Surface Elevation Dynamics of Lake Karakul from 1991 to 2020 Inversed by ICESat, CryoSat-2 and ERS-1/2

Abstract

High-altitude lakes are sensitive indicators of climate change, reflecting the hydrological impacts of global warming in alpine regions. This study investigates the long-term dynamics of the water level and surface area of Lake Karakul on the eastern Pamir Plateau from 1991 to 2020 using integrated satellite altimetry data from ERS-1/2, ICESat, and CryoSat-2. A multi-source fusion approach was applied to generate a continuous time series, overcoming the temporal limitations of individual missions. The results show a significant upward trend in both water level and area, with an average lake level rise of 8 cm per year and a surface area increase of approximately 13.2 km² per decade. The two variables exhibit a strong positive correlation (r = 0.84), and the Mann–Kendall test confirms the significance of the trends at the 95% confidence level. The satellite-derived water levels show high reliability, with an RMSE of 0.15 m when compared to reference data. These changes are primarily attributed to increased glacial meltwater inflow, driven by regional warming and accelerated glacier retreat, with glacier area shrinking by over 10% from 1978 to 2001 in the eastern Pamir. This study highlights the value of integrating multi-sensor satellite data for monitoring inland waters and provides critical insights into the climatic drivers of hydrological change in high-altitude endorheic basins.

1. Introduction

High-altitude lakes are widely recognized as sensitive indicators of climate change, owing to their strong responsiveness to temperature fluctuations and glacial meltwater dynamics [1]. As key components of alpine hydrological systems—among the most vulnerable ecosystems globally—they offer valuable insights into the effects of global warming [2]. In contrast to low-altitude lakes, which are more directly influenced by anthropogenic activities, land use changes, and groundwater interactions, high-altitude lakes are predominantly shaped by cryospheric processes such as glacier retreat, snowpack variability, and seasonal melt. This fundamental difference makes them especially suitable for detecting and tracking climate-induced hydrological changes. Over recent decades, rising global temperatures have accelerated glacial melting, triggering significant shifts in lake water levels and associated hydrological processes. These transformations are driven by a complex interplay among glacial melt, precipitation regimes, and regional climate variability, highlighting the critical importance of monitoring high-altitude lakes to understand broader cryospheric and hydrological responses to climate change [3].

Understanding the dynamic responses of high-altitude lakes to climate change requires robust and comprehensive monitoring methods. Traditional in situ measurements, while valuable, often face limitations in remote and difficult-to-access alpine environments. This is particularly true for high-elevation lakes surrounded by complex and rugged terrain, where field observations can be logistically challenging and time-consuming [2]. Consequently, the development and application of advanced remote sensing technologies have become essential for long-term and large-scale monitoring of these critical water bodies.

Remote sensing techniques offer several advantages for tracking changes in lake water levels over time. Satellite altimetry, in particular, provides a practical and efficient means to measure lake surface elevation with high precision [4,5]. By employing satellite missions such as ICESat, CryoSat-2, and ERS-1/2, scientists can obtain consistent and high-resolution data that cover vast and otherwise inaccessible areas [6,7,8]. These missions use laser and radar altimetry to collect elevation data, allowing for the detection of even subtle water level variations over extended temporal scales. However, relying on individual satellite missions for water level reconstruction presents certain limitations. For instance, while ICESat provides valuable elevation data, its operational period was limited to 2003 to 2010, which restricts the temporal scope of any analysis based solely on its data [9]. CryoSat-2, launched in 2010, has enhanced capabilities for measuring ice and water elevations but may also encounter challenges due to its polar orbit, which limits coverage of lower-latitude regions and specific lake sites like Karakul [10]. Similarly, ERS-1/2, while pioneering in radar altimetry, has not operated since the early 2000s, and the temporal gaps in the data can hinder comprehensive assessments of long-term water level changes [11,12]. To address these limitations, the integration of multiple satellite datasets becomes essential. Combining data from ICESat, CryoSat-2, and ERS-1/2 allows for a more comprehensive and continuous dataset that enhances the reliability and accuracy of water level measurements [13,14]. This multi-source approach can fill in temporal gaps and provide a more robust framework for analyzing hydrological changes over the decades. It enables scientists to leverage the strengths of each satellite mission while compensating for their individual weaknesses.

