept open between
active, i.e. above the regulator threshold. Assuming no water input June 1 and October 30 of 2006, roughly 5 hm3 of lake water would have
into the lake, the active volume will be fully discharged in only three been discharged. In this case, 5.5 hm3 of water would remain in the lake
days when the regulator channel serves at full capacity (15 m3/s). by the end of November 2007. The calculated lake level elevation at this
3 If the lake level drops to the regulator threshold and no water input time is only 15 cm above the minimum desired level of 971 m asl.
occurs, the lake volume (2.15 hm3) will evaporate entirely at the end Consequently, although not wholly desiccated, the lake would have
of the following year (i.e. duration of resistance of 2 years at max.) shrunk considerably in this period, threatening the ecology.
under average annual evaporation conditions (1.1 hm3). Years 2017-2018: By the end of May 2017, the level of Lake Mogan
was at 973.31 m asl. On June 1, the regulator was closed as is customary.
6. Back-analysis of severe dry years for natural conditions The lake level dropped to 972.85 m asl, while the water volume
diminished from 20 hm3 to 15.1 hm3 in 4 months at the end of
As mentioned earlier, the outlet of both lakes has been regulated September. If the regulator were not closed, the water level would have
since 1974 as a drought measure. The quantitative approach of the decreased further below 972.85 with downstream discharge at the end
present study allows us to get some insights into what the situation of September. Considering the amount of discharge via the regulator
would be if the regulators were not constructed (i.e. natural flow con­ channel (Fig. 8), the lake level is estimated to have been at 972.25 m, 60
ditions were kept). Below, we estimated possible states of the lakes for cm below the measured level (Table 3).
the marked dry years of 1974, 2007 and 2017. In this period, another unexpected input to Lake Mogan should also
The year 1974: At the end of May 1974 (water level = 978.1 m) the be considered. According to the data from the General Directorate of
regulator was closed to prevent the discharge from Lake Mogan to Lake Ankara Water and Sewerage Administration (ASKI), the water recharged
Eymir. By the end of September, the lake level dropped 30 cm to 971.51 from the Kesikköprü dam into the Lake Mogan totaled 2.87 hm3 between
m, which means the lake volume decreased from 9.1 hm3 to 7 hm3 in July–October in 2014, and 2.38 hm3 in 2015. In conclusion, the lake
four months. If the regulator were not closed, the water level would have level would have been roughly 971.70 m asl by the end of September
fallen further below the 971.51 m by the end of September due to 2017.
downstream discharge after early June. According to Fig. 8, the monthly As of September 2019, the regulators in both lakes were closed since
discharge decreased from 50,000 m3 to zero while the water level in the end of February 2017. Even in the wet May of 2017 and the
Lake Mogan decreased from 971.81 m to 971.00 m. Assuming discharge following summer and autumn, most of the streams feeding the lake,
rates of 35,000 m3/month between 971.81 m and 971.5 m and 35,000 except the Sukesen and Yavrucak streams, had no flow (Fig. 2). In the
m3/month between 971.5 m and 971 m, and taking into account the summer and autumn of 2017 and January 2018, the flow rate of Sukesen
monthly average evaporations between 1999 and 2004, the level and and Yavrucak streams was measured 30–40 l/s and 40–50 l/s, respec­
volume of the lake were calculated on monthly basis. Accordingly, lake tively. The contribution of these streams to the lake budget in December-
level dropped 60 cm below the acceptable threshold by the early January-February period was 1 hm3. Because the evaporation losses
September and reaches 970.90 m asl by the end of September. DMİ re­ were negligible in winter, and the water level was lower than 972 m asl,
cords show the lake level was decreasing throughout the relatively dry the expected downstream discharge rate (Fig. 8) would have been small.
autumn of 1974. Eventually, the lake level was calculated 970.70 m asl Consequently, the lake level at this time would have been 971.65 m asl,
(30 cm below the desired minimum level) by the end of November. slightly above the desired minimum level.
Water budget calculations for Lake Mogan show that the lake level The lowest elevation of the bottom of Lake Eymir is 964 m. When the
dropped to 970.70 and the lake volume decreased to roughly 4 hm3 at level of the lake drops to 967 m, the storage and extent of the lake are
this time. It is suggested that the lake level started to rise again following 2.15 hm3 and 1 km2 respectively. The water depths for Ln (968.5 m asl)
the precipitation in December. Considering the transpiration from the and Lm conditions (969.5 m asl) are 450 cm and 550 cm respectively.
swamp areas and the underestimated evaporation in the DMİ records, When the regulator was closed between end February and end 2017, the
the level of Lake Mogan would have fallen even further, to the edge of lake depth was measured as 5 m on March 20 and June 10, and 3 m on
desiccation. However, the lake level would have risen again with the October 9. Accordingly, in 2015–2016 lake water level was recorded as
rainfall and snowfall at the end of 1975. 968.83 m (50 cm above the Ln) when the regulator was closed. If the
The years 2007-2008: At the end of May 2007, the lake level was at regulator had not been closed in February, the lake level would have
972.72 m. On June 1, the regulator was closed. The lake level decreased decreased significantly beyond the measured level. At the Lm position,
to 972.06 m at the end of September (a drop of 66 cm). As a result, the the volume of Lake Eymir is 5.4 hm3 where 3.25 hm3 is active. At Ln
lake storage diminished from 14.3 hm3 to 10 hm3 in four months. If the position of lake level, 2.15 hm3 of water is lost at the end of the second
regulator was not closed, the lake level would have dropped below year, and the lake completely dries out assuming no water inflow more
972.06 m at the end of September due to downstream discharge. As seen than the average annual evaporation (1.1 hm3) occurs. In other words,
in Fig. 8, the monthly discharge diminished from 0.35 hm3 to 0 while the the lifespan of Lake Eymir at the Ln position is 2 years at maximum. As a

