Investigating Hydroclimatic Variables Trends on the Natural Lakes of Western Greece Using Earth Observation Data

Abstract

Expected global climate change is allegedly becoming more intense, and the impacts on water resources are being tracked in various hydroclimatic regimes. The present research investigates a hydrologically important area of Greece, where four natural lakes are concentrated. It aims to quantify any potential long-term trends in lake water area, precipitation, and temperature timeseries. Water area timeseries spanning four decades are estimated by the mNDWI from Landsat satellite imagery and used as an index of each lake’s water storage. Precipitation and temperature measurements are obtained from the open access datasets Hydroscope and ERA5-Land, respectively. All of the timeseries were tested seasonally and annually with the Pettitt and Mann–Kendal tests for statistically significant breakpoints and trends detection. No timeseries analysis resulted in a statistically significant (at 0.05 or 0.1 levels) annual or seasonal trend. The hydroclimatic regime over the past forty years in western Greece is found to have been relatively stable. Land use was also assessed to have been relatively unchanging, converging to the overall stability of the local water regime. However, the findings of this research should not be interpreted as a reassurance against climate change, but as a call to further research for the detailed regional and local assessment of climate change and hydroclimatic variability with acknowledged statistical approaches.

Full text

Citation: Gourgouletis, N.; Baltas, E.
Investigating Hydroclimatic
Variables Trends on the Natural
Lakes of Western Greece Using Earth
Observation Data. Sensors 2023,23,
2056. https://doi.org/10.3390/
s23042056
Academic Editor: Ronghua Ma
Received: 22 December 2022
Revised: 4 February 2023
Accepted: 8 February 2023
Published: 11 February 2023
Copyright: © 2023 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
sensors
Article
Investigating Hydroclimatic Variables Trends on the Natural
Lakes of Western Greece Using Earth Observation Data
Nikolaos Gourgouletis and Evangelos Baltas *
Department of Water Resources & Environmental Engineering, School of Civil Engineering,
National Technical University of Athens, Str. Iroon Politexniou 9, 157 80 Zografou, Greece
*Correspondence: baltas@chi.civil.ntua.gr
Abstract:
Expected global climate change is allegedly becoming more intense, and the impacts on
water resources are being tracked in various hydroclimatic regimes. The present research investigates
a hydrologically important area of Greece, where four natural lakes are concentrated. It aims to
quantify any potential long-term trends in lake water area, precipitation, and temperature time-
series. Water area timeseries spanning four decades are estimated by the mNDWI from Landsat
satellite imagery and used as an index of each lake’s water storage. Precipitation and temperature
measurements are obtained from the open access datasets Hydroscope and ERA5-Land, respectively.
All of the timeseries were tested seasonally and annually with the Pettitt and Mann–Kendal tests
for statistically significant breakpoints and trends detection. No timeseries analysis resulted in a
statistically significant (at 0.05 or 0.1 levels) annual or seasonal trend. The hydroclimatic regime over
the past forty years in western Greece is found to have been relatively stable. Land use was also
assessed to have been relatively unchanging, converging to the overall stability of the local water
regime. However, the findings of this research should not be interpreted as a reassurance against
climate change, but as a call to further research for the detailed regional and local assessment of
climate change and hydroclimatic variability with acknowledged statistical approaches.
Keywords: climate change; lakes; remote sensing; water area; trend analysis; ERA5; Landsat
1. Introduction
Natural lakes and manmade reservoirs play a key role in the Earth’s water cycle, satis-
fying the water demands of human activities along with important ecosystem services [
1
–
4
].
They also are important indexes of climate conditions, trends, and abrupt changes, with
deviations of their physical, chemical, and biological features [
1
,
5
]. Lakes and reservoirs
regulate river flows, mitigate floods, provide groundwater recharge flows, improve and
stabilize water quality, as well as frequently serving as the primary water supply source for
human use [
3
,
6
]. Additionally, lakes and reservoirs hold a significant share in the global
carbon cycle and contribute to methane emissions [1,3,7]
Climate change is projected to augment the pressures and intensify the impacts on
water resources, especially in the mid latitudes—Mediterranean region [
8
–
11
]. Thus, alter-
ations of the water cycle are expected, highlighting its importance on providing ecosystem
services and regulating extreme flows [
8
,
9
]. Furthermore, land cover/use changes subject
to economic growth, increased livestock, increased need of pasture and agricultural pro-
duction, as well as increased water demands, constitute another key pressure on water
resources [2,12–15].
Greece is a typical Mediterranean country, where in recent decades, land resources are
documented to be overexploited, while noteworthy shares of land have been transformed
into artificial surfaces, intensifying the pressures on water and land resources [
16
]. More-
over, after the mid-20th century, an increasing number of droughts is being recorded in
Greece, and relevant studies project an intensification of the drought phenomena, along
Sensors 2023,23, 2056. https://doi.org/10.3390/s23042056 https://www.mdpi.com/journal/sensors

Sensors 2023,23, 2056 2 of 19
with incremental water stress [
17
–
19
]. It is worth noting a drought occurrence that extended
to the whole country in 1990, which is considered the most severe of the 20th century [
20
].
By the time of the drought’s peak, in December of 1990, the water reserves of the country’s
capital, Athens, were expected to only last for a few weeks more [21].
Considering the issues mentioned above, the quantitative pressures exerted on water
resources are expected to increase in the near future. The latter will also be required to
contribute to an increasing demand of the human environment, as local regulators of climate
change and as pillars sustaining ecosystems. A major role of any toolkit of sustainable water
resource management in defence of water security, is stated to be their monitoring [
22
–
24
].
Water resource monitoring is feasible with in-situ observations, hydrological modelling, and
remote sensing techniques [
25
]. From the 1990s, remote sensing approaches towards the
monitoring of inland waters have been receiving increasing consideration from researchers
and governance bodies. The most important aspects of the lakes’ and reservoirs’ water
cycles, their water levels, water areas, and water storages have been attracting the highest
research interest.
The Normalized Difference Water Index (NDWI) was introduced in 1996, and con-
stitutes the most widely adopted index for delineating water features from satellite im-
agery [
26
–
30
]. However, the modified NDWI (mNDWI) has gained significant approval,
and is considered to be more accurate in the distinction of water bodies [
31
–
36
]. The
modification consists of replacing the Near Infrared Band spectrum used in the NDWI,
with a higher bandwidth, middle infrared band. Thus, the mNDWI has a higher accuracy
and less noise, when delineating open water features, compared to the original NDWI [
37
].
Since the first launch of the Landsat-1 satellite in 1972, environmental earth observation
entered a new phase. Optical satellite imagery provides an almost 50 year-long record of
important measurements over water bodies. Although altimetry missions became adequate
sources of water level information a few decades ago, water area estimations are feasible
for a much longer timeseries, constituting a significant source of information for long-term
hydroclimatic analysis [14].
Accordingly, this research aims to assess the long-term trends in hydroclimatic vari-
ables in western Greece, where several natural lakes are located. Despite the large amounts
of available consistent remote sensing data concerning quantitative indices of the water
cycle, only a couple of studies have focused on these aspects of the lake water cycle, e.g.,
water area, water level, or water storage [
30
,
38
]. However, there have been several ap-
plications of remote sensing techniques on the qualitative characteristics of inland water
bodies [
39
–
46
]. Hence, this research aims to target the long-term climatic trends of the natu-
ral lakes of western Greece, Trichonis, Lysimacheia, Ozeros, and Amvrakia, using remote
sensing water area estimations as the main water content indicator. The abovementioned
choice, of the water area as an indirect index of water storage, is considered justified due
to the converging conclusions of various studies towards the high correlation between
water area and water storage timeseries [
2
,
24
,
25
,
30
,
47
–
50
]. Moreover, this research aims to
analyse the simultaneous trends of importance to the water cycle climate variables, such
as precipitation and temperature, as well as human interference in the land use pattern of
each lake basin.
2. Materials and Methods
2.1. Study Area
The four natural lakes, Trichonis, Lysimacheia, Ozeros, and Amvrakia are located
in western Greece, a rich hydrologic part of Greece, and are presented along with their
corresponding basins in Figure 1. All four lakes belong to the River Basin District of
western Greece, which is characterised by the second highest average precipitation in the
country. Specifically, low elevation regions receive 800–1000 mm annual precipitation,
while mountainous regions receive 1400–1800 mm of annual precipitation. The dominant
land cover/land uses are forests (> 45%) and pastures (> 30%), followed by crop cultivation
and roads [51].

Sensors 2023,23, 2056 3 of 19
Sensors 2023, 23, x FOR PEER REVIEW 3 of 19
cover/land uses are forests (> 45%) and pastures (> 30%), followed by crop cultivation and
roads [51].
Figure 1. Upper: Basins of lakes Trichonis, Lysimacheia, Ozeros, and Amvrakia, and their locations
in Greece. Lower: Detailed view of the above lakes (LWB) and basins in SW part of western Greece
with a digital elevation model background.
Lake Trichonis is the largest lake in Greece, with a nominal water area of 96.51 km
2
and water perimeter of around 53.5 km. It is a protected area of aquatic species of eco-
nomic importance under the Directive 2006/44/EC, due to its significant production of
smelt. Its annual inflows are 254 hm
3
from its hydrologic basin, as well as 131.5 hm
3
com-
ing from a diversion channel from the hydroelectric reservoir Stratos II. About 329.6 hm
3
on average are diverted from lake Trichonis to lake Lysimacheia, while the direct water
abstractions from lake Trichonis account for 12.96 hm
3
. Under the 1st Revision of the River
Basin Management Plan of Western Greece, lake Trichonis is considered to have a good
ecological and chemical status. Lake Trichonis is considered to have an average depth of
30.5 m and a maximum depth of 57.0 m, about 40 m below sea level [52,53]. Moreover,
lake Trichonis is indexed as a NATURA 2000 protection area, as an area of high potential
research value [52].
Lake Lysimacheia has a nominal water area of 13.04 km
2
and water perimeter of
around 22.9 km. Its annual inflows are 123 hm
3
from its hydrologic basin, as well as 329.6
hm
3
coming from a diversion channel from lake Trichonis. About 264.7 hm
3
on average
are diverted from lake Lysimacheia to the nearby Acheloos river basin, while the direct
water abstractions from lake Lysimacheia account for 6.3 hm
3
. Under the 1st Revision of
Figure 1. Upper
:Basins of lakes Trichonis, Lysimacheia, Ozeros, and Amvrakia, and their locations
in Greece.
Lower
:Detailed view of the above lakes (LWB) and basins in SW part of western Greece
with a digital elevation model background.
Lake Trichonis is the largest lake in Greece, with a nominal water area of 96.51 km
2
and water perimeter of around 53.5 km. It is a protected area of aquatic species of economic
importance under the Directive 2006/44/EC, due to its significant production of smelt. Its
annual inflows are 254 hm
3
from its hydrologic basin, as well as 131.5 hm
3
coming from a
diversion channel from the hydroelectric reservoir Stratos II. About 329.6 hm
3
on average
are diverted from lake Trichonis to lake Lysimacheia, while the direct water abstractions
from lake Trichonis account for 12.96 hm
3
. Under the 1st Revision of the River Basin
Management Plan of Western Greece, lake Trichonis is considered to have a good ecological
and chemical status. Lake Trichonis is considered to have an average depth of 30.5 m and a
maximum depth of 57.0 m, about 40 m below sea level [
52
,
53
]. Moreover, lake Trichonis is
indexed as a NATURA 2000 protection area, as an area of high potential research value [
52
].
Lake Lysimacheia has a nominal water area of 13.04 km
2
and water perimeter of
around 22.9 km. Its annual inflows are 123 hm
3
from its hydrologic basin, as well as
329.6 hm
3
coming from a diversion channel from lake Trichonis. About 264.7 hm
3
on
average are diverted from lake Lysimacheia to the nearby Acheloos river basin, while the
direct water abstractions from lake Lysimacheia account for 6.3 hm
3
. Under the 1st Revision
of the River Basin Management Plan of Western Greece, lake Lysimacheia is considered to
have a moderate ecological and good chemical status.
Lake Ozeros has a nominal water area of 9.39 km
2
and water perimeter of around
13.6 km. Its annual inflows are 24.3 hm
3
from its hydrologic basin. Lake Ozeros is not