As a critical high-altitude lake located in the Pamir Plateau, Lake Karakul exemplifies the challenges and opportunities presented by climate change in fragile alpine ecosystems [15,16]. Its hydrological dynamics are intricately linked to glacial melt and regional climatic conditions, making it an ideal case study for examining the effects of environmental changes on water levels [17]. Given the lake’s importance as a water resource for local ecosystems and communities, understanding its historical water level trends is essential for effective management and conservation strategies.

The prevailing focus of research on high-altitude lakes is on the analysis of single satellite data or on the changes in lake water levels during only part of the time period. There is a paucity of systematic research on the fusion of long time series and multi-satellite data [18]. Existing studies on lakes such as Lake Karakul still lack knowledge of the trend of water level changes, the relationship with area changes and the mechanism of the influence of climatic factors [15]. There is also a paucity of unified and effective methods for integrating multi-source satellite data to accurately reconstruct lake water level changes. Therefore, considering the current limitations in long-term lake monitoring and the importance of accurately quantifying lake dynamics under climate change, this study aims to address these challenges through the following specific research objectives: first, to integrate the data from ICESat, CryoSat-2 and ERS-1/2 to accurately reconstruct the water level changes in Lake Karakul. Secondly, the trend of water level and area changes in Lake Karakul during 1991–2020 will be analyzed and the relationship between the two changes will be explored. Thirdly, the main climatic influences on the water level and area changes in Lake Karakul and their mechanisms will be further explored. By establishing a comprehensive dataset for Lake Karakul, the findings will provide insights into the responses of closed alpine lakes to environmental stressors, informing both regional water management practices and global hydrological models. Ultimately, this study aims to enhance our understanding of the interplay between climate dynamics and high-altitude water resources, supporting sustainable management in a rapidly changing world.

2. Methods and Data

2.1. Study Area

Lake Karakul is situated in the eastern part of Tajikistan (Figure 1), within the Pamir Plateau, often referred to as the Roof of the World [19]. This high-altitude lake lies at an elevation of approximately 4000 m above sea level, making it one of the highest lakes in Central Asia. Covering an area of about 380 square kilometers, the lake is fed by numerous small rivers and streams, most notably the influx of meltwater from the nearby Fedchenko Glacier, one of the longest glaciers in the world [20]. This glacial meltwater is critical for sustaining the hydrology of the lake, particularly during the warmer months when temperatures rise and ice begins to melt. The Pamir Plateau is characterized by a harsh alpine climate, with cold winters and short, cool summers [21]. The region experiences significant seasonal variations in temperature and precipitation, which can profoundly impact water levels in the lake. In addition to its hydrological significance, Lake Karakul holds cultural and ecological importance. Given the ongoing impacts of climate change—manifested through rising temperatures and altering precipitation patterns—understanding the water level dynamics of Lake Karakul is increasingly important.

2.2. Data Sources

The analysis of water level dynamics of Lake Karakul relies on three primary satellite missions: ICESat, CryoSat-2, and ERS-1/2. Each of these missions provides unique capabilities and datasets that contribute to the comprehensive assessment of lake water levels. Launched by NASA in January 2003, ICESat operated until October 2009 and focused primarily on measuring ice sheet elevations, as well as cloud and land topography [22,23]. Its laser altimetry system provided high-resolution elevation measurements with a vertical accuracy of approximately 10 cm. The data of ICESat are particularly valuable for tracking changes in ice and water elevations, making it relevant for assessing the water levels of alpine lakes [24,25]. CryoSat-2 was launched by the European Space Agency (ESA) in April 2010, succeeding the original CryoSat mission. It features a radar altimeter designed specifically for measuring the thickness of ice and monitoring changes in ice sheets, glaciers, and sea ice [26]. The satellite operates in a polar orbit, allowing it to collect elevation data across a wide range of latitudes, including the high-altitude regions where Lake Karakul is located. CryoSat-2 has the capability to provide accurate measurements of water surface elevations with a vertical precision of around 2 cm [27]. The ERS-1 and ERS-2 missions, launched by ESA in 1991 and 1995, respectively, were among the first satellites to employ radar altimetry for monitoring the Earth’s surface. These missions collected valuable data on various hydrological features, including lakes, before ceasing operations in the early 2000s. ERS satellites have provided a wealth of historical data that can be utilized for long-term trend analysis [28]. The available time periods for the three satellite altimetry missions used in this study are: ERS-1/2 (1991–2002), ICESat-1 (2003–2009), and CryoSat-2 (2010–2020), ensuring comprehensive temporal coverage from 1991 to 2020. In addition, the JRC Global Surface Water dataset, developed by the European Commission’s Joint Research Centre, provides long-term surface water data from 1984 to the present, based on Landsat imagery [29]. This dataset includes metrics such as water occurrence, seasonality, and changes, which are essential for validating the satellite-derived water level estimates. By comparing the integrated satellite altimetry data with the JRC Global Surface Water measurements, we can ensure the accuracy and reliability of the reconstructed water level trends for Lake Karakul. A specially optimized global reanalysis dataset, ERA5-Land (ERA5L), was also utilized at https://cds.climate.copernicus.eu (accessed on 18 September 2023) to obtain mean annual temperature (MAT), mean annual precipitation (MAP), mean annual evapotranspiration (MAET), and mean annual snow depth (MASD) data for the years 1991–2020.