9
A. Apaydın and F. Ocakoğlu Journal of Arid Environments 183 (2020) 104299

Table 3
Monthly decrease of lake level (m.asl) and water volume (hm3) for some critical dry years in Lake Mogan due to the downstream flow and evaporation losses.
Year Lake May’31 June’30 July’30 August’30 Sept’30 Oct’31 Nov’30

1974 Volume 9.1 7.95 6.6 5.37 4.69 4.25 4.06

Elevation 971.81 971.65 971.35 971.10 970.92 970.75 970.70
2007 Volume 14.3 12.93 11.36 10.0 9.37 8.9 –
Elevation 972.72 972.55 972.30 972.05 971.90 971.85 –
2017 Volume 19.8 15.89 14.08 12.38 11.23 10.71 10.53
Elevation 973.30 972.94 972.71 972.45 972.25 972.20 972.10
Excluding the additional water from the Kesikköprü Dam in 2014 and 2015 971.65

Fig. 9. Comparison of the actual (a), and most likely (if human intervention had not been occurred) (b) extent of Lake Mogan in selected dry years.

result, the water level, water volume and lake area of the lakes were 8. Conclusion and discussion
maintained at certain levels in dry periods. If the regulators have been
not constructed, the shoreline of the lakes would be quite different Our budget estimations indicate that the Mogan and Eymir lakes in
compared to their current position, especially after the dry years (Fig. 9). the semi-arid central Anatolia are highly susceptible to rainfall and
evaporation conditions. Since these lakes are shallow and dominantly
7. Major extreme events and their impacts on the lakes in the fed by surface waters, they respond very quickly to even short-term
last 100 years meteorological extremes. As a result, Lake Mogan experienced several
desiccation events in the early decades of the 20th century and some
During the period of instrumentation in the Ankara region, the catastrophic floods in the 1950s and 1960s. Two regulators at the
notable dry years are 1926–1934, 1938, 1986–1987, 1992–1994, 2001, downstream of each lake were constructed in 1974. Since then, both
2006–2008, 2014 and 2017–2018. Flood events occurred in 1951, 1957, regulators have been closed to reduce the impact of drought during dry
1961, 1968, 2011 and 2012. Data from the Ankara meteorological periods. However, in some dry years (such as 1974, 2007 and 2017), the
gauging station show very reduced rainfall in the years 1926–1929, water level was still significantly reduced and the lakes shrunk. These
1932–1934, 1944–1945, 1950, 1956, 1977, 1992–1994, 2004, responses in the last 100 years are in line with the archival, testimonial
2007–2008 (Fig. 6). These data also demonstrate that hot and long and geological records extending back to the 18th century.
summers prevailed in the years 1927, 1929, 1930, 1931, 1935, 1937, The limited previous studies have suggested a long-term impact of
1986, 1998, 2001, 2004, 2007 and 2008. These years are characterised climate change on the levels of Mogan and Eymir lakes. Yağbasan and
by less rainfall and higher evaporation than the average of summer Yazıcıgil (2012) estimated that, under the constant climatic conditions
seasons (Fig. 6). The winter seasons were remarkably cold and dry in the of the years 1998–2002, Mogan and Eymir lakes would be completely
years 1942, 1949, 1950, 1954, 1964, 1976, 1992 and 1993. desiccated in 2075 and 2067, respectively. Moreover, the A2 (higher
The extreme dry, cold and flood events around Mogan and Eymir emission) and B1 (low emission) scenarios shift the complete desiccation
lakes are listed in Table 4; data from before 1926 were obtained from to much earlier: these are 2023 and 2037 for Lake Mogan, and
archival records, and hydro-meteorological data after 1926 are from XX 2021–2032 for Lake Eymir. Conversely, our analysis demonstrates that
source. Also, the extent of the Mogan and Eymir lakes under severe dry the main danger to the lakes is not an overall subtle change of rainfall
and flood conditions in the last 60 years and their desired state are also and temperature, but their rapid sub-decadal fluctuations. The flat ba­
shown in Fig. 10 and in Fig. 11. To prevent flooding, the regulator in thymetry, especially apparent in Lake Mogan, accentuates the impact of
Lake Mogan is kept open in winter and spring times. However, in order climatic factors so that they rapidly shrink to desiccation in drought, and
not to let the water level fall below 971 m asl, the water level is fixed via recover again during rainy periods. A remarkable feature of the Mogan
the regulator at 972.5 m asl at the beginning of the summer (DSİ, 2007). and Eymir lakes under natural conditions is their ability to discharge
Construction of the Mogan Flood Control Structure upstream of Lake quickly downstream. While the levels are at Lm (maximum level) in the
Mogan in 2015 is the last intervention to the lake system. This structure Mogan and Eymir lakes (973.5 m and 969.5 m, respectively), water level
was built to temporarily store the waters of Kepir Stream (or Çölova naturally falls to the current regulator base level (971 m and 967 m
Stream) during heavy rainfalls (Fig. 2). Since the rainfall has been respectively) in 6 and 3 days, respectively. Before the regulators were
relatively scarce after construction, its function has not been tested yet. built, the natural discharge of the lakes is estimated as lower than the

10
A. Apaydın and F. Ocakoğlu Journal of Arid Environments 183 (2020) 104299

Table 4
Shrinkage, desiccation and flood events that occurred in the last 100-year period in the Mogan and Eymir lakes.
Year Event Description Explanation