Sensors 2023,23, 2056 4 of 19
considered to receive any water abstractions. Under the 1st Revision of the River Basin Man-
agement Plan of Western Greece, lake Ozeros is considered to have a moderate ecological
and good chemical status.
Lake Amvrakia has a nominal water area of 14.53 km
2
and water perimeter of around
34.35 km. Its annual inflows are 84.0 hm
3
from its hydrologic basin. Lake Amvrakia is not
considered to receive any known water abstractions. However, there is an evident decrease
of its water content in its north compartment, which has not been attributed to a specific
cause. Under the 1st Revision of the River Basin Management Plan of Western Greece, lake
Ozeros is considered to have a good ecological and chemical status [51,54,55].
2.2. Data Used
2.2.1. Landsat Family IMAGERY
Satellite images from the Landsat missions 4, 5, 7, 8, and 9, covering the four lakes’
basins, were collected from the United States Geological Survey (USGS) online portal
(https://earthexplorer.usgs.gov/, accessed on 1 December 2022). Landsat 4, launched in
1982, was the first environmental remote sensing satellite to provide band products at the
resolution of 30 m, which is sufficient to capture the examined lakes. Thus, the selected
dataset covers the temporal period between 1984–2022, with a spatial resolution of 30 m,
for every Landsat mission mentioned. The selection of 1984 as a start year is based on
satellite products’ availability and temporal consistency. Moreover, the selected dataset had
cloud covered images removed, resulting in 375 satellite images throughout the examined
years, which leaves approximately 10 images per year. Regarding Landsat 7’s SLC failure,
the present research followed the gap-filling methodology prescribed by the USGS [
56
].
Regarding the rest of the Landsat missions, 4, 5, 8, and 9, it should be noted that their
products selected are Level-2, therefore already preprocessed and ready to use for the scope
of the present research.
2.2.2. In-Situ Precipitation Measurements
In-situ daily precipitation measurements for the four lakes’ basins were collected
from the “Hydroscope” project’s online portal (http://www.hydroscope.gr/, accessed on 1
December 2022). Fourteen hydrometeorological stations were selected within the basins of
the four lakes and their precipitation data were downloaded from the portal. The stations’
locations are Analipsi, Gavalou, Thermo, Kallithea, Lepenou, Poros Riganiou, Stamna,
Ag. Vlasios, Achira, Monastirakion, Sargiada, Stanos, Lesinio, and Trikorfo. Their altitude
varies from 2 m.a.s.l. to 852 m.a.s.l., and they are all operated by the Hellenic Ministry of
Environment and Energy. The available datasets cover almost the entirety of the water area
timeseries, containing the years from 1984 until 2016–2019.
2.2.3. ERA5-Land Monthly Temperature Data
ERA5-Land reanalysis data for the four lakes’ basins were collected from the EU
Copernicus Climate Change Service online portal (https://cds.climate.copernicus.eu, ac-
cessed on 1 December 2022). The ERA5-Land reanalysis dataset provides 2 m above ground
temperature data at a maximum temporal resolution of one hour, and a spatial resolution
of 0.1 degree. For the purpose of the present research, the downloaded data consisted of the
monthly average temperatures, containing the years from 1984 until 2022, fully covering
the water area timeseries. ERA5-Land reanalysis products are considered ideal in cases of
scarce in-situ data, as in the case of western Greece where there is a lack of open access
temperature data. Moreover, ERA5-Land data have a growing usage in hydrology and
remote sensing applications [
57
,
58
], thus they are considered suitable for the purposes of
the present research.
2.2.4. CORINE Land Cover (CLC)
Land cover data for the four lakes’ basins were collected from the EU Coperni-
cus Land Monitoring Service online portal (https://land.copernicus.eu/, accessed on

Sensors 2023,23, 2056 5 of 19
1 December 2022
). The CLC dataset is characterized by a minimum mapping unit of 25 ha
and a minimum mapping width of 100 m [
59
]. Two datasets covering western Greece were
downloaded from the abovementioned portal, one covering the time period around year
1990 and the other covering the time period around year 2018.
2.3. Methodology
2.3.1. Water Area Extraction
For the calculation of the mNDWI, the Green and SWIR bands were used and the water
and non-water pixels were distinguished following Equation (1) [
32
–
35
]. The mNDWI is
characterized by a spatial resolution of 30 m, deriving from the resolution of the Landsat
bands used. For each Landsat mission used, the different bands’ names representing Green
and SWIR are presented in Table 1.
mNDWI =Green −SWIR
Green +SWIR (1)
Table 1. Landsat 4, 5, 7, 8, and 9 missions’ Green and SWIR bands used.
Landsat Green [µm] SWIR [µm]
4, 5—Thematic Mapper B2 (0.52–0.60) B5 (1.55–1.75)
7—ETM+ B2 (0.52–0.60) B5 (1.55–1.75)
8, 9—OLI B3 (0.53–0.59) B6 (1.57–1.65)
Moreover, the nominal water–non water threshold value of 0, was substituted from an
optimum threshold following the minimum thresholding method, initially proposed by
Prewitt and Mendelsohn [
60
]. The selected approach has been proven to be more accurate
than other thresholding approaches [
29
], and has also been tested by the authors for the
calculation of NDWI in Yliki reservoir [
30
]. Each date’s initial mNDWI’s histogram is
smoothed until it shows only two local maxima, and the optimum threshold lies in the
minimum value between them. The extracted wet perimeter is converted to a lake water
polygon and the area is calculated for each date.
2.3.2. Statistical Point of Change and Trend Analysis
A widely accepted approach to detect statistically significant trends and points of
abrupt change in hydrological variables is the combination of the
Mann–Kendall [61,62]
and Pettitt [
63
] tests, respectively. The abovementioned statistical tests have been applied
extensively to hydroclimatic variables, including discharge timeseries [
64
–
66
], lake water
level timeseries [
14
], as well as climatic variables including precipitation and tempera-
ture [14,64–67].
The statistic value U
t,N
of a timeseries x
t
(t = 1, 2, 3,
. . .
, N) is calculated as shown in
Equation (2).
Ut,N =Ut−1,N +Vt,N ; t =2, 3, . . . , N (2)
where N is the timeseries size and Vt,N as described in Equation (3).
Vt,N =
N
∑
i=1
sgn(xi−xt)(3)
The location of the breakpoint K
N
, which represents the point of abrupt change, is
defined by Equation (4) [63].
KN=Max |Ut,N|; t =1, 2, 3, . . . , N (4)

Sensors 2023,23, 2056 6 of 19
while the statistical significance pof KNis approximated by Equation (5).
p∼
=2×exp"−6×K2
N
N3+N2#(5)
The null hypothesis “H
o
: no change point exists” is rejected at low p-values, and thus
there is a significant point of abrupt change that separates the timeseries into two parts, pre
and post change.
The Mann–Kendall test statistics SMK are defined in Equation (6).
SMK =
N
∑
i
N
∑
j=i+1
sgnxj−xi(6)
For N
≥
8, S
MK
follows an approximately normal distribution and its variance,
σ2
MK
,
is calculated as in Equation (7).
σ2
MK =N×(N−1)×(2×N+5)
18 (7)
Finally, the Mann–Kendall main test statistic ZMK is defined in Equation (8).
ZMK =






SMK−1
σ2
MK
, SMK >0
0, SMK =0
SMK+1
σ2
MK
, SMK <0
(8)
The null hypothesis “H
o
: no trend detected” can be rejected at 5% significance level,
p-value
< 0.05, when |Z
MK
| > 1.96. If the above is true, then a positive value of Z
MK
indicates a positive (incremental) trend, while a negative value of Z
MK
indicates a nega-
tive trend.
The Mann–Kendall test, as defined in Equations (6)–(8), does not account for season-
ality in the examined data. Hence, since many hydrologic and climatic variables display
seasonality, an alteration of the MK test is applied. The alteration, proposed by Hirsch and
Slack [
68
], is defined again by Equations (6)–(8), by substituting N (total timeseries sample
size) with n
k
, which stands for the measurements size in season k. For the purposes of the
present research, two seasons are assessed, the dry hydrological period (May to October)
and the wet hydrological period (November to April).
3. Results
3.1. Water Area Timeseries and Trends
3.1.1. Water Area Timeseries
The extraction of the lakes’ water areas, spanning from 1984 to mid-2022, resulted in
377, 376, 379, and 369 water area measurements for lakes Trichonis, Lysimacheia, Ozeros,
and Amvrakia, respectively.
Lake Trichonis’ water area measurements are presented in Figure 2. The maximum
area observed is 93.43 km
2
, whilst the minimum area observed is 88.60 km
2
. The average
water area of lake Trichonis during the observed period of 39 years is found to be 92.16 km
2
.
During the period examined, lake Trichonis shows small percentages of water area changes,
hence the maximum water extent is 1.38% larger than the average, and the minimum is
3.86% smaller.