2.3. Workflow Overview

To provide a clear overview of the methodological framework, a comprehensive technical workflow is presented in Figure 2. This diagram outlines the primary data inputs (satellite altimetry, optical imagery, MODIS LST, and reanalysis climate data), the respective data processing steps (e.g., format conversion, quality control, and spatial filtering), and the final outputs including lake area, water level, and land surface temperature time series, as well as their response to climate variability.

2.4. Water Level Estimation

The extraction of Lake Karakul’s water levels from ICESat-1 (GLA14) data for 2006–2012 involved selecting elevation points within the lake boundaries and converting them to a unified reference frame. A 100-m buffer zone was applied around the lake to minimize shoreline interference, and only points within this zone were retained. Water surface elevations were adjusted to WGS84 ellipsoidal heights by subtracting the geoid height and applying a 0.7-m correction. To ensure data accuracy, outliers were identified through visual inspection and statistical deviation analysis and subsequently removed. Daily average elevations of the filtered points were calculated as the lake water level for each observation day, forming a consistent time series for further analysis. The Ice-1 retracker was applied for ERS-1/2, which is commonly used in inland water studies. Due to the coarse temporal and spatial resolution and older waveform processing algorithms, greater uncertainties were expected in these observations.

CryoSat-2 data for 2010–2020 were used to derive annual water levels of Lake Karakul. Data were filtered to retain only valid surface-type records (surf_type = 1) within a 100-m buffered area around the lake boundary. Corrections were applied using predefined parameter ranges to eliminate unreliable measurements. Outliers caused by signal noise or instrument errors were removed through visual and statistical screening. Valid elevation points were averaged annually to construct a consistent water level trend, enabling analysis of the lake’s hydrological response to climate variability. For CryoSat-2, the OCOG retracker (Offset Centre of Gravity) was used, which is suitable for specular water surfaces and commonly adopted in hydrological applications.

For the ERS-1/2 data, water levels were derived by extracting valid altimetry points within a buffered zone surrounding Lake Karakul. Measurements were filtered using alt_state_flag values of 2 or 3 and ice1_qual_flag of 0 to ensure retracked, high-quality data. Corrections were verified against valid thresholds to remove any erroneous values. Final water level estimates were calculated using standard altimetry formulas, and outliers were removed to produce a reliable lake water level series from 1991 to 2000. ICESat-1 data were processed using the centroid algorithm embedded in GLA14 product, which ensures high accuracy for water surface elevation retrieval.

To account for spatial geoid variation across Lake Karakul, all satellite altimetry-derived elevations were converted to WGS84 ellipsoidal heights using the EGM2008 geoid model. The geoid undulation across the lake was found to be within ±0.2 m, based on EGM2008 grid values [30]. For CryoSat-2, the drifting orbit results in varying ground track intersections across different cycles. However, due to the relatively small size and stable surface of Lake Karakul, the spatial effect on water level estimation is minimal. This is further supported by the low RMSE (0.147 m) between CryoSat-2 and ICESat-1 water levels during their overlapping period (2010–2012), indicating good consistency and minimal geoid-related bias.