1910 Desiccation of Lake Lake Mogan dried out in 1910 (a statement in the DSİ archive). No confirmation in the official records.
Mogan
1929 Severe shrinkage A dry period between 1927 and 1929. Rainfall deficit covering the whole year of 1928 (precipitation is 218 mm,
there is no significant rainfall until November) and the first 7 months of
1929 (except for April). There is no record of drying of the lakes, but a
considerable shrinkage in Lake Mogan is anticipated.
1934 Desiccation of Lake According to Calvi (1936), Lake Mogan is totally dried out. Lake Mogan was completely dry on between November 29 and December
Mogan 9, 1934, and the lake was covered with a thin layer of salt.
1938 Desiccation of Lake Lake Mogan completely dried out (data from DSİ archive). During the 1935–1937 period, the summers were hot and long though the
Mogan rainfall is average, but this may have contributed slightly to the flow. Due
to hot and dry summers of the previous years, the lakes should have been
withdrawn as a result of evapotranspiration. After three years of drought,
there was a small amount of excessive rainfall (102.6 mm) in July 1938.
1951 Flood İncesu and Dikmen streams overflowed as a result of excessive Thirty houses as well as the Atatürk Avenue and the Sıhhiye region was
rainfall on June 12–15. flooded in Ankara.
1957 Flood After a heavy rainfall on September 11, Hatip Stream overflowed. The heavy rain (a quarter of the annual average) lasted for 2 h in a large
(250 km2) area of. 169 people were died due to the flooding.
1961 Flood The Hatip, Incesu, Dikmen and Cubuk stream overflowed due to Three people have died. The total rainfall in June is estimated at 121.9
heavy rain on June 18–21. mm.
1968 Flood Lake Mogan overflowed following long-lasting heavy rains (the Annual rainfall of the year is high (571 mm). In Lake Mogan, the water
facilities around the Bülbül Stream-Opera and a small nearby lake level increased to 973.60 m asl (35 cm higher than the Lm-safe maximum
turned into wide lake). level).
1974 Severe shrinkage Lake withdrawal between 1971 and 1974. In September 1974: No record of drying. Eymir and Mogan regulators were closed in the early
971.5 m in Lake Mogan, and 966,04 m in Lake Eymir. summer of 1973 and 1974. Otherwise, they would have likely faced
drying.
1983 Decrease of the The level decreased from 978 m to 966.6 m in February and A similar decline in lake level probably occurred in Lake Mogan.
level and shrinkage September in Lake Eymir; no data in Lake Mogan. Surprisingly, the rainfall data in 982–1983 was above the average.
2001 Level drop and In November, the water level decreased to 971.95 m in the Mogan In the first week of July, the regulator was closed. Evaporation loss was
shrinkage and to 967.18 m in Eymir. high due to the hot and long summer. According to EİEİ data, the surface
area of Lake Mogan decreased to 6.3 km2 in November 2001 (8 km2 in the
previous year).
2007 Level decrease and In November, the water level decreased to 971.93 m in Lake In early June, the regulator was closed. Evaporation loss is probably high
shrinkage Mogan and to 966.86 m in Lake Eymir. due to the hot and long summer.
2008 Level decrease and Discontinuous lake level monitoring in March. 2006-2007-2008 Regulators are closed as of June 2007. The water level in both lakes
shrinkage receives precipitation below average. Spring precipitation in 2008 remained unchanged until March despite the hot and long summer, and
allows lake level rise. insufficient rainfall.
2011 Flood Lake Mogan overflowed on March 18 due to the snowmelt. During the flood, lake levels were well above the Lm; 37 cm in Lake
Mogan and 32 cm in Lake Eymir.
2012 Flood Lake Mogan overflowed on March 15. Level in Lake Mogan reached at 973.90 m in March, and 973.94 m in