Sensors 2023,23, 2056 7 of 19
Sensors 2023, 23, x FOR PEER REVIEW 7 of 19
significant trend can be illustrated in the moving average trendline (red colored) shown
in Figure 2.
Figure 2. Landsat family derived lake Trichonis water area timeseries (red line; 2-year moving av-
erage).
Lake Lysimacheia’s water area measurements are presented in Figure 3. The maxi-
mum area observed is 14.42 km2, whilst the minimum area observed is 7.00 km2. The av-
erage water area of lake Lysimacheia during the observed period of 39 years is found to
be 10.33 km2. During the period examined, lake Lysimacheia shows significant water area
changes, with the maximum water extent being 39.61% larger than the average, and the
minimum being 32.37% smaller.
Regarding the Pettitt and Mann–Kendall tests conducted on the timeseries, there
does not exist a statistically significant point of change or trend, hence the calculated val-
ues of p ≈ 1 and ZMK = −5.07. The absence of a statistically significant trend is illustrated in
the moving average trendline (red colored) in Figure 3.
Figure 3. Landsat family derived lake Lysimacheia water area timeseries (red line; 2-year moving
average).
Lake Ozeros’ water area measurements are presented in Figure 4. The maximum area
observed is 9.43 km2, whilst the minimum area observed is 8.60 km2. The average water
area of lake Ozeros during the observed period of 39 years is found to be 9.00 km2. During
the period examined, lake Ozeros shows relatively small water area changes, with the
88.0
89.0
90.0
91.0
92.0
93.0
94.0
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
Area [km2]
Year
Lake Trichonis Total Area
6.0
7.0
8.0
9.0
10.0
11.0
12.0
13.0
14.0
15.0
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
Area [km2]
Year
Lake Lysimacheia Total Area
Figure 2.
Landsat family derived lake Trichonis water area timeseries (red line; 2-year moving average).
Regarding the Pettitt and Mann–Kendall tests conducted on the timeseries, there
does not exist a statistically significant trend, although a breakpoint is identified in 2014.
The calculated value of pis equal to 0.00 and Z
MK
=
−
0.33. The absence of a statistically
significant trend can be illustrated in the moving average trendline (red colored) shown in
Figure 2.
Lake Lysimacheia’s water area measurements are presented in Figure 3. The maximum
area observed is 14.42 km
2
, whilst the minimum area observed is 7.00 km
2
. The average
water area of lake Lysimacheia during the observed period of 39 years is found to be
10.33 km
2
. During the period examined, lake Lysimacheia shows significant water area
changes, with the maximum water extent being 39.61% larger than the average, and the
minimum being 32.37% smaller.
Sensors 2023, 23, x FOR PEER REVIEW 7 of 19
significant trend can be illustrated in the moving average trendline (red colored) shown
in Figure 2.
Figure 2. Landsat family derived lake Trichonis water area timeseries (red line; 2-year moving av-
erage).
Lake Lysimacheia’s water area measurements are presented in Figure 3. The maxi-
mum area observed is 14.42 km2, whilst the minimum area observed is 7.00 km2. The av-
erage water area of lake Lysimacheia during the observed period of 39 years is found to
be 10.33 km2. During the period examined, lake Lysimacheia shows significant water area
changes, with the maximum water extent being 39.61% larger than the average, and the
minimum being 32.37% smaller.
Regarding the Pettitt and Mann–Kendall tests conducted on the timeseries, there
does not exist a statistically significant point of change or trend, hence the calculated val-
ues of p ≈ 1 and ZMK = −5.07. The absence of a statistically significant trend is illustrated in
the moving average trendline (red colored) in Figure 3.
Figure 3. Landsat family derived lake Lysimacheia water area timeseries (red line; 2-year moving
average).
Lake Ozeros’ water area measurements are presented in Figure 4. The maximum area
observed is 9.43 km2, whilst the minimum area observed is 8.60 km2. The average water
area of lake Ozeros during the observed period of 39 years is found to be 9.00 km2. During
the period examined, lake Ozeros shows relatively small water area changes, with the
88.0
89.0
90.0
91.0
92.0
93.0
94.0
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
Area [km2]
Year
Lake Trichonis Total Area
6.0
7.0
8.0
9.0
10.0
11.0
12.0
13.0
14.0
15.0
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
Area [km2]
Year
Lake Lysimacheia Total Area
Figure 3.
Landsat family derived lake Lysimacheia water area timeseries (red line; 2-year moving
average).
Regarding the Pettitt and Mann–Kendall tests conducted on the timeseries, there does
not exist a statistically significant point of change or trend, hence the calculated values of
p≈1
and Z
MK
=
−
5.07. The absence of a statistically significant trend is illustrated in the
moving average trendline (red colored) in Figure 3.
Lake Ozeros’ water area measurements are presented in Figure 4. The maximum
area observed is 9.43 km
2
, whilst the minimum area observed is 8.60 km
2
. The average
water area of lake Ozeros during the observed period of 39 years is found to be 9.00 km
2
.
During the period examined, lake Ozeros shows relatively small water area changes, with

Sensors 2023,23, 2056 8 of 19
the maximum water extent being 4.81% larger than the average, and the minimum being
4.43% smaller.
Sensors 2023, 23, x FOR PEER REVIEW 8 of 19
maximum water extent being 4.81% larger than the average, and the minimum being
4.43% smaller.
Regarding the Pettitt and Mann–Kendall tests conducted on the timeseries, there
does not exist a statistically significant point of change or trend, hence the calculated val-
ues of p ≈ 1 and ZMK = +0.36. The absence of a statistically significant trend can be seen in
the moving average trendline (red colored) shown in Figure 4.
Figure 4. Landsat family derived lake Ozeros water area timeseries (red line; 2-year moving aver-
age).
Lake Amvrakia’s water area measurements are presented in Figure 5. The maximum
area observed is 13.24 km2, whilst the minimum area observed is 8.91 km2. The average
water area of lake Amvrakia during the observed period of 39 years is found to be 11.10
km2. During the period examined, lake Amvrakia shows moderate water area changes,
with the maximum water extent being 19.34% larger than the average, and the minimum
being 19.71% smaller.
Regarding the Pettitt and Mann–Kendall tests conducted on the timeseries, there
does not exist a statistically significant point of change or trend, hence the calculated val-
ues of p ≈ 1 and ZMK = +1.87. The absence of a statistically significant trend can be seen in
the moving average trendline (red colored) shown in Figure 5.
Figure 5. Landsat family derived lake Amvrakia water area timeseries (red line; 2-year moving av-
erage).
8.5
8.7
8.9
9.1
9.3
9.5
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
Area [km2]
Year
Lake Ozeros Total Area
8.0
9.0
10.0
11.0
12.0
13.0
14.0
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
Area [km2]
Year
Lake Amvrakia Total Area
Figure 4.
Landsat family derived lake Ozeros water area timeseries (red line; 2-year moving average).
Regarding the Pettitt and Mann–Kendall tests conducted on the timeseries, there does
not exist a statistically significant point of change or trend, hence the calculated values of
p
≈
1 and Z
MK
= +0.36. The absence of a statistically significant trend can be seen in the
moving average trendline (red colored) shown in Figure 4.
Lake Amvrakia’s water area measurements are presented in Figure 5. The maximum
area observed is 13.24 km
2
, whilst the minimum area observed is 8.91 km
2
. The average
water area of lake Amvrakia during the observed period of 39 years is found to be 11.10 km
2
.
During the period examined, lake Amvrakia shows moderate water area changes, with
the maximum water extent being 19.34% larger than the average, and the minimum being
19.71% smaller.
Sensors 2023, 23, x FOR PEER REVIEW 8 of 19
maximum water extent being 4.81% larger than the average, and the minimum being
4.43% smaller.
Regarding the Pettitt and Mann–Kendall tests conducted on the timeseries, there
does not exist a statistically significant point of change or trend, hence the calculated val-
ues of p ≈ 1 and ZMK = +0.36. The absence of a statistically significant trend can be seen in
the moving average trendline (red colored) shown in Figure 4.
Figure 4. Landsat family derived lake Ozeros water area timeseries (red line; 2-year moving aver-
age).
Lake Amvrakia’s water area measurements are presented in Figure 5. The maximum
area observed is 13.24 km2, whilst the minimum area observed is 8.91 km2. The average
water area of lake Amvrakia during the observed period of 39 years is found to be 11.10
km2. During the period examined, lake Amvrakia shows moderate water area changes,
with the maximum water extent being 19.34% larger than the average, and the minimum
being 19.71% smaller.
Regarding the Pettitt and Mann–Kendall tests conducted on the timeseries, there
does not exist a statistically significant point of change or trend, hence the calculated val-
ues of p ≈ 1 and ZMK = +1.87. The absence of a statistically significant trend can be seen in
the moving average trendline (red colored) shown in Figure 5.
Figure 5. Landsat family derived lake Amvrakia water area timeseries (red line; 2-year moving av-
erage).
8.5
8.7
8.9
9.1
9.3
9.5
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
Area [km2]
Year
Lake Ozeros Total Area
8.0
9.0
10.0
11.0
12.0
13.0
14.0
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
Area [km2]
Year
Lake Amvrakia Total Area
Figure 5.
Landsat family derived lake Amvrakia water area timeseries (red line; 2-year moving average).
Regarding the Pettitt and Mann–Kendall tests conducted on the timeseries, there does
not exist a statistically significant point of change or trend, hence the calculated values of
p
≈
1 and Z
MK
= +1.87. The absence of a statistically significant trend can be seen in the
moving average trendline (red colored) shown in Figure 5.
3.1.2. Annual Water Area Trend Analysis
The annual average of water area for lake Trichonis is presented in Figure 6. It is calcu-
lated that no significant trend is detected, despite a graphically observed slight positive