2.5. Data Integration Methodology

The integration of water level data for Lake Karakul from different satellite altimetry missions was carried out through a standardized calibration and correction workflow to ensure consistency and continuity across datasets. Three main datasets were used: ERS-1/2 (1991–2002), ICESat-1 (2003–2009), and CryoSat-2 (2010–2020). Due to differences in data acquisition periods, sensor types, and reference systems, systematic corrections were necessary.

First, the ICESat-1/GLA14 data were compared against the inland water surface heights (IWSH, 2006–2009) [31] as a reference during their overlapping period. The RMSE between the datasets was approximately 0.3 m, indicating a moderate level of uncertainty but not a systematic bias that could be directly corrected. Therefore, no uniform offset was applied to ICESat-1 data, and this uncertainty was taken into account in subsequent analyses. In the next step, the overlapping period between ICESat-1 and CryoSat-2 (2010–2012) was used to evaluate consistency. The CryoSat-2 dataset exhibited a systematic deviation of +1.2 m, which was corrected by subtracting 1.2 m from all CryoSat-2 water level values. The root mean square error (RMSE) between the adjusted CryoSat-2 and ICESat-1 values during the overlapping years was 0.147 m, with residuals showing no obvious spatial clustering, indicating good agreement between the datasets.

For ERS-1/2, which lacked temporal overlap with either ICESat-1 or CryoSat-2, an indirect correction was applied. We identified comparable multi-year mean water levels between ERS-1/2 (1991–2002) and ICESat-1 (2003–2009) through statistical extrapolation of overlapping climate variables and stable elevation patterns. A mean offset of 0.6 m was calculated and applied to ERS-1/2 data. The relatively higher uncertainty in ERS-1/2 measurements may result from the combined effects of sensor limitations, orbital characteristics, and geophysical correction quality, rather than the waveform retracking method alone. To quantify this uncertainty, we analyzed inter-mission variance, resulting in an overall RMSE of ±0.189 m for the integrated series.

Compared with previous long-term lake water level reconstructions that typically rely on single-mission data or limited overlap periods, our approach presents three innovations: (1) a 30-year continuous time series with cross-sensor calibration; (2) quantitative uncertainty assessment using RMSE and residuals in overlapping periods; and (3) the integration of historical ERS-1/2 data via indirect validation, expanding the temporal coverage to 1991. These improvements enhance the reliability and temporal resolution of the lake hydrology assessment, especially in data-scarce, high-altitude regions like the Pamir Plateau.

2.6. Trend Analysis

The trend analysis of water level and area changes in Lake Karakul from 1991 to 2020 employs both linear regression and the Mann–Kendall test to provide a robust understanding of temporal dynamics. These methodologies facilitate the detection of trends while accounting for potential non-stationarity in the data.

3. Results

3.1. Analysis of Climate Change in the Lake District

To better understand the hydrological dynamics of Lake Karakul, it is essential to analyze the long-term trends in regional climatic factors that influence lake level and area changes. Therefore, this section examines the interannual variations and trends in key climate variables, including MAT, MAP, MAET, and MASD, from 1991 to 2020. This analysis provides the foundational context for interpreting glacial meltwater contributions and hydrological responses discussed in later sections.

As Figure 3 illustrates, the MAT of Karakul Lake exhibited an overall slow increasing trend from 1991 to 2020, with an average annual rate of change of 0.134 °C/year (Figure 3). The MAT of Karakul Lake over the last 30 years was –4.02 °C, with the highest recorded temperature of −2.89 °C occurring in 2016 and the lowest of −5.34 °C in 1991 (Figure 3). A Mann–Kendall trend test for MAT (

Table 1. Results of M-K trend test for climate factors.