April.
2017–2018 Level drop and Despite the drought, lake levels were slightly above the average. The regulators were closed in both lakes after February 2017. No notable
shrinkage Until September, the water level did not fall below 973 m in Lake steam flow occurred into the lakes due to insufficient rainfall in the whole
Mogan and below 969 m in Lake Eymir. year.

recent regulator bottom elevations. So, the lakes can efficiently both sides in the downstream direction. Lake Mogan is exposed more to
discharge downstream in dry years, and quickly shrink and even dry out the impacts of wind and sun compared to Lake Eymir. These charac­
with further evapotranspiration. Historical records of frequent drying teristics imply that evaporation losses will be less in Lake Eymir and this
out in Lake Mogan before the intervention support this conclusion. would help it to survive longer than Lake Mogan. Furthermore, inflow
According to the historical records, Lake Mogan dried out three times from Lake Mogan and steep lake margin topography free from water-
in the first half of the 20th century. There is no record after the last demanding vegetation and human settlements are additional advan­
known desiccation in 1938, but the monitoring and control period, tages for the longevity of Lake Eymir.
which began in 1960, shows no evidence of the lakes drying out. The Flood risk in the Mogan and Eymir lakes is closely linked to land use,
most significant aspect of the present study is to figure out how Lake urbanisation and downstream conditions in the area. Unfortunately,
Mogan would have reacted to drought without intervention in the last these factors were not fully considered in accordance with the boundary
50 years. Our estimates indicate that, if Lake Mogan had been kept under and hydrology of the lakes in the future plans and practices. For this
natural conditions over the last 50 years, it would have dried out once reason, the lakes were kept under control to prevent flooding. In other
entirely and would have shrunk significantly several times. It is note­ words, instead of keeping urbanisation under control to protect the
worthy that these conclusions rely on the data from the beginning to the lakes, the lakes were intended to be kept under control to protect the
end of dry periods. If the water level measurements, downstream dis­ city. In such systems, the ideal approach would be to keep the system in
charges and evaporation data of the lakes had been fully recorded for the its natural state and to take the necessary precautions in advance to
whole instrumental period, more reliable estimates could have been prevent flooding in the lake periphery and downstream. Unfortunately,
made. it is no longer possible to reverse the negative, restrictive impacts of
There historical records and hydrologic data available for Lake urbanisation on the Mogan and Eymir lakes. However, some actions can
Mogan are more abundant than those for Lake Eymir. Additionally, the be taken to ensure the sustainability of the lakes. Online monitoring of
historical records of Lake Mogan do not mention the state of Lake Eymir. the stream discharges, evaporation losses, downstream discharges and
Although the fate of Lake Eymir is mostly dependent on that of Lake lake level measurements is of great importance for the ecosystem, rec­
Mogan, there are some morphological advantages for Lake Eymir. Lake reation and flood control. A dataset of this kind would help to predict the
Mogan is more extensive, shallower and its surroundings are flatter future behaviour of the lakes. Although this is not the focus of the study,
compared to Lake Eymir. Lake Eymir is surrounded by steep slopes on in the light of the climate scenarios that emphasise increasing drought