Sensors 2023,23, 2056 9 of 19
trend. Lake Trichonis shows insignificant interannual water area variations during the
examined 39 years, with the annual water area varying between
−
0.53% and +0.49% from
the annual water area average of 92.17 km
2
. The Pettitt test conducted on the available
annual timeseries does not show a statistically significant point of change. Moreover,
the Mann–Kendall test resulted in p
≈
1 and Z
MK
= 1.84, depicting a nonsignificant posi-
tive trend.
Sensors 2023, 23, x FOR PEER REVIEW 9 of 19
3.1.2. Annual Water Area Trend Analysis
The annual average of water area for lake Trichonis is presented in Figure 6. It is
calculated that no significant trend is detected, despite a graphically observed slight pos-
itive trend. Lake Trichonis shows insignificant interannual water area variations during
the examined 39 years, with the annual water area varying between −0.53% and +0.49%
from the annual water area average of 92.17 km2. The Pettitt test conducted on the availa-
ble annual timeseries does not show a statistically significant point of change. Moreover,
the Mann–Kendall test resulted in p ≈ 1 and ΖΜΚ = 1.84, depicting a nonsignificant positive
trend.
Figure 6. Annual water area average and linear trendline of lake Trichonis (red line; linear trend-
line).
The annual average of the water area for lake Lysimacheia is presented in Figure 7.
It is calculated that no significant trend is observed, despite a graphically observed nega-
tive trend. Lake Lysimacheia shows small interannual water area variations during the
examined 39 years, with the annual water area varying between −6.27% and +13.94% from
the annual water area average of 10.35 km2. The Pettitt test conducted on the available
annual timeseries does not show a statistically significant point of change. Moreover, the
Mann–Kendall test resulted in p ≈ 1 and ΖΜΚ = −2.06, depicting a nonsignificant negative
trend.
Figure 7. Annual water area average and linear trendline of lake Lysimacheia (red line; linear trend-
line).
91.2
91.4
91.6
91.8
92.0
92.2
92.4
92.6
92.8
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014
2016
2018
2020
2022
Area [km2]
Year
Lake Trichonis Annual average Area
8.5
9.0
9.5
10.0
10.5
11.0
11.5
12.0
12.5
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014
2016
2018
2020
2022
Area [km2]
Year
Lake Lysimacheia Annual average Area
Figure 6.
Annual water area average and linear trendline of lake Trichonis (red line; linear trendline).
The annual average of the water area for lake Lysimacheia is presented in Figure 7.
It is calculated that no significant trend is observed, despite a graphically observed neg-
ative trend. Lake Lysimacheia shows small interannual water area variations during the
examined 39 years, with the annual water area varying between
−
6.27% and +13.94% from
the annual water area average of 10.35 km
2
. The Pettitt test conducted on the available
annual timeseries does not show a statistically significant point of change. Moreover,
the Mann–Kendall test resulted in p
≈
1 and Z
MK
=
−
2.06, depicting a nonsignificant
negative trend.
Sensors 2023, 23, x FOR PEER REVIEW 9 of 19
3.1.2. Annual Water Area Trend Analysis
The annual average of water area for lake Trichonis is presented in Figure 6. It is
calculated that no significant trend is detected, despite a graphically observed slight pos-
itive trend. Lake Trichonis shows insignificant interannual water area variations during
the examined 39 years, with the annual water area varying between −0.53% and +0.49%
from the annual water area average of 92.17 km2. The Pettitt test conducted on the availa-
ble annual timeseries does not show a statistically significant point of change. Moreover,
the Mann–Kendall test resulted in p ≈ 1 and ΖΜΚ = 1.84, depicting a nonsignificant positive
trend.
Figure 6. Annual water area average and linear trendline of lake Trichonis (red line; linear trend-
line).
The annual average of the water area for lake Lysimacheia is presented in Figure 7.
It is calculated that no significant trend is observed, despite a graphically observed nega-
tive trend. Lake Lysimacheia shows small interannual water area variations during the
examined 39 years, with the annual water area varying between −6.27% and +13.94% from
the annual water area average of 10.35 km2. The Pettitt test conducted on the available
annual timeseries does not show a statistically significant point of change. Moreover, the
Mann–Kendall test resulted in p ≈ 1 and ΖΜΚ = −2.06, depicting a nonsignificant negative
trend.
Figure 7. Annual water area average and linear trendline of lake Lysimacheia (red line; linear trend-
line).
91.2
91.4
91.6
91.8
92.0
92.2
92.4
92.6
92.8
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014
2016
2018
2020
2022
Area [km2]
Year
Lake Trichonis Annual average Area
8.5
9.0
9.5
10.0
10.5
11.0
11.5
12.0
12.5
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014
2016
2018
2020
2022
Area [km2]
Year
Lake Lysimacheia Annual average Area
Figure 7.
Annual water area average and linear trendline of lake Lysimacheia (red line; linear trendline).
The annual average of water area for lake Ozeros is presented in Figure 8. It is clearly
depicted that no significant trend is observed. Lake Ozeros shows small interannual
water area variations during the examined 39 years, with the annual water area varying
between
−
3.14% and +2.31% from the annual water area average of 9.0 km
2
. The Pettitt test
conducted on the available annual timeseries does not show a statistically significant point

Sensors 2023,23, 2056 10 of 19
of change. Moreover, the Mann–Kendall test resulted in p
≈
1 and Z
MK
=
−
0.60, showing a
nonsignificant negative trend.
Sensors 2023, 23, x FOR PEER REVIEW 10 of 19
The annual average of water area for lake Ozeros is presented in Figure 8. It is clearly
depicted that no significant trend is observed. Lake Ozeros shows small interannual water
area variations during the examined 39 years, with the annual water area varying between
−3.14% and +2.31% from the annual water area average of 9.0 km2. The Pettitt test con-
ducted on the available annual timeseries does not show a statistically significant point of
change. Moreover, the Mann–Kendall test resulted in p ≈ 1 and ΖΜΚ = −0.60, showing a
nonsignificant negative trend.
Figure 8. Annual water area average and linear trendline of lake Ozeros (red line; linear trendline).
The annual average of water area for lake Amvrakia is presented in Figure 9. It is
calculated that no significant trend is observed, despite a graphically observed slight neg-
ative trend. Lake Amvrakia shows moderate interannual water area variations during the
examined 39 years, with the annual water area varying between −16.62% and +13.69%
from the annual water area average of 11.2 km2. The Pettitt test conducted on the available
annual timeseries does not show a statistically significant point of change. Moreover, the
Mann–Kendall test resulted in p ≈ 1 and ΖΜΚ = −0.73, showing a nonsignificant negative
trend.
Figure 9. Annual water area average and linear trendline of lake Amvrakia (red line; linear trend-
line).
8.0
8.5
9.0
9.5
10.0
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014
2016
2018
2020
2022
Area [km2]
Year
Lake Ozeros Annual average Area
8.0
9.0
10.0
11.0
12.0
13.0
14.0
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014
2016
2018
2020
2022
Area [km2]
Year
Lake Amvrakia Annual average Area
Figure 8. Annual water area average and linear trendline of lake Ozeros (red line; linear trendline).
The annual average of water area for lake Amvrakia is presented in Figure 9. It
is calculated that no significant trend is observed, despite a graphically observed slight
negative trend. Lake Amvrakia shows moderate interannual water area variations during
the examined 39 years, with the annual water area varying between
−
16.62% and +13.69%
from the annual water area average of 11.2 km
2
. The Pettitt test conducted on the available
annual timeseries does not show a statistically significant point of change. Moreover,
the Mann–Kendall test resulted in p
≈
1 and Z
MK
=
−
0.73, showing a nonsignificant
negative trend.
Sensors 2023, 23, x FOR PEER REVIEW 10 of 19
The annual average of water area for lake Ozeros is presented in Figure 8. It is clearly
depicted that no significant trend is observed. Lake Ozeros shows small interannual water
area variations during the examined 39 years, with the annual water area varying between
−3.14% and +2.31% from the annual water area average of 9.0 km2. The Pettitt test con-
ducted on the available annual timeseries does not show a statistically significant point of
change. Moreover, the Mann–Kendall test resulted in p ≈ 1 and ΖΜΚ = −0.60, showing a
nonsignificant negative trend.
Figure 8. Annual water area average and linear trendline of lake Ozeros (red line; linear trendline).
The annual average of water area for lake Amvrakia is presented in Figure 9. It is
calculated that no significant trend is observed, despite a graphically observed slight neg-
ative trend. Lake Amvrakia shows moderate interannual water area variations during the
examined 39 years, with the annual water area varying between −16.62% and +13.69%
from the annual water area average of 11.2 km2. The Pettitt test conducted on the available
annual timeseries does not show a statistically significant point of change. Moreover, the
Mann–Kendall test resulted in p ≈ 1 and ΖΜΚ = −0.73, showing a nonsignificant negative
trend.
Figure 9. Annual water area average and linear trendline of lake Amvrakia (red line; linear trend-
line).
8.0
8.5
9.0
9.5
10.0
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014
2016
2018
2020
2022
Area [km2]
Year
Lake Ozeros Annual average Area
8.0
9.0
10.0
11.0
12.0
13.0
14.0
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014
2016
2018
2020
2022
Area [km2]
Year
Lake Amvrakia Annual average Area
Figure 9.
Annual water area average and linear trendline of lake Amvrakia (red line; linear trendline).
3.1.3. Seasonality Trend Analysis
The investigation of water area seasonality in lake Trichonis, during the available
timeseries of 39 years, shows that there is no significant variation between wet and dry
periods. More explicitly, the average water area of the lake for the wet periods is 92.17 km
2
,
while for the dry ones is 92.16 km
2
. Furthermore, following the results of the seasonality
trend analysis, no significant trends or points of change are identified for either period. The
wet periods show a p
≈
1 and Z
SK
= +1.06, depicting a statistically insignificant positive
trend, while the dry periods show a p
≈
1 and Z
SK
= +1.45, depicting again a statistically
insignificant positive trend.

