Acceso abierto

Estimation of Morphometric Parameters in Lakes Based on Satellite Imagery Data: Implications of Relationships Between Lakes in the Arid Region of Western Mongolia, Central Asia

, , , , , , , ,  y   
21 ene 2025

Cite
Descargar portada

Introduction

Object-based satellite image classification to identify pre- and post-event changes is the primary method for quickly retrieving evaluation information (Valeyev et al. 2019, Liu et al. 2020). While remote sensing technology has advanced significantly with the development of modern technology, its accessibility remains relatively limited in less developed countries (Sheffield et al. 2018). However, the demand and requirements for research areas using remote sensing technology and research using satellite image data are essential for those countries (Lebedev et al. 2020). Using satellite image data with various spatial resolutions, modelling of lake area changes, monitoring of shoreline dynamics, monitoring of water depth changes, analysis of water index and calculations of water evaporation are performed (Busker et al. 2019, Valeyev et al. 2019, Liu et al. 2020, Luo et al. 2020, Emami, Zarei 2021, Şerban et al. 2022, Xu et al. 2022). Integrating spatial and temporal data from space imagery provides key evidence for detailing lake area changes and interrelationships (Valeyev et al. 2019, Liu et al. 2020, Lehner et al. 2022, Xu et al. 2022). Based on satellite imagery, the study analyses the various changes occurring in the lake and determines the dominant factors (Melesse et al. 2007, Dörnhöfer, Oppelt 2016, Shen et al. 2022). Variations in lake volume, area, their interactions and human activities could exert future influence on the area’s climate and environment.

Morphometric changes in the lakes have the potential to trigger a series of ecological and environmental effects, both positive and negative, making it imperative to investigate these transformations. The area of lakes has significantly changed in modern times due to human activities (Seyoum et al. 2015, Qi et al. 2020). Also, due to global climate change, surface water resources and distribution are changing over time (Wang, Qin 2017, Dorjsuren et al. 2018); based on human and natural activities, the lake area is shrinking and increasing, and even new artificial lakes are being created (Zhang et al. 2019). The impact of the processes on both natural ecosystems and human activities associated with the lake has been manifested. The importance of providing reservoirs and ponds in arid climatic regions has been well-proved in numerous studies (Busker et al. 2019, Chen et al. 2020, Mady et al. 2020, Stringer et al. 2021, Lehner et al. 2022, Pi et al. 2022, Shen et al. 2022, Rousta et al. 2023). A synthesis of this research emphasises that developing reservoirs and ponds is crucial for the region’s climate, environment, biodiversity, agriculture, and economy.

In recent years, lakes in arid regions have shown a tendency to shrink due to climate change and global environmental shifts (Chen et al. 2020, Stringer et al. 2021). Air temperature plays a significant role in reducing lake volume (Yu et al. 2021). The increase in average annual air temperature has led to greater surface evaporation, resulting in smaller size of the lakes across Central Asia (Sumiya et al. 2020, Enkhbold et al. 2024). Additionally, the trend of decreasing precipitation in recent years has a direct impact on lake volumes in arid regions (Rousta et al. 2023). However, human-driven actions to create reservoirs and ponds help counter this trend by increasing the number of lakes (Pi et al. 2022).

As the scientific and practical challenges of utilising lakes as a primary natural resource are highly crucial, conducting in-depth research on human impacts on lakes is of paramount importance. Mongolia, where nearly 90% of its surface water is stored in lakes (Tserensodnom 2000, Sato et al. 2007), emphasises the significance of investigating lakes to address water resource management and environmental and socioeconomic development concerns (Oyunbaatar et al. 2017). In the future, the results of this research may serve as a foundational model for studying the chemical composition, water characteristics, biological resources, ecosystems and biodiversity of lakes in arid regions. Despite its significance, few studies have previously evaluated the relationship between reservoirs and naturally formed lakes in Mongolia (Sukhbaatar et al. 2020, Baterdene et al. 2022). This research is necessary to raise new questions, introduce them to the scientific community and advance the study of Mongolian lakes.

Mongolia’s geographical landscape, located in Central Asia, is defined by a transition zone extending from the southern Siberian Mountains to the Central Asian steppes (Tserensodnom 1971, Yembuu 2021). This region experiences a gradual shift from a cold, humid northern climate to an arid southern one (Lehmkuhl et al. 2016, Yembuu 2021). Most of Mongolia’s lakes are situated in dry and arid areas, primarily in the southern and western parts of the country (Tserensodnom 2000, Enkhbold et al. 2021, 2022, Dorjsuren et al. 2024). This arid region is a focal point for changes in surface water, heavily influenced by both natural and human activities.