	Z	p	Sslope	Trend
MAT	1.054	0.292	0.013	no trend
MAP	−1.156	0.248	−0.179	no trend
MAET	1.530	0.126	0.002	no trend
MASD	−1.530	0.126	−0.736	no trend

) revealed a weak increasing trend from 1991 to 2020 onwards. The inter-annual variation in MAT exhibited a fluctuating downward trend, with a mean value of 57.85 mm over the last 30 years. The lowest annual precipitation occurred in 2019 with a minimum value of 42.39 mm, and the maximum precipitation occurred in 1999 (68.92 mm). The MAET has been observed to gradually increase, with an average of 0.444 mm from 1991 to 2020, annual evapotranspiration reaching a maximum of 0.569 mm in 2000 and a minimum of 0.336 mm in 1991 (Figure 3). Conversely, the MASD exhibited a downward trend, with an average of 76.70 mm from 1991 to 2020. Correspondingly, annual snow depth reached a maximum of 123.14 mm in 1994 and a minimum of 43.98 mm in 2020 (Figure 3).

3.2. Trends in Lake Changes

3.2.1. Changes in the Lake from 1991 to 2020

The water level change series of Lake Karakul from 1991 to 2020 was obtained by integrating elevation values extracted from three satellite altimetry datasets (Figure 4). The figure illustrates a segmented upward trend in the lake’s water level over the three decades. The lowest recorded level, at 3915.78 m, occurred in 1991, as indicated by the green bar on the left. From this point, the water level steadily increased, with some fluctuations, ultimately reaching its highest value of 3917.99 m in 2018, marked in red on the right.

The linear regression analysis, represented by the red trend line, yields the equation y = 0.079x + 3759.167 and an R² value of 0.874, demonstrating a strong fit to the observed data. The overall rate of water level change during the study period was 8 cm per year. Additionally, the Mann–Kendall (M-K) trend test was applied to the 1991–2020 time series, yielding a Z statistic of 6.458, which is greater than the critical value of 2.58 at a 99% confidence level, and a slope (S_slope) of 0.082. This indicates that the increase in water level over the past 30 years is highly significant (p < 0.001). These results collectively highlight a pronounced and statistically significant upward trend in Lake Karakul’s water level, emphasizing the impact of environmental changes on the lake’s hydrology.

Using the lake extraction results from 1991 to 2020 to analyze the long time series change in the lake area of Karakul Lake, the lake area of Karakul Lake was found to have changed a lot over the past 30 years, with an overall trend of increasing and then decreasing and then slowly increasing; the lake area was the smallest in 1991 with a minimum area of 387.95 km² and reached a maximum value of 414.83 km² in 2020 (Figure 5). After analyzing the overall trend of lake area change by M-K trend test and Sen’s slope analysis (

Table 2. Changes in lake area of Lake Karakul.

Time	Area/km2	Annual Change in Area/km2	K	Time	Area/km2	Annual Change in Area/km2	K
1991	387.95	2.62	0.68%	2006	402.85	2.25	0.56%
1992	390.57	2.62	0.67%	2007	403.09	−0.11	−0.03%
1993	393.18	2.62	0.67%	2008	403.58	0.60	0.15%
1994	395.72	2.62	0.67%	2009	402.78	−1.00	−0.25%
1995	397.67	1.87	0.47%	2010	404.74	1.96	0.49%
1996	398.80	1.87	0.47%	2011	404.78	0.12	0.03%
1997	402.80	3.68	0.92%	2012	407.98	3.20	0.79%
1998	403.39	0.17	0.04%	2013	409.19	1.34	0.33%
1999	406.97	3.58	0.89%	2014	409.76	0.79	0.19%
2000	406.56	−0.41	−0.10%	2015	412.14	2.16	0.53%
2001	402.71	−3.85	−0.95%	2016	410.69	−1.45	−0.35%
2002	399.66	−3.05	−0.76%	2017	409.93	−0.76	−0.19%
2003	398.61	−1.05	−0.26%	2018	414.03	4.10	1.00%
2004	398.37	−0.24	0.62%	2019	414.82	0.79	0.19%
2005	400.85	2.48	0.56%	2020	414.83	0.01	0.00%

), it was found that the lake area showed a highly significant increasing trend (Z = 5.949, S_slope = 0.801, p < 0.001), and the linear fit result of R² = 0.79, which was a better fit, further illustrated that the lake area had a significant increasing trend of lake area.