Sensors 2023,23, 2056 11 of 19
Lake Lysimacheia, during the examined timeseries, shows a slight variation between
wet and dry periods. The average wet period water area is calculated at 10.84 km
2
, while the
dry periods show an average water area of 10.07 km
2
. Furthermore, following the results
of the seasonality trend analysis, no significant trends or points of change are identified
for either period. The wet periods show a p
≈
1 and Z
SK
=
−
1.46, depicting a statistically
insignificant negative trend, while the dry periods show a p
≈
1 and Z
SK
=
−
2.85, depicting
a statistically insignificant negative trend.
Lake Ozeros, during the examined timeseries, shows a slight variation between wet
and dry periods. The average wet period water area is calculated at 9.06 km
2
, while the
dry periods show an average water area of 8.97 km
2
. Furthermore, following the results
of the seasonality trend analysis, no significant trends or points of change are identified
for either period. The wet periods show a p
≈
1 and Z
SK
=
−
1.72, depicting a statistically
insignificant negative trend, while the dry periods show a p
≈
1 and Z
SK
= +0.33, depicting
a statistically insignificant positive trend.
Lake Amvrakia, during the examined timeseries, shows a slight variation between wet
and dry periods. The average wet period water area is calculated at 11.19 km
2
, while the
dry periods show an average water area of 11.05 km
2
. Furthermore, following the results
of the seasonality trend analysis, no significant trends or points of change are identified
for either period. The wet periods show a p
≈
1 and Z
SK
=
−
0.04, depicting a statistically
insignificant negative trend, while the dry periods show a p
≈
1 and Z
SK
=
−
0.48, depicting
a statistically insignificant negative trend.
3.2. Further Investigation
3.2.1. Precipitation
Lake Trichonis’ basin’s annual precipitation throughout the available examined years,
1984–2016, is found to be on average 1009 mm. The annual precipitation presents a large
variability, showing a maximum value of 1633 mm and a minimum of 595 mm, with a
standard deviation of 262 mm. The annual precipitation of lake Trichonis’ basin presents
none significant points of change or trends, as the Pettitt and Mann–Kendal tests resulted
in p
≈
1 and Z
MK
= 2.80. Seasonal precipitation trends were also tested and no statistically
significant trend was found in either the wet or dry periods. Moreover, the Pearson
correlation between the annual precipitation and lake area is calculated at 0.05, implying a
negligible relationship between the abovementioned variables.
Lake Lysimacheia’s basin’s annual precipitation throughout the available examined
years, 1984–2016, is found to be on average 954 mm. The annual precipitation presents
a large variability, showing a maximum value of 1325 mm and a minimum of 606 mm,
with a standard deviation of 205 mm. The annual precipitation of lake Lysimacheia’s basin
presents no significant points of change or trends, as the Pettitt and Mann–Kendal tests
resulted in p
≈
1 and Z
MK
= 2.37. Seasonal precipitation trends were also tested and no
statistically significant trend was found in either the wet or dry periods. Moreover, the
Pearson correlation between the annual precipitation and lake area is calculated at
−
0.03,
implying a negligible relationship between the abovementioned variables.
Lake Ozeros’ basin’s annual precipitation throughout the available examined years,
1984–2018, is found to be on average 1028 mm. The annual precipitation presents a large
variability, showing a maximum value of 1400 mm and a minimum of 632 mm, with a
standard deviation of 247 mm. The annual precipitation of lake Ozeros’ basin presents
no significant points of change or trends, as the Pettitt and Mann–Kendal tests resulted
in p
≈
1 and Z
MK
= 3.13. The seasonal precipitation trends have also been tested and no
statistically significant trend was found in either the wet or dry periods. Moreover, the
Pearson correlation between the annual precipitation and lake area is calculated at
−
0.05,
implying a negligible relationship between the abovementioned variables.
Lake Amvrakia’s basin’s annual precipitation throughout the available examined
years, 1984–2018, is found to be on average 963 mm. The annual precipitation presents a
large variability, showing a maximum value of 1410 mm and a minimum of 587 mm, with a

Sensors 2023,23, 2056 12 of 19
standard deviation of 198 mm. The annual precipitation of lake Amvrakia’s basin presents
no significant points of change or trends, as the Pettitt and Mann–Kendal tests resulted
in p
≈
1 and Z
MK
= 1.60. The seasonal precipitation trends have also been tested and no
statistically significant trend was found in either the wet or dry periods. Moreover, the
Pearson correlation between the annual precipitation and lake area is calculated at 0.05,
implying a negligible relationship between the abovementioned variables.
Overall, regarding the four lakes’ basins’ annual and seasonal precipitation, there is no
significant change detected, as far as points of change and trends are concerned. Moreover,
the available timeseries of precipitation and the remotely sensed lake water area, do not
show a significant value of Pearson correlation between them. The annual precipitation
timeseries of the four basins are presented below, in Figure 10.
Sensors 2023, 23, x FOR PEER REVIEW 12 of 19
variability, showing a maximum value of 1400 mm and a minimum of 632 mm, with a
standard deviation of 247 mm. The annual precipitation of lake Ozeros’ basin presents no
significant points of change or trends, as the Pettitt and Mann–Kendal tests resulted in p
≈ 1 and ZMK = 3.13. The seasonal precipitation trends have also been tested and no statisti-
cally significant trend was found in either the wet or dry periods. Moreover, the Pearson
correlation between the annual precipitation and lake area is calculated at −0.05, implying
a negligible relationship between the abovementioned variables.
Lake Amvrakia’s basin’s annual precipitation throughout the available examined
years, 1984–2018, is found to be on average 963 mm. The annual precipitation presents a
large variability, showing a maximum value of 1410 mm and a minimum of 587 mm, with
a standard deviation of 198 mm. The annual precipitation of lake Amvrakia’s basin pre-
sents no significant points of change or trends, as the Pettitt and Mann–Kendal tests re-
sulted in p ≈ 1 and ZMK = 1.60. The seasonal precipitation trends have also been tested and
no statistically significant trend was found in either the wet or dry periods. Moreover, the
Pearson correlation between the annual precipitation and lake area is calculated at 0.05,
implying a negligible relationship between the abovementioned variables.
Overall, regarding the four lakes’ basins’ annual and seasonal precipitation, there is
no significant change detected, as far as points of change and trends are concerned. More-
over, the available timeseries of precipitation and the remotely sensed lake water area, do
not show a significant value of Pearson correlation between them. The annual precipita-
tion timeseries of the four basins are presented below, in Figure 10.
Figure 10. Annual precipitation of the four lakes’ basins.
3.2.2. Temperature
Lake Trichonis’ basin’s annual average temperature throughout the available exam-
ined years, 1984–2021, is found to be on average 15.36 °C. The annual average temperature
presents a moderate variability, showing a maximum value of 16.49 °C and a minimum
of 14.37 °C, with a standard deviation of 0.58 °C. The annual average temperature of lake
Trichonis’ basin presents no significant points of change or trends, as the Pettitt and
Mann–Kendal tests resulted in p ≈ 1 and ZMK = 4.65. The seasonal temperature trends have
also been tested and no statistically significant trend was found in either the wet or dry
periods. Moreover, the Pearson correlation between the annual average temperature and
lake area is calculated at 0.42, implying a weak relationship between the abovementioned
variables.
Lake Lysimacheia’s basin’s annual average temperature throughout the available ex-
amined years, 1984–2021, is found to be on average 15.36 °C. The average temperature
presents a moderate variability, showing a maximum value of 16.49 °C and a minimum
of 14.37 °C, with a standard deviation of 0.58 °C. The annual average temperature of lake
0
500
1000
1500
2000
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
Precipitation [mm]
Year
Lake's basin annual precipitation
Ozeros Lysimacheia Trichonis Amvrakia
Figure 10. Annual precipitation of the four lakes’ basins.
3.2.2. Temperature
Lake Trichonis’ basin’s annual average temperature throughout the available examined
years, 1984–2021, is found to be on average 15.36
◦
C. The annual average temperature
presents a moderate variability, showing a maximum value of 16.49
◦
C and a minimum
of 14.37
◦
C, with a standard deviation of 0.58
◦
C. The annual average temperature of lake
Trichonis’ basin presents no significant points of change or trends, as the Pettitt and Mann–
Kendal tests resulted in p
≈
1 and Z
MK
= 4.65. The seasonal temperature trends have also
been tested and no statistically significant trend was found in either the wet or dry periods.
Moreover, the Pearson correlation between the annual average temperature and lake area
is calculated at 0.42, implying a weak relationship between the abovementioned variables.
Lake Lysimacheia’s basin’s annual average temperature throughout the available
examined years, 1984–2021, is found to be on average 15.36
◦
C. The average temperature
presents a moderate variability, showing a maximum value of 16.49
◦
C and a minimum
of 14.37
◦
C, with a standard deviation of 0.58
◦
C. The annual average temperature of lake
Lysimacheia’s basin presents no significant points of change or trends, as the Pettitt and
Mann–Kendal tests resulted in p
≈
1 and Z
MK
= 4.52. The seasonal temperature trends
have also been tested and no statistically significant trend was found in either the wet or
dry periods. Moreover, the Pearson correlation between the annual average temperature
and lake area is calculated at
−
0.22, implying a weak inverse relationship between the
abovementioned variables.
Lake Ozeros’ basin’s annual average temperature throughout the available examined
years, 1984–2021, is found to be on average 15.36
◦
C. The annual average temperature
presents a large variability, showing a maximum value of 16.49
◦
C and a minimum of
14.37
◦
C, with a standard deviation of 0.58
◦
C. The annual average temperature of lake
Ozeros’ basin presents no significant points of change or trends, as the Pettitt and Mann–