11
A. Apaydın and F. Ocakoğlu Journal of Arid Environments 183 (2020) 104299

Fig. 10. Extent of Lake Mogan during (a) the floods in 1968, 2003, 2011 and 2012, (b)optimum elevation at any time, (c) the severe draughts in 1974.

Fig. 11. Extent of Lake Eymir (a) during the severe droughts in 1974, (b) optimum conditions at any time (b), and (c) the maximum level.

12
A. Apaydın and F. Ocakoğlu Journal of Arid Environments 183 (2020) 104299

and elevated temperatures in the coming decades, the desiccation fre­ DSİ, 2007. Mogan Ve Eymir Gölü Havzası (Ankara) Hidrojeolojik Etüt Raporu, DSİ 5, 62
s. Bölge Müdürlüğü, Ankara.
quency for the lakes is predicted to increase. So, we suggest that the
Dursun, S., 2010. Effect of global climate change on water balance of Beyşehir Lake
future will bring about a highly unstable hydrological behaviour in the (Konya–Turkey). In: Proceeding from BALWOIS, vol. 25. Ohrid, Republic of
lakes. Macedonia. (Accessed 29 April 2010).
EİEİ, 2005. Mogan Ve Eymir Gölleri Havzasının Hidrometeorolojik Özellikleri, 60 s. EİEİ
Raporu.
Credit author contribution statement Fyfe, J.C., Gillett, N.P., Zwiers, F.W., 2013. Overestimated global warming over the past
20 years. Nat. Clim. Change 3 (9), 767.
Ahmet Apaydın: (first author), Conceptualization, Writing - original Hoegh-Guldberg, O., Jacob, D., Taylor, M., Bindi, M., Brown, S., Camilloni, I., et al.,
2018. Impacts of 1.5 ◦ C Global Warming on Natural and Human Systems.
draft, Writing - review & editing, (lead), writing (equal), original draft Jeppesen, E., Brucet, S., Naselli-Flores, L., Papastergiadou, E., Stefanidis, K., Nõges, T.,
(equal), review and editing (equal). Faruk Ocakoğlu: (co-author, cor­ Nõges, P., Attayde, J.L., Zohary, T., Coppens, J., Menezes, R.F., Freitas, F.R.S.,
responding author), Conceptualization, Writing - original draft, Writing Kernan, M., Søndergaard, Beklioglu, M., 2015. Ecological impacts of global warming
and water abstraction on lakes and reservoirs due to changes in water level and
- review & editing, (secondary role), writing (equal), original draft related changes in salinity, 750 (1), 201–227.
(equal), review and editing (equal), Methodology, (secondary). Karabulut Dogan, Ö., Akyürek, Z., Beklioglu, M., 2009. Mapping and identification of
submerged plants in a shallow lake, Turkey using Quickbird satellite data.
J. Environ. Manag. 90, 2138–2143.
Declaration of competing interest Kıymaz, S., 2009. Seyfe gölü sulak alanı ve su kaynakları yönetimine i?lişkin sorunlar ve
çözüm önerileri. e-Journal of New World Sciences Academy2010 5 (2), 174–180.
The authors declare that they have no known competing financial Article Number: 5A0036, ISSN:1306–3111.
Lioubimtseva, E., 2014. Impact of climate change on the Aral Sea and its basin. In:
interests or personal relationships that could have appeared to influence
Micklin, Philip, Aladin, N.V. (Eds.), Book: the Aral Sea: the Devastation and Partial
the work reported in this paper. Rehabilitation of a Great Lake, Edition: 1st Chapter: Ch.17: Impact of Climate
Change on the Aral Sea and its Basin. Publisher: Springer-Verlag, Igor Plotnikov,
Acknowledgement pp. 405–427.
Mahsafar, H., Maknoon, R., Saghafian, B., 2017. The impact of climate change on water