In Mongolia, the utilisation of hydropower plants (HPPs) requires careful planning to balance hydropower generation, water supply, pasture irrigation and agricultural needs. Gegeen Lake was created in 2007 by constructing a hydropower dam 50 m deep, 75 m wide and 190 m long in the Ulaan Boom Canyon of the Zavkhan River (Oyunbaatar et al. 2011, Mendsaihan et al. 2016). In 2018, it was registered as part of an international satellite imagery database under the name Gegeen Lake (Saint Lake in English). The Taishir HPP generates 12 MW of electricity annually and serves 8–10 towns in the Zavkhan and Govi-Altai provinces (Mendsaihan et al. 2016). Unlike fossil fuel power, HPPs do not emit harmful waste or CO2, playing a key role in reducing greenhouse gas emissions (Kumar et al. 2019). However, the construction of the HPP and the creation of Gegeen Lake have caused a notable reduction in the wetland area around Airag Lake, a Ramsar Convention site (Purevdorj et al. 2019, 2023). This case represents how human activities can negatively impact the environment. The area surrounding Airag Lake hosts 183 species of waterfowl, whose habitat is now at risk due to the shrinking wetland area (Purevdorj et al. 2019, 2023).

This research aims to calculate morphometric changes in the lakes based on satellite imagery data and to assess its effect on the hydrological relationships among lakes in Mongolia’s arid regions.

Study area

The catchment area in the Zavkhan River–Khyargas Lake (Fig. 1) belongs to the closed basin of Central Asia (Ochir et al. 2013). The Zavkhan River originates in the Khangay Mountains and flows 808 km into Airag Lake (Fig. 1A, B). The total area of the catchment is 98822–99040 km2 (Ochir et al. 2013, Dorjsuren et al. 2023), the average annual flow is 0.4 km3, the average runoff is 1.3 m · s−1 and the flow modulus is 1.13 m · s−1 (Oyunbaatar et al. 2011, Davaa 2015, Dorjsuren et al. 2023). The three lakes included in this study are situated in the Zavkhan River–Khyargas Lake basin (Fig. 1C, D and Table 1).

Fig. 1.

A – The geographical location of the study area. B – Topographic map of Mongolia and the Zavkhan River–Khyargas Lake basin in western Mongolia. C – Satellite images of the Airag and Khyargas lakes. D – Gegeen Lake.

Morphometric parameters of the study area (Modified after Tserensodnom 2000; Davaa 2018; Enkhbold et al. 2022; Dorjsuren et al. 2023, 2024).

Description Unit Khyargas Lake Airag Lake Gegeen Lake
Location DMS 49°11'05"N, 93°16'16"E 48°53'12"N, 93°26'09"E 46°41'02"N, 96°46'17"E
Elevation (m a.s.l.) m 1028 1030 1727
Average area ha 138,306 15,179 3794
Shape Direction W–E W–E NW–SE
Average length km 75 18 17.2
Average width km 19 13 3.1
Depth (max) m 78.7 10.2 31.5
Origin Tectonic Tectonic Human
Chemistry Salt Fresh Fresh
Water source Surface Surface Surface

The study area is surrounded by Tagna Mountains in the north, Khangai Mountains in the east, Gobi Altai Mountains in the south, and Mongolian Altai Mountains in the west (Fig. 1). Elevations in this area range from 760 to 1800 m a.s.l. Hydrologically, a network of rivers originates from the Altai and Khangai mountain ranges in Mongolia. Extreme weather conditions in the region include low precipitation, resulting in arid and desert landscapes (Lehmkuhl et al. 2016, Yembuu 2021, Dorjsuren et al. 2023).

Data and methods
Satellite data

The relationships between various parameters were determined by analysing spatial image data. We obtained Landsat 5 TM, Landsat 7 +ETM and Landsat 8 OLI/TIRS satellite imagery data with a 30-m resolution from USGS (2023). These data sources were used to calculate the surface area of Airag and Khyargas Lake water over 30 years from 1991 to 2021, as well as the surface area of Gegeen Lake water over 15 years from 2007 to 2021. The volume of the lakes is updated using three-dimensional bathymetric mapping (Tserensodnom 2000, Davaa 2018). In 2021, field measurements were conducted to assess changes in the lakes’ shorelines.

Climate data

The climate data of the Khyargas Lake basin in western Mongolia from 1991 to 2021 were provided by the Database Center at the National Agency of Meteorology and Environment of Mongolia. The monthly air temperature and precipitation data, measured at the Taishir (Gegeen Lake) and Zavkhan (Airag Lake and Khyargas Lake) weather stations, were used to estimate the impact of climate factors on lake volume changes.