Sensors 2023,23, 2056 13 of 19
Kendal tests resulted in p
≈
1 and Z
MK
= 4.45. The seasonal temperature trends have also
been tested and no statistically significant trend was found in either the wet or dry periods.
Moreover, the Pearson correlation between the annual average temperature and lake area is
calculated at 0.07, implying a negligible relationship between the abovementioned variables.
Lake Amvrakia’s basin’s annual average temperature throughout the available exam-
ined years, 1984–2021, is found to be on average 14.69
◦
C. The annual average temperature
presents a large variability, showing a maximum value of 16.88
◦
C and a minimum of
14.69
◦
C, with a standard deviation of 0.58
◦
C. The annual average temperature of lake
Amvrakia’s basin presents no significant points of change or trends, as the Pettitt and Mann–
Kendal tests resulted in p
≈
1 and Z
MK
= 4.50. The seasonal temperature trends have also
been tested and no statistically significant trend was found in either the wet or dry periods.
Moreover, the Pearson correlation between the annual average temperature and lake area is
calculated at 0.01, implying a negligible relationship between the abovementioned variables.
Overall, regarding the four lakes’ basins’ annual average temperatures, there is no sig-
nificant change detected, as far as points of change and trends are concerned. Moreover, the
available timeseries of temperature and the remotely sensed lake water area, do not show a
significant value of Pearson correlation between them. The annual average temperature
timeseries of the four basins are presented below, in Figure 11.
Sensors 2023, 23, x FOR PEER REVIEW 13 of 19
Lysimacheia’s basin presents no significant points of change or trends, as the Pettitt and
Mann–Kendal tests resulted in p ≈ 1 and ZMK = 4.52. The seasonal temperature trends have
also been tested and no statistically significant trend was found in either the wet or dry
periods. Moreover, the Pearson correlation between the annual average temperature and
lake area is calculated at −0.22, implying a weak inverse relationship between the above-
mentioned variables.
Lake Ozeros’ basin’s annual average temperature throughout the available examined
years, 1984–2021, is found to be on average 15.36 °C. The annual average temperature
presents a large variability, showing a maximum value of 16.49 °C and a minimum of
14.37 °C, with a standard deviation of 0.58 °C. The annual average temperature of lake
Ozeros’ basin presents no significant points of change or trends, as the Pettitt and Mann–
Kendal tests resulted in p ≈ 1 and ZMK = 4.45. The seasonal temperature trends have also
been tested and no statistically significant trend was found in either the wet or dry peri-
ods. Moreover, the Pearson correlation between the annual average temperature and lake
area is calculated at 0.07, implying a negligible relationship between the abovementioned
variables.
Lake Amvrakia’s basin’s annual average temperature throughout the available ex-
amined years, 1984–2021, is found to be on average 14.69 °C. The annual average temper-
ature presents a large variability, showing a maximum value of 16.88 °C and a minimum
of 14.69 °C, with a standard deviation of 0.58 °C. The annual average temperature of lake
Amvrakia’s basin presents no significant points of change or trends, as the Pettitt and
Mann–Kendal tests resulted in p ≈ 1 and ZMK = 4.50. The seasonal temperature trends have
also been tested and no statistically significant trend was found in either the wet or dry
periods. Moreover, the Pearson correlation between the annual average temperature and
lake area is calculated at 0.01, implying a negligible relationship between the abovemen-
tioned variables.
Overall, regarding the four lakes’ basins’ annual average temperatures, there is no
significant change detected, as far as points of change and trends are concerned. Moreo-
ver, the available timeseries of temperature and the remotely sensed lake water area, do
not show a significant value of Pearson correlation between them. The annual average
temperature timeseries of the four basins are presented below, in Figure 11.
Since temperature is considered the climate variable mostly connected to climate
change, there was an additional test conducted. The entire ERA-5 Land monthly average
temperature timeseries are tested for significant points of change and trends detection.
Yet, none are found, at the 95% or 90% confidence interval, as the Pettitt and Mann–Ken-
dal test concluded in p = (0.29, 0.28, 0.34, 0.40) and ZMK = (1.49, 1.49, 1.40, 1.36) for the basins
of Trichonis, Lysimacheia, Ozeros, and Amvrakia, respectively.
Figure 11. Annual average temperature of the four lakes’ basins.
14
14.5
15
15.5
16
16.5
17
1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020
Temperature [oC]
Year
Lake's basin annual average temperature
Ozeros Lysimacheia Trichonis Amvrakia
Figure 11. Annual average temperature of the four lakes’ basins.
Since temperature is considered the climate variable mostly connected to climate
change, there was an additional test conducted. The entire ERA-5 Land monthly average
temperature timeseries are tested for significant points of change and trends detection. Yet,
none are found, at the 95% or 90% confidence interval, as the Pettitt and Mann–Kendal test
concluded in p= (0.29, 0.28, 0.34, 0.40) and Z
MK
= (1.49, 1.49, 1.40, 1.36) for the basins of
Trichonis, Lysimacheia, Ozeros, and Amvrakia, respectively.
3.2.3. Land Cover
The examined land cover datasets, describing years 1990 and 2018, show a generally
stable land use pattern across the four lakes’ basins. Specifically, human environment land
uses, such as motorways and urban areas, gain small portions, from 1% to 10% of each
basin’s total area. Additionally, natural vegetation, such as grasslands and forests, also gain
small portions, accounting for 1% to 3% of each basin’s total area. In all four lakes’ basins,
cultivated areas seem to decline, favoring the abovementioned human environment and
natural vegetation uses. The reduction of cultivated areas ranges from 2% to 12% across
the four lakes’ basins. As no significant change is detected in any lake basin, it is evident
that there will be negligible or small differentiations of the water use regimes across the
four basins. This is confirmed by the water area analysis conducted in Section 3.1, where
no significant change or trend patterns were detected.

Sensors 2023,23, 2056 14 of 19
4. Discussion
Before discussing the key findings of this research, a few constraining factors should
be remarked on as the study limitations. As described in Section 3.1, the four lakes’ water
area are found to have been relatively stable. Only small fluctuations are recorded, ranging
from
−
16% to +16% in lake Amvrakia, and from
−
0.5% to +0.3% in lake Trichonis. A
sensitivity analysis was conducted by the authors, but its results are explicitly limited by
the small water area change range. All four lakes showed near zero sensitivity
[−5%, +5%]
to precipitation and temperature, when examined annually. Even when water area sensi-
tivity is examined for larger temporal periods, no specific pattern can be identified. For
instance, when examined for the 5-year average water area sensitivity to precipitation
and temperature, the results are incoherent. Consequently, this research argues that the
small range of change recorded during the period 1984–2022, poses a limit to presenting a
well-established sensitivity analysis. Additionally, there is a lack of tools to examine the
lakes relationships and interconnections with groundwater, which may also contribute
to the observed water content stability. Another aspect which should be addressed in
future research is the evaluation of the underlying uncertainties of the data used in the
present study. However, since all data sources have been extensively evaluated and used in
different case studies and scopes, the underlying uncertainty is judged to be minor.
The present research attempted to identify and evaluate potential quantitative changes
in the water content of four natural lakes of western Greece and links to climate variation.
Furthermore, the investigation assessed climate variables, such as precipitation and tem-
perature, perceived as highly connected to climate change. Finally, land cover and land use
changes were also assessed, since they have strong interconnections with water resources.
The overall results are concisely presented in Table 2and discussed below.
Table 2. Overall statistical analysis results; annual timeseries.
Lake/Basin Water Area Precipitation Temperature
pZMK pZMK pZMK
Trichonis ≈1 1.84 ≈1 2.80 ≈1 4.65
Lysimacheia ≈1−2.06 ≈1 2.37 ≈1 4.52
Ozeros ≈1−0.60 ≈1 3.13 ≈1 4.45
Amvrakia ≈1−0.73 ≈1 1.60 ≈1 4.50
Neither the water area of the four lakes, nor the climatic variables of their basins
assessed—precipitation and temperature—show statistically significant annual or seasonal
trends or breakpoints during the examined timeseries from 1984 to 2016–2022. The findings
of the present research are in agreement with a long-term drought trend analysis over
Greece, where no statistically significant trend was identified [
20
]. Moreover, land cover
has remained relatively stable between 1990 and 2018 for the four lake basins, with only
small increases in artificial and natural vegetation areas and decreases of cultivated ones.
Taking also into account the seasonal stability of the water area timeseries, it is argued that
the hydrologic regime is stable within the lake basins.
Similar research, conducted in different case studies, has associated lake water level
decrease with an intensification of agricultural land and water usage [
14
,
15
], and not with
climatic variables, when the latter remain relatively stable. Such are the present cases of
the four lake basins of western Greece, where no significant change of climatic variables
is observed. Additionally, the fact that land cover, and especially agricultural land, has
remained stable are converging to the long-term stability of the lake water areas, and thus
water contents. The nearest case study examined by similar research was lake Prespa,
a transboundary natural lake, located in the borders of Greece, Albania, and Northern
Macedonia [
38
]. Although lake Prespa is subject to a century-long decrease in water content,
it is suggested that increased irrigation of agricultural land has significantly contributed

Sensors 2023,23, 2056 15 of 19
to a more intense rate of water storage loss [
38
]. Moreover, another study, examining arid
lakes in China, has indicated human uses, and alterations of the basin’s hydrology, as major
drivers of lake water loss [69].
According to the 6th IPCC report, the majority of Mediterranean basins are projected to
have reduced discharges, while lake water levels are also expected to decline [
11
]. Moreover,
the IPCC report argues for significant (at 0.1 level) trends regarding annual precipitation
(decreasing) and annual average temperature (increasing) over western Greece [
10
]. The
abovementioned findings could not be verified by the present research, either regarding
annual precipitation, which was based on in-situ data, or the annual temperature of
the basins, which was calculated from the ERA5-Land reanalysis data. Therefore, an
interesting finding emerges, that different datasets and statistical approaches over the
same variables, may lead to divergent results. The present research’s attempt to capture
potential quantitative changes of lake water exploited the finest resolution of open access
data, 30 m Landsat imagery for lake water area and a dense network of rain gauge stations.
Taking into account, the divergent results of this case specific study and of global or regional
datasets, the need for extensive validation of the latter is stressed. Furthermore, the need for
systematic and efficient downscaling and upscaling approaches is evident, in order to close the
gap between local, regional, and global monitoring, and forecasting of hydroclimatic variables.
5. Conclusions
The main findings of this research are concentrated below:
•
Between 1984 and 2022, the maximum and minimum water area variations compared
to average water area for lakes Trichonis, Lysimacheia, Ozeros and Amvrakia, were
+1.38% to
−
3.86%, +39.61% to
−
32.37%, +4.81% to 4.43% and +19.34% to
−
19.71%
respectively.
•
Annual average trend analysis conducted on the Landsat derived water area timeseries,
resulted in p
≈
1 and Z
MK
equal to
−
0.33,
−
5.07, +0.36 and +1.87 for the four lakes
respectively, depicting the absence of statistically significant quantitative trends.
•
Seasonal trend analysis conducted on the Landsat derived water area timeseries did
not again show a statistically significant quantitative trend.
•
The climate variable of precipitation did not reveal a statistically significant trend
during years from 1984 to 2016–2019, in annual and seasonal scale.
•
The climate variable of temperature did not reveal a statistically significant trend
during years from 1984 to 2022, in annual, seasonal and monthly scale.
•
Land use change from 1990 to 2018 revealed a generally stable land environment, with
neglible to small increases of artificial and natural areas with a simultaneous decrease
of agricultural areas.
The abovementioned points align towards the observed stability of the water area,
which was used as an indicator of the lake water storage. Additionally, the stability
of the climatic variables of precipitation and temperature was depicted with the use of
statistics. Furthermore, the land use changes were assessed and found mostly stable in the
catchments of the four lakes examined. These findings strengthen recent similar research,
which examined catchments with intense land use alternations and relatively stable climatic
variables, resulting in significant lake water storage change. Thus, it is argued that land
uses and human interventions to the water cycle are a major driver of pressures to water
resources. Regarding the hydrologic implications of the present research, the results depict
a stable natural and human environment in the examined catchments. However, the
constant monitoring of the associated variables is necessary in order to ensure early and
solid identification of change and breakpoints.