level of Urmia Lake. Research in Marine Sciences 2 (2), 83–94.
This study is financially supported by the TUBİTAK (Scientific and METU, 1995. Gölbaşı Mogan-Eymir Gölleri İçin Su Kaynakları Ve Çevre Yönetim Planı
Technological Research Council of Turkey) research project “Paleo­ Projesi, Kesin Rapor-Cilt 1, Cilt 2 (Ekler), Orta Doğu Teknik Üniversitesi. ASKİ Genel
Müdürlüğü. No: 93-03-03-04-01, Temmuz 1995, Ankara.
climatological and Genetic Investigation of the Lakes Mogan and Eymir Micklin, P., 2007. The Aral Sea disaster. Annu. Rev. Earth Planet Sci. 35, 47–72.
(Ankara)” (grant number: 114Y557). The authors are grateful to an Mooij, W.M., Hülsmann, S., Domis, L.N.D.S., Nolet, B.A., Bodelier, P.L., Boers, P.C., et al.,
anonymous reviewer who opened new perspectives in the paper. 2005. The impact of climate change on lakes in The Netherlands: a review. Aquat.
Ecol. 39 (4), 381–400.
NFB, 2015. Ankara İli Taşkın Tehlike Alanları Belirlenmesi Planlama Raporu, DSİ 5.
References Bölge Müdürlüğü, Ankara.
Ocakoğlu, F., Dönmez, E.O., Tunoğlu, C., Akbulut, A., Apaydın, A., Tün, M., Görüm, T.,
Altın, T.B., Barak, B., Altın, B.N., 2012. Change in precipitation and temperature Tuncer, A., 2018. Paleoclimatological and genetic investigation of the lakes mogan
amounts over three decades in central Anatolia, Turkey. Atmos. Clim. Sci. 2, and Eymir (Ankara). TÜBİTAK, ÇAYDAG Araştırma Projesi 371, 114Y557.
107–125. Özaydın, V., 1997. Water Balance of Lakes by Using Sta- Ble Isotope Mass Balance
Ataol, M., 2010. Burdur gölü’nde seviye değişimleri. Coğrafi Bilimler Dergisi Cbd 8 (1), Method. Ph.D. Thesis. METU Civil Eng. Dept.. June 1997.
77–92. Reyer, C.P., OttoIM, Adams S., Albrecht, T., Baarsch, F., Cartsburg, M., et al., 2017.
Muluk, Ç., Beklioglu, M., 2005. Lack of a typical diel vertical migration: varying role of Climate change impacts in Central Asia and their implications for development. Reg.
water clarity, food, and dissolved oxygen in Lake Eymir, Turkey. Hydrobiologia 537 Environ. Change 17 (6), 1639–1650.
(1–3), 139–149. Rodell, M., Famiglietti, J.S., Wiese, D.N., Reager, J.T., Beaudoing, H.K., Landerer, F.W.,
Beklioğlu, M., Romo, S., Kagalou, I., Quintana, X., Bécares, E., 2007. State of the art in Lo, M.H., 2018. Emerging trends in global freshwater availability. Nature 557
the functioning of shallow Mediterranean lakes: workshop conclusions. (7707), 651.
Hydrobiologia 584 (1), 317–326. Sanchez, E., Gallardo, C., Gaertner, M.A., Castro, M., 2004. Future climate extreme
Bucak, T., Trolle, D., Andersen, H.E., Thodsen, H., Erdoğan, Ş., Levi, E.E., et al., 2017. events in the Mediterranean simulated by a regional climate model. EGS-AGU-EUG
Future water availability in the largest freshwater Mediterranean lake is at great risk Joint Assembly.
as evidenced from simulations with the SWAT model. Sci. Total Environ. 581, Scott, J., Rosen, M.R., Saito, L., Decker, D.L., 2011. The influence of irrigation water on
413–425. the hydrology and lake water budgets of two small arid-climate lakes in Khorezm,
Çakıroğlu, A.İ., Levi, E.E., Tavşanoğlu, Ü.N., Bezirci, G., Erdoğan, Ş., Filiz, N., Uzbekistan. J. Hydrol. 410, 114–125, 2011.