Morphological analysis

Morphological analysis (MA) is a method of detecting the external signs and changes in various surface elements and determining their shape (Finkl et al. 2005, Harmar et al. 2005). The method is unique in its ability to identify changes in surface shapes and the external processes that influence them. This study will show how the morphology of the lake’s water surface changes in response to alternations in the lake’s surface using satellite imagery results (Soille, Pesaresi 2002, Shang 2013). MA offers the advantage of quantifying average morphological change by measuring surface alternations. This study is used to determine the average size of the area change of the lake.

If the area of the lake changes according to the satellite map calculations, the average graphical change in the lake area, as determined by MA, is calculated using Eq. (1): Sa=SF+SLtnSa = {{SF + SL} \over {tn}} where:

Sa is the average change in the area of the lake,

SF is the amount of the first estimated area of the lake,

SL is the amount of the area of the lake calculated in the last period,

tn is the sum of the time series.

This is essentially the linear regression coefficient of the time series.

Delineation of water surface from multispectral images

Several extracting techniques can be applied to identify and highlight water surface in multispectral images. These methods include approaches that utilise (1) reflected solar radiation, (2) emitted thermal radiation and (3) active microwave emissions. Among these, techniques relying on reflected solar radiation have the longest history of use. In this approach, researchers have either used a single spectral band or ratio of two bands to distinguish and enhance open water features in the imagery. Normalised difference water index (NDWI) is one of the most commonly used water indices to calculate changes in the lake shape and water surface area. It was first introduced by McFeeters (1996), utilising the green (GREEN) and near-infrared (NIR) spectral bands of Landsat TM. The wavelengths in challenges were selected for the following reasons: (1) to achieve the greatest possible reflectance of water features using GREEN wavelengths; (2) to reduce the typically low reflectance of water features in the NIR wavelengths and (3) to utilise the high NIR reflectance characteristic of vegetation and soil. The method is used to distinguish not only water features but also man-made structures that produce light reflections of the same colour as water (Gao 1996, McFeeters 2013, Bijeesh, Narasimhamurthy 2020, Luo et al. 2021). The NDWI on multispectral images is determined using the following Eq. (2): NDWI=GREENNIRGREEN+NIRNDWI = {{GREEN - NIR} \over {GREEN + NIR}} where:

NDWI is the normalised difference water index,

GREEN is the green wavelength (0.52–0.60 μm) in the visible spectrum and

NIR (0.76–0.90 μm) is NIR wavelengths in the infrared spectrum.

The central wavelengths of the GREEN bands on Landsat 5, 7 and 8 are similar at 0.56 μm, while those of the NIR bands are quite different at 0.83, 0.84 and 0.86 μm. Spatial resolution of the GREEN and NIR bands is 30 m. NDWI is dimensionless and varies between –1 and +1, depending on the leaf water content, vegetation type and surface cover characteristics. The NDWI values > 0.5 usually correspond to water bodies. Vegetation usually corresponds to much smaller values, and built-up areas correspond to values between 0 and 0.2. NDWI estimation of satellite images was performed by ENVI 5.0 (Exelis Visual Information Solutions, Boulder, Colorado), and comparison maps were applied on qGIS 3.18.

Statistical analysis

We examined the volume fluctuations in the three lakes and computed them using a one-factor regression equation. The linear regression equation for one factor is defined as shown in Eq. (3): y=p1x+p0y\;{\rm{ = }}\;{p_1}x{\rm{ + }}{p_0}

The fluctuations in lake volume over time are determined by climate trends (Sumiya et al. 2020, Dorjsuren et al. 2024).

The regression parameter a and the regression coefficient b were determined using the least square method and computed according to Eq. (4): b=i=1n(xix¯)(yiy¯)i=1n(xix¯)b = {{\sum\nolimits_{i = 1}^n {\left( {{x_i} - \bar x} \right)} \left( {{y_i} - \bar y} \right)} \over {\sum\nolimits_{i = 1}^n {\left( {{x_i} - \bar x} \right)} }}

Here, x=1Ni=1nxi x = {1 \over N}\sum\nolimits_{i = 1}^n {{x_i}} , y=1Ni=1nyi y = {1 \over N}\sum\nolimits_{i = 1}^n {{y_i}} b > 0 indicates an increase in climate factors and b< 0 indicates a decreasing trend.