Sensors 2023,23, 2056 16 of 19
The present research successfully demonstrates the combination of different Earth
Observation tools and practices, towards the monitoring of regional hydrologic regimens.
More specifically, this research makes ground to novel remote sensing and hydrological
approaches in Greece, a climatically diverse Mediterranean country. It demonstrates an
efficient practice of merging earth observation and in-situ data to create consistent and
long timeseries of hydrologic and climatic variables. The exclusive use of open access data
promotes the applicability of the proposed approach. The capabilities of open access data
and water resources monitoring are expected to be more vital for the years to come, as water
demand and climatic pressures are expected to intensify, even if their impacts are not yet
traceable in several regions. The use of statistical tools is considered important, in order to
trace significant alternations of water regimes and correctly track, each specific case’s causes.
Moreover, the examined Earth Observation data sources, when validated with ground
truth data, may be significant inputs for successful climate and hydrological forecasting.
Author Contributions:
Conceptualization, N.G. and E.B.; methodology, N.G.; software, N.G.; valida-
tion, N.G. and E.B.; data curation, N.G.; writing—original draft preparation, N.G.; writing—review
and editing, N.G. and E.B.; visualization, N.G.; supervision, E.B. All authors have read and agreed to
the published version of the manuscript.
Funding: This research received no external funding.
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement:
All freely available data are mentioned in section on Data and Methods.
Conflicts of Interest: The authors declare no conflict of interest.
References
1.
Williamson, C.E.; Saros, J.E.; Vincent, W.F.; Smol, J.P. Lakes and Reservoirs as Sentinels, Integrators, and Regulators of Climate
Change. Limnol. Oceanogr. 2009,54, 2273–2282. [CrossRef]
2.
Busker, T.; de Roo, A.; Gelati, E.; Schwatke, C.; Adamovic, M.; Bisselink, B.; Pekel, J.-F.; Cottam, A. A Global Lake and Reservoir
Volume Analysis Using a Surface Water Dataset and Satellite Altimetry. Hydrol. Earth Syst. Sci. Discuss.
2018
,23, 1–32. [CrossRef]
3.
Frappart, F.; Zeiger, P.; Betbeder, J.; Gond, V.; Bellot, R.; Baghdadi, N.; Blarel, F.; Darrozes, J.; Bourrel, L.; Seyler, F. Automatic
Detec-Tion of Inland Water Bodies along Altimetry Tracks for Estimating Surface Water Storage Variations in the Congo Basin.
Remote Sens. 2021,10, 3804. [CrossRef]
4.
Zhong, S.; Geng, Y.; Qian, Y.; Chen, W.; Pan, H. Analyzing Ecosystem Services of Freshwater Lakes and Their Driving Forces: The
Case of Erhai Lake, China. Environ. Sci. Pollut. Res. 2019,26, 10219–10229. [CrossRef] [PubMed]
5.
Kostianoy, A.G.; Lebedev, S.A.; Kostianaia, E.A.; Prokofiev, Y.A. Interannual Variability of Water Level in Two Largest Lakes of
Europe. Remote Sens. 2022,14, 659. [CrossRef]
6.
Gourgouletis, N.; Bariamis, G.; Baltas, E. Estimation of Characteristics of Surface Water Bodies Based on Sentinel-2 Images: The
Case Study of Yliki Reservoir. In Proceedings of the Eigth International Conference on Environmental Management, Engineering,
Planning & Economics, Thessaloniki, Greece, 20–24 July 2021; pp. 551–558.
7.
Schmitz, O.J.; Raymond, P.A.; Estes, J.A.; Kurz, W.A.; Holtgrieve, G.W.; Ritchie, M.E.; Schindler, D.E.; Spivak, A.C.; Wilson, R.W.;
Bradford, M.A.; et al. Animating the Carbon Cycle. Ecosystems 2014,17, 344–359. [CrossRef]
8.
Jimenez Cisneros, B.E.; Oki, T.; Arnell, N.W.; Benito, G.; Cogley, J.G.; Döll, P.; Jiang, T.; Mwakalila, S.S. Freshwater Resources. In
Climate Change 2014: Impacts, Adaptation, and Vulnerability. Part A: Global and Sectoral Aspects. Contribution of Working Group II
to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change; Field, C.B., Barros, V.R., Dokken, D.J., Mach, K.J.,
Mastrandrea, M.D., Bilir, T.E., Chatterjee, M., Ebi, K.L., Estrada, Y.O., Genova, R.C., et al., Eds.; Cambridge University Press:
Cambridge, UK; New York, NY, USA, 2014; pp. 1199–1265.
9.
Kovats, R.S.; Valentini, R.; Bouwer, L.M.; Georgopoulou, E.; Jacob, D.; Martin, E.; Rounsevell, M.; Soussana, J.-F. Europe. In
Climate Change 2014: Impacts, Adaptation, and Vulnerability. Part B: Regional Aspects. Contribution of Working Group II to the Fifth
Assessment Report of the Intergovernmental Panel on Climate Change; Field, C.B., Barros, V.R., Dokken, D.J., Mach, K.J., Mastrandrea,
M.D., Bilir, T.E., Chatterjee, M., Ebi, K.L., Estrada, Y.O., Genova, R.C., et al., Eds.; Cambridge University Press: Cambridge, UK;
New York, NY, USA, 2014; pp. 1267–1326.
10. Gutiérrez, J.M.; Mernild, S.H.; Jones, R.G.; Narisma, G.T.; Alves, L.M.; Amjad, M.; Gorodetskaya, I.V.; Grose, M.; Klutse, N.A.B.;
Krakovska, S.; et al. Atlas. In Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment
Report of the Intergovernmental Panel on Climate Change; Masson-Delmotte, V., Zhai, P., Pirani, A., Connors, S., Péan, C., Berger, S.,
Caud, N., Chen, Y., Goldfarb, L., Gomis, M., Eds.; Cambridge University Press: Cambridge, UK, 2021; pp. 1927–2058.

Sensors 2023,23, 2056 17 of 19
11.
Ali, E.; Cramer, W.; Carnicer, J.; Georgopoulou, E.; Hilmi, N.J.M.; le Cozannet, G.; Lionello, P.; Pörtner, H.-O.; Roberts, D.C.; Tignor,
M.; et al. Cross-Chapter Paper 4: Mediterranean Region. In Climate Change 2022: Impacts, Adaptation and Vulnerability. Contribution
of Working Group II to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change; Cambridge University Press:
Cambridge, UK, 2022; pp. 2233–2272.
12.
Weatherhead, E.K.; Howden, N.J.K. The Relationship between Land Use and Surface Water Resources in the UK. Land Use Policy
2009,26, S243–S250. [CrossRef]
13.
Thornton, P.K.; Herrero, M. The Inter-Linkages between Rapid Growth in Livestock Production, Climate Change, and the Impacts on Water
Resources, Land Use, and Deforestation Background Paper to the 2010 World Development Report; 2010; The World Bank: Washington,
DC, USA, 2010.
14.
Khazaei, B.; Khatami, S.; Alemohammad, S.H.; Rashidi, L.; Wu, C.; Madani, K.; Kalantari, Z.; Destouni, G.; Aghakouchak, A.
Climatic or Regionally Induced by Humans? Tracing Hydro-Climatic and Land-Use Changes to Better Understand the Lake
Urmia Tragedy. J. Hydrol. 2019,569, 203–217. [CrossRef]
15.
Satgé, F.; Espinoza, R.; Zolá, R.P.; Roig, H.; Timouk, F.; Molina, J.; Garnier, J.; Calmant, S.; Seyler, F.; Bonnet, M.P. Role of Climate
Variability and Human Activity on PoopóLake Droughts between 1990 and 2015 Assessed Using Remote Sensing Data. Remote
Sens. 2017,9, 218. [CrossRef]
16.
Bariamis, G.; Paschos, G.; Baltas, E. Land Accounts in the River Basin Districts of Greece. Environ. Process.
2018
,5, 213–237.
[CrossRef]
17. OECD. Water and Climate Change Adaptation; OECD Studies on Water; OECD: Paris, France, 2013; ISBN 9789264200432.
18.
Karavitis, C.A.; Chortaria, C.; Alexandris, S.; Vasilakou, C.G.; Tsesmelis, D.E. Development of the Standardised Precipitation
Index for Greece. Urban Water J. 2012,9, 401–417. [CrossRef]
19.
Livada, I.; Assimakopoulos, V.D. Spatial and Temporal Analysis of Drought in Greece Using the Standardized Precipitation Index
(SPI). Appl. Clim. 2007,89, 143–153. [CrossRef]
20.
Kalamaras, N.; Michalopoulou, H.; Byun, H.R. Detection of Drought Events in Greece Using Daily Precipitation. Hydrol. Res.
2010,41, 126–133. [CrossRef]
21.
Terzidis, G. Water Scarcity; Causes, Forecasting Potential and Mitigation Measures in the Urban and Agricultural Development of
Greece. In Water Scarcity and Floods; Geotechnical Chamber of Greece: Thessaloniki, Greece, 1992; pp. 25–40.
22. Biswas, A.K. Integrated Water Resources Management: A Reassessment. Water Int. 2004,29, 248–256. [CrossRef]
23.
Schwatke, C.; Dettmering, D.; Bosch, W.; Seitz, F. DAHITI—An Innovative Approach for Estimating Water Level Time Series over
Inland Waters Using Multi-Mission Satellite Altimetry. Hydrol. Earth Syst. Sci. 2015,19, 4345–4364. [CrossRef]
24.
Chawla, I.; Karthikeyan, L.; Mishra, A.K. A Review of Remote Sensing Applications for Water Security: Quantity, Quality, and
Extremes. J. Hydrol. 2020,585, 124826. [CrossRef]
25.
Baup, F.; Frappart, F.; Maubant, J. Combining High-Resolution Satellite Images and Altimetry to Estimate the Volume of Small
Lakes. Hydrol. Earth Syst. Sci. 2014,18, 2007–2020. [CrossRef]
26.
Yang, X.; Zhao, S.; Qin, X.; Zhao, N.; Liang, L. Mapping of Urban Surface Water Bodies from Sentinel-2 MSI Imagery at 10 m
Resolution via NDWI-Based Image Sharpening. Remote Sens. 2017,9, 596. [CrossRef]
27.
Cavallo, C.; Papa, M.N.; Gargiulo, M.; Palau-Salvador, G.; Vezza, P.; Ruello, G. Continuous Monitoring of the Flooding Dynamics
in the Albufera Wetland (Spain) by Landsat-8 and Sentinel-2 Datasets. Remote Sens. 2021,13, 3525. [CrossRef]
28.
McFeeters, S.K. The Use of the Normalized Difference Water Index (NDWI) in the Delineation of Open Water Features. Int. J.
Remote Sens. 1996,17, 1425–1432. [CrossRef]
29.
Sekertekin, A. A Survey on Global Thresholding Methods for Mapping Open Water Body Using Sentinel-2 Satellite Imagery and
Normalized Difference Water Index. Arch. Comput. Methods Eng. 2020,28, 1335–1347. [CrossRef]
30.
Gourgouletis, N.; Bariamis, G.; Anagnostou, M.N.; Baltas, E. Estimating Reservoir Storage Variations by Combining Sentinel-2
and 3 Measurements in the Yliki Reservoir, Greece. Remote Sens. 2022,14, 1860. [CrossRef]
31.
Du, Y.; Zhang, Y.; Ling, F.; Wang, Q.; Li, W.; Li, X. Water Bodies’ Mapping from Sentinel-2 Imagery with Modified Normalized
Difference Water Index at 10-m Spatial Resolution Produced by Sharpening the Swir Band. Remote Sens.
2016
,8, 354. [CrossRef]
32.
Acharya, T.D.; Subedi, A.; Lee, D.H. Evaluation of Water Indices for Surface Water Extraction in a Landsat 8 Scene of Nepal.
Sensors 2018,18, 2580. [CrossRef]
33.
Mehmood, H.; Conway, C.; Perera, D. Mapping of Flood Areas Using Landsat with Google Earth Engine Cloud Platform.
Atmosphere 2021,12, 866. [CrossRef]
34.
Ichsan Ali, M.; Darma Dirawan, G.; Hafid Hasim, A.; Rais Abidin, M. Detection of Changes in Surface Water Bodies Urban Area
with NDWI and MNDWI Methods. Int. J. Adv. Sci. Eng. Inf. Technol. 2019,9, 946.
35.
Deoli, V.; Kumar, D.; Kuriqi, A. Detection of Water Spread Area Changes in Eutrophic Lake Using Landsat Data. Sensors
2022
,22,
6827. [CrossRef] [PubMed]
36.
Schmitt, M. Potential of Large-Scale Inland Water Body Mapping from Sentinel-1/2 Data on the Example of Bavaria’s Lakes and
Rivers. PFG-J. Photogramm. Remote Sens. Geoinf. Sci. 2020,88, 271–289. [CrossRef]
37.
Xu, H. Modification of Normalised Difference Water Index (NDWI) to Enhance Open Water Features in Remotely Sensed Imagery.
Int. J. Remote Sens. 2006,27, 3025–3033. [CrossRef]
38.
Soria, J.; Apostolova, N. Decrease in the Water Level of Lake Prespa (North Macedonia) Studied by Remote Sensing Methodology:
Relation with Hydrology and Agriculture. Hydrology 2022,9, 99. [CrossRef]