To better illustrate the spatial dynamics of Lake Karakul, we analyzed the annual water boundaries extracted from Landsat imagery from 1991 to 2022 (Figure 6). The lake’s surface extent has undergone considerable interannual fluctuations, with evident shoreline expansion and contraction patterns over the past three decades. During the 1990s (Figure 6a), lake boundaries exhibited relatively frequent shifts, particularly in the southwestern and southeastern parts. From 1999 to 2004 (Figure 6b), the lake area continued to fluctuate but maintained a more stable contour, with minor expansions observed mainly in the northern margin. In the third period (2015–2020, Figure 6c), lake shrinkage became more pronounced, especially along the eastern and southern shores, indicating a long-term trend of water loss. The superimposed boundary colors reveal that the interannual water body changes predominantly occurred in peripheral shallow zones, suggesting high sensitivity of these areas to hydrometeorological fluctuations. Overall, the spatial pattern confirms that Lake Karakul has experienced substantial dynamic changes over time, aligning with the trends detected in surface area and water level time series.

By calculating the lake area motivation (K), it was found (Table 2) that the total lake motivation from 1991 to 2020 was 0.26%, and the lake area increased by 29.49 km²; among them, the fastest increase in lake area was observed from 2017 to 2018, which increased by 4.1 km² from 409.93 km² to 414.03 km², and the lake area motivation K was 1.00%, and the largest decrease in lake area from 2000 to 2001, from 406.56 km² in 2000 to 402.71 km² in 2001, for a total decrease in lake area of 0.95%.

3.2.2. The Relationship Between Lake Level and Area

The comparison between the surface area of Lake Karakul, extracted from Landsat imagery using the JRC Global Surface Water dataset, and the water level series derived from integrated satellite altimetry data reveals a strong similarity in their trends over the past 30 years (Figure 7). As shown in the figure, both the surface area (indicated by black points) and the water level (indicated by red points) generally increased from 1991 to 2020. The fluctuations in water level correspond closely to the variations in area, suggesting a clear linkage between these two parameters. Analyzing the trends more quantitatively, the Mann–Kendall (M-K) test results indicate that both the lake area (Z = 5.629) and the water level (Z = 6.458) exhibit highly significant upward trends. These findings highlight that Lake Karakul has experienced substantial and statistically significant growth in both area and water level over the three-decade study period. The synchronous increase in these metrics underscores the influence of climatic and hydrological changes, as expanding lake area is often associated with rising water levels, reflecting an overall positive hydrological balance.

The analysis of the regression relationship between Lake Karakul’s water level and area is presented in Figure 8. As shown, there is a relatively good fit between the two variables, with a coefficient of determination R² = 0.504. The regression equation, y = 5.95x − 22,912.70, indicates a positive relationship between lake area (y) and water level (x), suggesting that changes in one are closely associated with changes in the other. Further correlation analysis supports this relationship, revealing a correlation coefficient r = 0.710, which is significant at the 0.01 level. This strong positive correlation implies a significant interaction between the two variables: as the lake water level rises, the area also expands, and conversely, a drop in water level results in a reduction in the area. The shaded 95% confidence interval around the regression line indicates the range within which the true relationship between water level and surface area is likely to lie.

3.3. Lake Response to Climate Change

3.3.1. Response of Lake Levels to Climate Change

As shown in

Table 3. Correlation of lake level changes with climatic factors.

		Level	MAT	MAP	MAET	MASD
Level	Correlation	1	−0.130	0.041	0.121	0.167
	Significance (two-tailed)		0.500	0.831	0.531	0.388

Note: Correlations are significant at the 0.05 level (two-tailed).

, the results show that the lake level is positively correlated with MASD (r = 0.167), MAP (r = 0.041) and MAET (r = 0.121), and negatively correlated with MAT (r = −0.130), and the correlation is not significant. It indicates that climatic factors have less influence on the change in lake level in Karakul Lake, and the change in climatic factors is not the direct cause of the change in lake level. And the correlation coefficient r between lake area and water level is 0.710, indicating that the lake area is the main factor causing the change in lake level. Figure 9 shows the comparison of changes between lake water level and climate factors, from which it can be seen that there are differences between changes in lake water level and climate factors, with different fluctuation trends.

3.3.2. Response of Lake Area to Climate Change

As demonstrated in

Table 4. Correlation of lake area with climate change.

		Area	MAT	MAP	MAET	MASD
Area	Correlation	1	0.083	−0.115	0.137	−0.376 *
	Significance (two-tailed)		0.661	0.544	0.470	0.040

Note: *. Correlations are significant at the 0.05 level (two-tailed).