The influence of climate on the changes in the volume of the lakes under study was determined, and the correlation coefficient was calculated using Eq. (5): rxy=i=0n[ (xix¯)(yiy¯) ]i=1n(xix¯)2+i=1n(yiy¯)2{r_{xy}} = {{\sum\nolimits_{i = 0}^n {\left[ {\left( {{x_i} - \bar x} \right)\left( {{y_i} - \bar y} \right)} \right]} } \over {\sqrt {\sum\nolimits_{i = 1}^n {{{\left( {{x_i} - \bar x} \right)}^2}} + \sum\nolimits_{i = 1}^n {{{\left( {{y_i} - \bar y} \right)}^2}} } }} where:

rx, y is the correlation coefficient,

xi is the independent variable,

is the mean of independent variables,

yi is the dependent variable,

ȳi is the mean of dependent variables,

rxy > 0 indicates a positive correlation and rxy < 0 indicates a negative relationship.

The one-factor linear regression equation for the factor was calculated using the ‘polyfit’ command in the MATLAB software. Correlation and regression analysis confirmed the relationship between changes in area and volume in the Gegeen, Airag and Khyargas lakes.

Volume analysis

The non-parametric analysis was used to find significant differences between water areas and volumes validated by on-site field measurements (Lu et al. 2013, Chipman 2019, Ahmed et al. 2021). The water volume in the lake is calculated by combining data on water-level variations with accurate bathymetry lakes (Duan, Bastiaanssen 2013, Lin et al. 2020). A water volume change was calculated (Yue, Liu 2019, Zhang et al. 2021, Qi et al. 2022) by the following basic Eq. (6): ΔV=13(h2h1)(A1+A2+A1×A2)\Delta V = {1 \over 3}\left( {{h_2} - {h_1}} \right)\left( {{A_1} + {A_2} + \sqrt {{A_1}} \times {A_2}} \right) where:

h1, A1 and h2, A2 are lake-level areas at the beginning and end of 2 years.

Saberioon et al. (2020) noted that utilising the R software represents the most convenient approach for integrating numerical data into mapping. In this study, the R software was employed to map the volume of lakes.

Figure 2 shows an outline of the study’s methodology.

Fig. 2.

Methodology flow chart.

Results and discussion
Lakes’ water surface area MA

The water surface area of the Khyargas Lake basin, influenced directly by the Zavkhan River’s flow, has experienced significant fluctuations in recent years. These variations are depicted in Table 2.

Estimation of water surface area in the Khyargas Lake basin.

Year Khyargas Lake area Airag Lake area Gegeen Lake area Satellite imagery considered in the analysis
[ha]
1991 137,482 13,190 0.00 LT05_L1TP_140026_19910811_20170125_01_T1
1992 137,726 13,419 0.00 LT05_L1TP_140026_19920813_20170122_01_T1
1993 138,084 16,501 0.00 LT05_L1TP_140026_19930816_20170117_01_T1
1994 139,201 17,811 0.00 LT05_L1TP_140026_19940718_20170114_01_T1
1995 139,913 19,295 0.00 LT05_L1TP_140026_19950822_20170107_01_T1
1996 140,011 19,319 0.00 LT05_L1TP_140026_19960621_20170104_01_T1
1997 139,967 19,039 0.00 LT05_L1TP_140026_19970827_20161230_01_T1
1998 140,105 19,064 0.00 LT05_L1TP_140026_19980611_20161224_01_T1
1999 140,317 18,266 0.00 LT05_L1TP_140026_19990716_20161217_01_T1
2000 139,728 18,023 0.00 LE07_L1TP_140026_20000827_20170210_01_T1
2001 140,007 17,859 0.00 LT05_L1TP_140026_20010806_20161210_01_T1
2002 139,699 17,562 0.00 LT05_L1TP_140026_20020809_20161207_01_T1
2003 139,697 17,392 0.00 LT05_L1TP_140026_20030727_20161205_01_T1
2004 139,629 17,272 0.00 LT05_L1TP_140026_20040713_20161130_01_T1
2005 139,349 16,901 0.00 LT05_L1TP_140026_20050817_20161124_01_T1
2006 139,347 16,545 0.00 LT05_L1TP_140026_20060719_20161120_01_T1
*LT05_L1TP_138027_20060619_20161121_01_T1
2007 139,124 15,923 89 LT05_L1TP_140026_20070823_20161112_01_T1
*LT05_L1TP_138027_20070825_20161111_01_T1
2008 138,609 15,125 1521 LT05_L1TP_140026_20080809_20161030_01_T1
*LT05_L1TP_138027_20080811_20161030_01_T1
2009 138,144 14,331 1801 LT05_L1TP_140026_20090711_20161024_01_T1
*LT05_L1TP_138027_20090713_20161023_01_T1
2010 138,167 13,808 3235 LT05_L1TP_140026_20100730_20161014_01_T1
*LT05_L1TP_138027_20100716_20161014_01_T1
2011 137,571 13,314 4749 LT05_L1TP_140026_20110818_20161007_01_T1
*LT05_L1TP_137028_20110728_20161008_01_T1
2012 137,453 13,096 4540 LE07_L1TP_140026_20120711_20161130_01_T1
*LE07_L1TP_138027_20120814_20161129_01_T1
2013 136,661 12,429 4431 LC08_L1TP_140026_20130823_20170502_01_T1
*LC08_L1TP_137028_20130701_20170503_01_T1
2014 136,428 11,992 4729 LC08_L1TP_140026_20140826_20170420_01_T1
*LC08_L1TP_138027_20140812_20170420_01_T1
2015 136,130 11,524 4363 LC08_L1TP_140026_20150813_20170406_01_T1
*LC08_L1TP_137028_20150808_20180524_01_T1
2016 136,300 12,108 5297 LC08_L1TP_140026_20160714_20170323_01_T1
*LC08_L1TP_137028_20160810_20170322_01_T1
2017 136,477 11,877 4715 LC08_L1TP_140026_20170717_20170727_01_T1
*LC08_L1TP_137028_20170712_20170726_01_T1
2018 136,298 11,585 4251 LC08_L1TP_140026_20180821_20180829_01_T1
*LC08_L1TP_137028_20180816_20180829_01_T1
2019 136,263 11,556 4710 LC08_L1TP_140026_20190824_20190903_01_T1
*LC08_L1TP_138027_20190826_20190903_01_T1
2020 137,064 11,529 3797 LE07_L1TP_140026_20200818_20200913_01_T1
*LC08_L1TP_138027_20200727_20200807_01_T1
2021 136,545 12,904 4686 LC08_L1TP_140026_20210813_20210819_01_T1
*LC08_L1TP_137028_20210824_20210831_01_T1