Sensors 2023,23, 2056 18 of 19
39.
Markogianni, V.; Kalivas, D.; Petropoulos, G.P.; Dimitriou, E. Modelling of Greek Lakes Water Quality Using Earth Observation
in the Framework of the Water Framework Directive (WFD). Remote Sens. 2022,14, 739. [CrossRef]
40.
Peppa, M.; Vasilakos, C.; Kavroudakis, D. Eutrophication Monitoring for Lake Pamvotis, Greece, Using Sentinel-2 Data. ISPRS
Int. J. Geoinf. 2020,9, 143. [CrossRef]
41.
Alexandridis, T.K.; Takavakoglou, V.; Crisman, T.L.; Zalidis, G.C. Remote Sensing and GIS Techniques for Selecting a Sustainable
Scenario for Lake Koronia, Greece. Environ. Manag. 2007,39, 278–290. [CrossRef]
42.
Papastergiadou, E.S.; Retalis, A.; Apostolakis, A.; Georgiadis, T. Environmental Monitoring of Spatio-Temporal Changes Using
Remote Sensing and GIS in a Mediterranean Wetland of Northern Greece. Water Resour. Manag. 2008,22, 579–594. [CrossRef]
43.
Psomiadis, E.; Soulis, K.X.; Zoka, M.; Dercas, N. Synergistic Approach of Remote Sensing and Gis Techniques for Flash-Flood
Monitoring and Damage Assessment in Thessaly Plain Area, Greece. Water 2019,11, 448. [CrossRef]
44.
Capolongo, D.; Refice, A.; Bocchiola, D.; D’Addabbo, A.; Vouvalidis, K.; Soncini, A.; Zingaro, M.; Bovenga, F.; Stamatopoulos,
L. Coupling Multitemporal Remote Sensing with Geomorphology and Hydrological Modeling for Post Flood Recovery in the
Strymonas Dammed River Basin (Greece). Sci. Total Environ. 2019,651, 1958–1968. [CrossRef] [PubMed]
45.
Kontopoulou, E.; Kolokoussis, P.; Karantzalos, K. Water Quality Estimation in Greek Lakes from Landsat 8 Multispectral Satellite
Data. Eur. Water 2017,58, 191–196.
46.
Elhag, M.; Yilmaz, N. Insights of Remote Sensing Data to Surmount Rainfall/Runoff Data Limitations of the Downstream
Catchment of Pineios River, Greece. Environ. Earth Sci. 2021,80, 35. [CrossRef]
47.
Valero-Jorge, A.; González-De Zayas, R.; Alcántara-Martín, A.; Álvarez-Taboada, F.; Matos-Pupo, F.; Brown-Manrique, O. Water
Area and Volume Calculation of Two Reservoirs in Central Cuba Using Remote Sensing Methods. A New Perspective. Rev.
Teledetec. 2022,2022, 71–87. [CrossRef]
48.
Walker, D.; Shutler, J.D.; Morrison, E.H.J.; Harper, D.M.; Hoedjes, J.C.B.; Laing, C.G. Quantifying Water Storage within the North
of Lake Naivasha Using Sonar Remote Sensing and Landsat Satellite Data. Ecohydrol. Hydrobiol. 2021,22, 12–20. [CrossRef]
49.
Lin, Y.; Li, X.; Zhang, T.; Chao, N.; Yu, J.; Cai, J.; Sneeuw, N. Water Volume Variations Estimation and Analysis Using Multisource
Satellite Data: A Case Study of Lake Victoria. Remote Sens. 2020,12, 3052. [CrossRef]
50.
Crétaux, J.F.; Abarca-del-Río, R.; Bergé-Nguyen, M.; Arsen, A.; Drolon, V.; Clos, G.; Maisongrande, P. Lake Volume Monitoring
from Space. Surv. Geophys. 2016,37, 269–305. [CrossRef]
51. 1st Revision RBDMP of Western Greece. Greek Gov. Gaz. 2017,4681 B, 65633–65920. (In Greek)
52.
Perivolioti, T.M.; Frouzova, J.; Tušer, M.; Bobori, D. Assessing the Fish Stock Status in Lake Trichonis: A Hydroacoustic Approach.
Water 2020,12, 1823. [CrossRef]
53.
Petriki, O.; Moutopoulos, D.K.; Tsagarakis, K.; Tsionki, I.; Papantoniou, G.; Mantzouni, I.; Barbieri, R.; Stoumboudi, M.T.
Assessing the Fisheries and Ecosystem Structure of the Largest Greek Lake (Lake Trichonis). Water 2021,13, 3329. [CrossRef]
54.
Special Secreteriat for Water (In Greek). 1st Revision RBDMP of Western Greece;Deliverable 05-M2; Analysis of Human Induced
Pressures and Impacts on Surface and Underground Water Bodies; Hellenic Ministry of Environment and Energy: Athens,
Greece, 2017.
55.
Special Secretariat for Water (In Greek). 1st Revision RBDMP of Western Greece;Deliverable 06-M2; Characterization, Typology and
Evaluation/Ranking of the Status of Surface Water Bodies; Hellenic Ministry of Environment and Energy: Athens, Greece, 2017.
56. Scaramuzza, P.; Micijevic, E.; Chander, G. SLC Gap-Filled Products;Phase One Methodology; USGS: Reston, VA, USA, 2004.
57.
Khaki, M.; Awange, J. The 2019–2020 Rise in Lake Victoria Monitored from Space: Exploiting the State-of-the-Art Grace-Fo and
the Newly Released Era-5 Reanalysis Products. Sensors 2021,21, 4304. [CrossRef]
58.
Cerlini, P.B.; Saraceni, M.; Silvestri, L.; Meniconi, S.; Brunone, B. Monitoring the Water Mass Balance Variability of Small Shallow
Lakes by an ERA5-Land Reanalysis and Water Level Measurement-Based Model. An Application to the Trasimeno Lake, Italy.
Atmosphere 2022,13, 949. [CrossRef]
59.
Büttner, G.; Kosztra, B.; Maucha, G.; Pataki, R.; Kleeschulte, S.; Hazeu, G.; Vittek, M.; Littkopf, A. Copernicus Land Monitoring
Service; European Environment Agency: Copenhagen, Denmark, 2021.
60. Prewitt, J.M.S.; Mendelsohn, M.L. The analysis of cell images. Ann. N. Y. Acad. Sci. 2006,128, 1035–1053. [CrossRef]
61. Mann, H.B. Nonparametric Tests Against Trend. Econometrica 1945,13, 245. [CrossRef]
62. Kendall, M.G. Rank Correlation Methods, 4th ed.; Griffin: London, UK, 1975; ISBN 9780852641996.
63. Pettitt, A.N. A Non-Parametric Approach to the Change-Point Problem. Appl. Stat. 1979,28, 126. [CrossRef]
64.
Gagnon, A.S.; Gough, W.A. Hydro-Climatic Trends in the Hudson Bay Region, Canada. Can. Water Resour. J.
2002
,27, 245–262.
[CrossRef]
65.
Tadese, M.T.; Kumar, L.; Koech, R.; Zemadim, B. Hydro-Climatic Variability: A Characterisation and Trend Study of the Awash
River Basin, Ethiopia. Hydrology 2019,6, 35. [CrossRef]
66.
Okafor, G.C.; Jimoh, O.D.; Larbi, K.I. Detecting Changes in Hydro-Climatic Variables during the Last Four Decades (1975-2014)
on Downstream Kaduna River Catchment, Nigeria. Atmos. Clim. Sci. 2017,7, 161–175. [CrossRef]
67.
Tekleab, S.; Mohamed, Y.; Uhlenbrook, S. Hydro-Climatic Trends in the Abay/Upper Blue Nile Basin, Ethiopia. Phys. Chem. Earth
2013,61–62, 32–42. [CrossRef]

Sensors 2023,23, 2056 19 of 19
68.
Hirsch, R.M.; Slack, J.R. A Nonparametric Trend Test for Seasonal Data With Serial Dependence. Water Resour. Res.
1984
,20,
727–732. [CrossRef]
69.
Lv, A.; Zhang, C. Analysis of the Characteristics and Driving Forces of Changes in Lake Water Volume in Inland Arid Basins in
China. Water 2022,14, 3141. [CrossRef]
Disclaimer/Publisher’s Note:
The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.