, there was a positive correlation between changes in lake area and MAT and MAET, and a negative correlation between changes in lake area and MAP and MASD throughout the study period from 1991 to 2020. However, the correlation between lake area and MASD was found to be significant. This indicates that MASD is the main factor influencing the change in lake area during the study period from 1991 to 2020. Figure 10 shows the comparison between changes in lake area and changes in climate factors from 1991 to 2020, from which it can be seen that the magnitude of change between lake area and each factor varies.

4. Discussion

4.1. Advantages and Limitations of Data Integration and Methodology

Integrating ICESat, CryoSat-2, and ERS-1/2 datasets enables a near-continuous reconstruction of Lake Karakul’s surface elevation from 1991 to 2020, effectively addressing the challenges of data scarcity in high-altitude, remote regions. Each satellite altimetry mission contributes unique advantages: ERS-1/2 provides early radar altimetry records that establish long-term historical baselines; ICESat, based on laser altimetry, offers high-precision spot elevation data essential for capturing interannual variability; and CryoSat-2 extends the observation period into recent years with dense spatial coverage and improved radar altimetry capabilities, particularly over small inland water bodies. Temporal overlaps between missions—such as between ICESat and CryoSat-2—facilitate inter-calibration, enabling correction of systematic biases and ensuring consistency across datasets. This cross-validation significantly enhances the reliability of the reconstructed time series. Compared with using a single altimetry source, the integrated approach not only captures short-term fluctuations but also reveals long-term elevation trends, offering a more comprehensive understanding of hydrological variations. The merged record highlights both seasonal dynamics and decadal-scale lake level changes, which are critical for assessing the impact of regional climate variability and cryospheric processes on closed-basin lake hydrology. Overall, this multi-source methodology establishes a robust foundation for subsequent analyses and provides valuable insights for monitoring environmental change in data-limited mountainous regions.

Despite these advantages, limitations remain. Spatial coverage is uneven—CryoSat-2’s polar orbit and ICESat’s fixed tracks reduce representativeness [32]. Atmospheric interference and complex lake geometries can further affect measurement quality. Moreover, differences in temporal and vertical resolution necessitate extensive calibration, and short-term dynamics like seasonal variations may still be underrepresented. Thus, while the integration improves temporal resolution and accuracy, further methodological refinement and complementary observations are essential for fully capturing surface elevation dynamics.

4.2. Relationship Between Lake Level and Area Changes and Its Influencing Factors

From 1991 to 2020, the water level and surface area of Lake Karakul showed an overall increasing trend, despite a general decrease in precipitation and only a slight rise in temperature. Evaporation exhibited a significant downward trend, reflecting a regional climate transition from cold-wet to warm-dry conditions. This aligns with observed climatic shifts across the Pamir Plateau, where rising temperatures and declining precipitation have accelerated glacier retreat and enhanced meltwater input [33,34]. Correlation analyses indicate that the lake’s surface area has a weak positive correlation with temperature (r = 0.083) and evapotranspiration (r = 0.137), and a negative correlation with precipitation (r = −0.115) and snow depth (r = −0.376, p < 0.05), suggesting that meltwater from glaciers and snow plays a dominant role in sustaining lake expansion. Similar patterns were reported for Rangkul and Shorkul lakes, where increased temperatures drove meltwater contributions from glaciers and permafrost [35,36]. The influence of orographic barriers and westerly winds restricts moisture input to the eastern Pamir, making Lake Karakul highly sensitive to changes in cryospheric dynamics [37].

The relationship between water level and surface area reflects typical hydrological responses of alpine lakes. A strong positive correlation was observed in this study, similar to that of Qinghai Lake on the Tibetan Plateau, where glacial melt and shifting precipitation regimes jointly led to lake expansion [38,39]. This indicates that in regions where glacial contributions are significant, lake area dynamics can be a sensitive indicator of cryospheric change. In contrast, some closed-basin lakes, such as Lake Urmia in Iran, experienced shrinking due to reduced inflow and intensive water use [40,41]. These comparisons underline the buffering role of glacier meltwater in sustaining high-altitude lake systems. Recent studies have confirmed the significant retreat of glaciers near Lake Karakul—such as the Garmo and Fedchenko glaciers—which contributes to increased meltwater inflow and alters local hydrology [20,42]. The continued rise in temperature may further intensify glacier melt, suggesting a persistent trend of lake expansion under future warming.