* Gegeen Lake sources.

The Airag Lake basin is interconnected with Khyargas Lake through the Khomiin Khooloi, and surplus water from Airag Lake contributes to the water level of Khyargas Lake. This hydrological pattern is expected to decline further, leading to a significant reduction in the size of the Khyargas Lake. Over the past three decades, the area of Khyargas Lake has decreased with an R2 value of 0.5174. Particularly since 2007, there has been an even greater reduction in the area of Khyargas Lake, with an R2 = 0.7434, a phenomenon closely associated with the formation of Gegeen Lake.

We have conducted lake water surface area mapping at 5-year intervals since 1991. This specific interval was chosen to clarify and align the trends in the water areas of Airag and Khyargas lakes with meteorological parameters, considering the unit of time. Since 2007, the annual lake area has been calculated based on the commencement of water accumulation in Gegeen Lake which began due to the construction of a dam for the Taishir HPP (Figs 35).

Fig. 3.

Morphological changes in the area of Airag Lake (1991–2021).

Fig. 4.

Morphological changes in the area of Gegeen Lake (2007–2021).

Fig. 5.

Morphological changes in the area of Khyargas Lake (1991–2021).

According to the satellite mapping analysis of the Airag and Khyargas lake areas and their morphological changes, the lake area remained relatively stable from 1990 to 2006. This stability can be attributed to the climate conditions during those years and variations in river inflow, which, in some years, contributed to an increase in the water surface area of the lake. However, between 2007 and 2011, the area of the lakes declined sharply, while since 2012, the water surface area has been relatively stable. This stability may be attributed to the intensive accumulation of water in Gegeen Lake from 2007 to 2011 and the subsequent maintenance of a stable water flow balance. MA of the lake area and satellite mapping changes revealed significant decreases in the lake area, particularly in the south and southwest of the Zavkhan River delta lakes, which are the primary sources of water for Airag Lake. Changes in the area of Gegeen Lake indicate a steady increase each year since August 2007 when water began to accumulate (Fig. 4). The increase in the area of Gegeen Lake since 2007 is remarkably similar in both size and duration to the decrease in the surface area of Airag and Khyargas lakes (Figs 35). Based on our measurements using the Landsat TM satellite data, the initial recorded area in 2007 was 8.9 km2. By 2021, this area had expanded to 46.85 km2, resulting in a 5.3-fold increase in the lake’s area.

Analysis of lake water surface area time series

Estimating the relationship between lake water surface areas is crucial in detecting changes over time. The time series of the water surface area of Airag, Khyargas and Gegeen lakes are calculated using NDWI (Fig. 6).