Andersen, T.J., Davidson, T.A., Jeppesen, E., Beklioğlu, M., 2016. Inferring past Şen, O.L., 2018. How does the global climate change affect the climate of Turkey? http://
environmental changes in three Turkish lakes from sub-fossil Cladocera. climatechangeinturkey.com/climate-change-basics-how-does-the-climate-of-turke
Hydrobiologia 778 (1), 295–312. ychange.html. (Accessed 13 August 2020).
Calvi, W.S., 1936. Çev. Şevket ahmet birand), türkiye cumhuriyetinde jeolojik görümler, Şener, E., Davraz, A., Şener, S., 2010. Investigation of Akşehir and Eber lakes (SW
ankara’nın su vaziyeti, (birinci kısım), Ankara yüksek ziraat enstitüsü. Sayı 20, 35 s. Turkey) coastline change with multitemporal satellite images. Water Resour. Manag.
Çatal, A., 2015. Akşehir gölünün Akşehir çölüne dönüşü süreci ve etki eden faktörler. 24, 727–745.
Toprak Su Dergisi 4 (1), 18–26. Temiz, F., Durduran, S.S., 2016. Monitoring coastline change using remote sensing and
Choobari, O.A., Ahmedi-Givi, F., Mirzaevi, N., Owlad, E., 2016. Climate change and GIS technology: a case study of Acıgöl Lake, Turkey. In: IOP Conference Series: Earth
anthropogenic impacts on the rapid shrinkage of Lake Urmia. Int. J. Climatol. 36 and Environmental Science, vol. 44. IOP Publishing, 042033. No. 4.
(13), 4276–4286. https://doi.org/10.1002/joc.4630. UKAM, 1998. Mogan gölü (Gölbaşı-Ankara) dip çamurunun i?ncelenmesi projesi.
Dalfes, H.N., Karaca, M., Sen, O.L., 2007. Climate change scenarios for Turkey. In: Hacettepe Üniv.
Çağlar, G. (Ed.), Climate Change & Turkey-impacts, Sectoral Analyses, Socio- USGS-NGA-NASA, 2015. Shuttle Radar Topography Mission 1 Arc-Second Global:
Economic Dimensions. United Nations Development Programme Turkey Office, SRTM1N37E036V3. U.S. Geological Survey (USGS) Earth Resources Observation and
Turkey. Science (EROS) Center, Sioux Falls, South Dakota. https://lta.cr.usgs.gov/SRTM1Arc
Dervişoğlu, A., Musaoğlu, N., Tanık, A., Şeker, D.Z., Kaya, S., 2017. Satellite-based . (Accessed 26 October 2016).
temporal assessment of a dried lake: case study of akgol wetland. Fresenius Environ. Xu, Y., Ramanathan, V., Victor, D.G., 2018. Global warming will happen faster than we
Bull. 26 (1/2017), 352–359. think. Nature 564 (7734), 30–32. https://doi.org/10.1038/d41586-018-07586-5.
Dişli, E., 2007. Eymir-Mogan Gölleri arasında yeraltısuyu akım ve Kütle taşınım modeli, Yağbasan, Ö., 2007. Modeling of Mogan Eymir Lakes Aquifer System, Thesis of Doctor of
Doktora tezi, Hacettepe Üniv. In: Jeoloji (Hidrojeoloji) Mühendisliği Anabilim Dalı, Philosophy. The Middle East Tecnical University (METU), Ankara, p. 163 s.
168 s (Ankara). Yağbasan, Ö., Yazıcıgil, H., 2009. Sustainable management of mogan and Eymir lakes in
DMI, 2019. Ankara Meteoroloji İstasyonu (No: 17130) 1926-2019 Aylık Meteorolojik Central Turkey. Environ. Geol. 56 (6), 1029–1040.
Verileri, Devlet Meteoroloji Genel Müdürlüğü (Turkish State Meteorological Service) Yağbasan, Ö., Yazıcıgil, H., 2012. Assessing the impact of climate change on Mogan and
yayımlanmamış. Eymir Lakes’ levels in Central Turkey. Environ Earth Sci. https://doi.org/10.1007/
Dokulil, M.T., 2014. Impact of climate warming on European inland waters. Inland s12665-011-1209-3.
Waters 4 (1), 27–40.

13