4.3. Ecological and Societal Implications

Changes in Lake Karakul’s water level and surface area over recent decades reflect the dynamic responses of high-altitude lake systems to climatic and hydrological drivers. The sustained rise in surface elevation not only reshapes the physical landscape but also triggers cascading ecological and societal impacts [43]. These hydrological changes are primarily driven by intensified glacier melt and increased precipitation, both linked to regional warming trends observed across the Pamir Plateau. Ecologically, increased water levels can expand habitat availability for aquatic and semi-aquatic species, similar to trends observed in Qinghai Lake. However, However, such expansion is not uniformly beneficial; the inundation of wetlands and shallow littoral zones may lead to the displacement of specialized plant communities and nesting grounds. Rapid elevation changes may disrupt shoreline ecosystems, alter nutrient cycling, and affect species composition, as seen in Lake Titicaca where fluctuating levels impacted bird nesting and fish spawning habitats [44]. In the context of Central Asia’s fragile ecosystems, even moderate hydrological variability can have disproportionate effects due to the limited resilience of alpine biodiversity. From a societal standpoint, water level rise can enhance resource availability but also pose risks. For communities near Lake Karakul, benefits such as increased water supply must be weighed against potential losses of arable land and infrastructure threats, echoing experiences around Issyk-Kul Lake.

4.4. Future Research Directions

This study provides valuable insights into the hydrological dynamics of Lake Karakul and its relationship with climate change. However, several directions remain for future research to enhance the robustness and scope of the findings. One potential avenue is the use of higher temporal and spatial resolution datasets to capture finer-scale variations in lake dynamics. Although satellite altimetry data from ICESat, CryoSat-2, and ERS-1/2 have offered long-term observations, their temporal frequency and limited spatial coverage constrain their effectiveness [13]. Future research could leverage advanced satellite missions such as Sentinel-3, which offers global coverage and enables more frequent monitoring of surface water levels [45]. In particular, Sentinel-3’s SAR altimetry and synergy with optical sensors allow simultaneous measurement of water extent, elevation, and reflectance properties, facilitating multi-parameter analysis of lake systems. Additionally, integrating optical and thermal remote sensing data—such as Landsat imagery—can improve the assessment of lake surface area dynamics and water temperature, offering deeper insights into the ecological consequences of hydrological changes. Combining these datasets with in situ observations, such as weather station data and glacier mass balance measurements, would also strengthen model validation and reduce uncertainties. Moreover, future studies should explore the socio-hydrological dimensions of lake changes, including community adaptation, water resource governance, and transboundary water cooperation, particularly given Lake Karakul’s strategic location near international borders.

5. Conclusions

This study provides a comprehensive analysis of the water level dynamics of Lake Karakul over the past three decades, utilizing satellite altimetry data from ICESat, CryoSat-2, and ERS-1/2. The results indicate a clear upward trend in the lake’s water level and surface area from 1991 to 2020, driven primarily by increased glacial meltwater inflow and regional climatic changes. The significant positive correlation between water levels and surface area expansion underscores the direct relationship between hydrological changes and the physical extent of the lake.

The findings of this research not only enhance our understanding of Lake Karakul’s hydrology but also contribute to broader knowledge of high-altitude lakes as sensitive indicators of climate change. The study highlights the critical role of glacial meltwater in sustaining the hydrological balance of such lakes, particularly in regions experiencing accelerated glacial retreat due to rising global temperatures. Moreover, the research demonstrates the utility of integrating multiple satellite datasets to reconstruct long-term lake level trends, addressing the limitations posed by individual missions and providing a more comprehensive and continuous time series.

This study underscores the importance of multidisciplinary research that integrates ecological, hydrological, and socio-economic perspectives. Future work should focus on expanding data sources, refining modeling techniques, and incorporating future climate scenarios to better understand how the lake’s hydrology will evolve under different environmental conditions. Additionally, long-term monitoring and cross-regional comparisons with other high-altitude lakes could provide further insights into the broader trends of climate-induced changes in such ecosystems.