Fig. 6.

A – Temporal changes in the water surface area of the Khyargas Lake basin. B – Surface area trends for the water accumulation (2007–2011) and after the accumulation period of Gegeen Lake (2011–2021).

The decrease in the area of Airag and Khyargas lakes and the increase in the area of Gegeen Lake can be attributed to changes in the average reduction and growth of the lake areas. These changes are calculated using the method of determining the regression coefficient, which reflects the temporal trends in the MA of specific parameters.

According to these estimations, over the last 15 years, the Airag and Khyargas lakes have been steadily declining in the area by an average of approximately 1.72 to 3.54 km2 per year. This is directly related to the reduction in the Zavkhan River, which has been the main source of the lake’s water since 2007 and its accumulation in Gegeen Lake. However, the area of Gegeen Lake has been increasing by an average of approximately 3.22 km2 per year (Fig. 6).

The coefficients of determination of the linear regression equation of the surface water area of the lakes in the time series from 1990 to 2021 and from 2007 to 2021 for Airag Lake were R2 = 0.455 and R2 = 0.694, and for Khyargas Lake, they were R2 = 0.517 and R2 = 0.688. The Gegeen Lake area time series had a coefficient of determination of R2 = 0.495. These values of the coefficient of determination exceed the average, suggesting that the linear approach to these parameters is plausible. Furthermore, this indicates a strong correlation between river flows in the lake area.

Examining the water accumulation period (2007–2011) and the period after accumulation (2011–2021) of Gegeen Lake in the Taishir HPP, we observed that during the lake water accumulation period, the lake area has decreased significantly for Airag Lake (R2 = 0.986) and increased for Gegeen lake (R2 = 0.965) (Fig. 6). However, the water area for all three lakes gradually decreased after the lake water accumulation. The annual areas of Airag and Gegeen lakes are graphically plotted in Figure 7A.

Fig. 7.

A – Relationship of surface area changes between the Gegeen and Airag lakes. The light green background represents the water accumulation period for Gegeen Lake, while the light orange background corresponds to the water accumulation period of the lake. B – Relationship of surface area changes for Khargas Lake with the Gegeen and Airag lakes.

The correlation between the temporal changes in the area of these lakes is highly negative (R = –0.96, p < 0.01), indicating that approximately 92% of the causes for the decrease in the area of Airag Lake can be attributed to the increase in the area of Gegeen Lake during the period from 2007 to 2011, which corresponds to the water accumulation period of the artificial lake. However, from 2011 to 2021, the relationship between the lake areas weakened and became positive (R2 = 0.104, p = 0.36) under the influence of recent climate change conditions. In other words, the changes in the lake area of Airag Lake after the water accumulation period cannot be explained by the changes in the area of Gegeen Lake.

As shown in Figure 7B, the decrease in the surface area of Khyargas Lake exhibits a highly positive correlation (R = 0.94, p< 0.0001) with the decrease in the surface area of Airag Lake, while it demonstrates a strong negative correlation (R = –0.88, p< 0.0001) with an increase in the surface area of Gegeen Lake during the period from 2007 to 2021.

Relationship between volumes of lakes

The changes in the volume of the lakes were compared for the period from 2007 to 2021. In August 2007, water was initially stored in the HPP. The Khyargas Lake basin (2007–2021) shows changes in the water volume of lakes (Table 3, Fig. 8).

Fig. 8.

Variations in morphometric changes in the lakes. A – Depth change; B – volume change of the Airag and Gegeen lakes and C – depth and volume changes of Khyargas Lake.

Changes in the depth and volume of the Khyargas Lake basin (2007–2021).

Serial No. Year Airag Lake Gegeen Lake Khyargas Lake
Depth Volume Depth Volume Depth Volume
[m] [km3] [m] [km3] [m] [km3]
1 2007 12.587 0.694 0.598 0.001 78.66 66.034
2 2008 11.956 0.596 10.224 0.064 78.37 65.630
3 2009 11.491 0.527 12.107 0.095 78.10 65.266
4 2010 10.905 0.444 21.746 0.335 78.12 65.283
5 2011 10.643 0.408 31.923 0.738 77.78 64.819
6 2012 10.333 0.367 30.519 0.673 77.71 64.727
7 2013 9.834 0.304 29.786 0.640 77.27 64.113
8 2014 9.465 0.259 31.789 0.732 77.13 63.934
9 2015 9.107 0.216 29.329 0.620 76.97 63.704
10 2016 9.572 0.271 35.607 0.923 77.06 63.835
11 2017 9.385 0.249 31.695 0.727 77.16 63.972
12 2018 9.169 0.224 28.576 0.588 77.06 63.833
13 2019 9.136 0.220 31.661 0.726 77.04 63.806
14 2020 9.113 0.217 25.524 0.465 77.49 64.426
15 2021 10.200 0.350 31.500 0.718 77.20 64.024

Similar to the area comparison, there is a highly negative correlation between the changes in the volume of these lakes (R = –0.85 to –0.86, p = 0.06). This means that approximately 73% to 74% of the decrease in the volume of Airag and Khyargas lakes can be explained by the increase in the volume of Gegeen Lake during the water accumulation period of Gegeen Lake. Another contributing factor to the decrease in the water volume of Airag and Khyargas lakes is the influence of climate.

The Griewank model, as utilised by Bacanin et al. (2020), was used to map the nonlinear volume of the lake depression. The volume change of the lake was estimated using both field data (Tserensodnom 2000, Davaa 2018) and satellite image data. The most recent three-dimensional changes were computed to determine changes in lake volume, as depicted in Figure 9.

Fig. 9.

Three-dimensional model of water volume changes in lakes. Morphometric changes in A – Airag Lake, B – Gegeen Lake and C – Khyargas Lake.

The findings demonstrated a clear correlation between the variations in lake volume. It was observed that the volumes of Airag and Khyargas lakes tended to decrease, while Gegeen Lake showed an upward trend in volume. Additionally, the shoreline changes resulting from lake water level fluctuations were observed through field measurements (Fig. 10).

Fig. 10.

Photograph of the lakes. A – Airag Lake aerial photo, photo by Batbold Dorj; B – Gegeen Lake aerial photo, photo by Dorjzovd Enkhtur; C – Airag Lake water level decline; D – The level of water increase in the Gegeen Lake and E – Khyargas Lake water level decline, photo by Altanbold Enkhbold.

Morphometric changes in the lakes and the effects of climate

Over the past three decades, almost all of the lake areas in the Mongolian arid region have been shrinking or even drying up, as documented by Amgalan et al. (2020), Sumiya et al. (2020), Enkhbold et al. (2021), Dingjun et al. (2023) and Enkhbold et al. (2024). The reasons for the decline in the volume of Khyargas and Airag lakes can be attributed to factors not considered in the regression analysis, such as changes in surface water and climate parameters influenced by global climate change. Currently, the primary factor affecting the lake areas in Mongolian arid regions is the influence of climate (Amgalan et al. 2020, Dingjun et al. 2023, Dorjsuren et al. 2023, Enkhbold et al. 2024).

An analysis of the area and climate trends of the lakes included in the study is depicted in Figure 11.

Fig. 11.

Climate trends in years 1991–2021.

In the past 30 years, climate trends have had noticeable effects on the regions surrounding Gegeen Lake. The average temperature has risen from 1.1°C to 1.4°C, and precipitation has increased from 84 mm to 105 mm. The slight increase in precipitation could be attributed to the more temperate climate following the water accumulation at Taishir HPP in Gegeen Lake. However, in the vicinity of Airag and Khyargas lakes, the average temperature over the same period has slightly increased from 1.4°C to 1.5°C. Precipitation, on the other hand, has dropped significantly from 79 mm to 61 mm, as shown in Figure 11. Despite a weak correlation with lake surface area, these temperature changes correspond to reductions in lake surface area and volume.

Based on calculations from a satellite data study, it is evident that the volumes of Airag and Khyargas lakes have decreased, while Gegeen Lake has shown an increasing trend in volume. Notably, the air temperature and precipitation in the vicinity of these lakes did not exhibit significant trends. The hydrological changes in the Khyargas Lake basin were caused by dams constructed for hydropower stations rather than by climate change.

Impact of artificial lakes on the environment

In arid regions, it is crucial to monitor changes in lake areas and their impacts on resource availability (Kang et al. 2015, Zhang et al. 2019). One such region is the arid climate around Airag Lake, a remnant of a once vast ancient lake situated in a tectonic depression. Owing to its relatively shallow water, the lake’s water level has experienced significant fluctuations due to both modern climate change and human activities. As a consequence, the lake’s area and water volume have witnessed a notable decline. In contrast, Gegeen Lake has an artificial origin, formed more than a decade ago through the construction of a hydroelectric dam. Remarkably, both the area and water level of the lake have been on a steady increase year by year. In the arid regions of western Mongolia, the natural flow of rivers experiences fluctuations, and the impact of human activities has brought about a marked transformation in the surface area and appearance of the Airag, Khyargas and Gegeen lakes, as evidenced by satellite imagery. Therefore, it becomes essential not only to consider the morphometric parameters of lake areas but also to delve into various aspects of hydrology, such as changes in water characteristics, the influence of climate, ecological effects, artificial interventions and alternations in the physical properties of lake water. This comprehensive approach helps understand the broader environmental dynamics and challenges faced by these lakes.

The establishment of new artificial lakes has been associated with a significant positive impact on local hydrology, as noted in studies by Gurnell (1998), Wang et al. (2017) and Zhang et al. (2019). However, this practice may also disrupt the natural hydrological system and ecological patterns (Shang 2013, Fang et al. 2015, Mendsaihan et al. 2016, Wang et al. 2021). Our study reveals that a substantial reduction in the area of a natural lake can indeed have adverse implications for the hydrological and ecological environment of the lake. Climate change has led to a notable reduction in the flow of the Zavkhan River, as documented in studies by Oyunbaatar et al. (2011), Ochir et al. (2013) and Dorjsuren et al. (2024). Moreover, the construction of the hydroelectric power station since 2007 has further exacerbated the situation, leading to a 10-fold reduction in water flow downstream of the Zavkhan River compared to the long-term average (Oyunbaatar et al. 2011, Dorjsuren et al. 2024). The consequences within the Zavkhan River basin are dire, including severe drying, sand migration, desertification, a substantial decline in pasture plant growth and notable shifts in species composition.

This phenomenon is substantiated by satellite imagery, which reveals a significant increase in the area of the Zavkhan and Khungui rivers as they flow into the lower reaches of Airag Lake. The reduction in the lake’s surface area at the confluence of Airag Lake and rivers may be a primary factor disrupting the lake’s ecological balance. The growing disparity between the supply and demand of water resources has become increasingly evident. These rivers are important components of the surface water cycle, serving as a lifeline for maintaining the ecological equilibrium and facilitating social development.

The economic and social development in western Mongolia imposes significant demands on water resources, emphasising the growing disparity between supply and demand. The findings of this research not only address national needs but also represent a collective success vital for the long-term growth and sustainability of local communities.

Taishir HPP has the potential to enhance the region’s water supply, ensuring access to clean water for rural residents and livestock farming. This transformation can lead to a shift from Mongolia’s traditional nomadic way of life to a semi-settled one, facilitating the construction of urban houses and farms with necessary engineering infrastructure. Additionally, the associated increase in employment offers substantial benefits from social, economic and ecological standpoints. This study provides a model for identifying, analysing and categorising issues that may influence future changes in the lake region using satellite mapping technologies. A notable feature and novelty of our study is its estimation of relationships within the Khyargas Lake basin, addressing a significant gap in the research on natural and artificial lakes in western Mongolia. By examining temporal variations in satellite data, this study determines the relationships between lake volume, water surface area and morphological changes.

Conclusions

Variations in the area of lakes in Mongolia’s arid regions were illuminated using satellite imagery. Estimates derived from satellite data indicate the significant impact of recent human activity on lake water balance, primarily due to the observed adverse correlation between changes in lake surface area and volume.

Over a relatively short 15-year period, the Khyargas Lake basin has undergone substantial changes in its field and volume. This transformation is primarily attributed to the accumulation of water in the Taishir HPP dam, initiated in 2007 (within Gegeen Lake). Satellite imagery reveals a rapid expansion of the lake area, originating at the confluence of Airag and Gegeen lakes, and extending downstream along the Zavkhan River and its valley.

A strong correlation exists between changes in morphometric parameters of the lakes over time, particularly in the case of Airag Lake (R = –0.96, p< 0.01) and Khyargas Lake (R = 0.94, p < 0.0001). Volume changes are significantly explained by the growth of Gegeen Lake in area and volume (R = –0.88, p < 0.0001).

The Taishir HPP, with its annual capacity to generate 12 MW of power, plays a vital role in supplying electricity to nearby towns, driving employment, reducing greenhouse gas emissions and developing tourism. Nonetheless, the decline in the water area and volume of Airag Lake, a Ramsar Convention-registered site, has led to significant degradation of the wetland ecosystem. These factors pose significant constraints on environmental changes, as well as on economic and social development in western Mongolia.

This study’s distinctive focus lies in examining how artificial and naturally formed lakes in arid regions are related to one another and to the natural hydrological network, ultimately contributing to the formation of artificial reservoirs and lakes.

Idioma:
Inglés
Calendario de la edición:
4 veces al año
Temas de la revista:
Geociencias, Geografía