A satellite remote sensing-based ice lake morphological information monitoring method for an area with insufficient data

Abstract

The application discloses a kind of ice lake form information monitoring methods of lack of data area based on satellite remote sensing, comprising the following steps:1) height-measuring satellite data acquisition and pretreatment;2) ice lake water level extraction based on adaptive threshold waveform resetting method;3) multi-source optical satellite data acquisition and pretreatment;4) water body index calculation;5) ice lake contour extraction based on machine learning method;6) water level-area-water quantity relationship construction.The application improves the precision of water level extraction of height-measuring satellite data by combining adaptive threshold waveform resetting method, and combines several traditional water body indexes, improves the precision of ice lake contour extraction by means of random forest classification method, to a certain extent, makes up for the blank of ice lake form information monitoring work in lack of data area, and provides efficient technical support for plateau area ice lake evolution analysis and disaster warning.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of remote sensing and geographic information, and relates to a remote sensing monitoring technology for ice lakes in an area lacking data, in particular to an ice lake morphological information monitoring method based on satellite remote sensing. BACKGROUND

[0002] An ice lake is a water body in an alpine region formed by the collection of glacial melt water, the obstruction of ice bodies or the formation of moraine dams, and is widely distributed in the Qinghai-Tibet Plateau, the Himalayas and other polar and high-altitude regions, and is one of the important freshwater resources. In recent years, with global warming, glacial retreat has intensified, and the number and area of ice lakes have continued to grow, which has made the problem of glacial lake outburst flood (GLOF) increasingly serious, threatening the ecological safety and infrastructure safety downstream. Ice lake morphological information is the focus of its change monitoring and risk assessment, including the outline, water level and water volume of the ice lake, which respectively reflect its spatial range, water body elevation and water storage, and is of great significance to GLOF early warning, regional water resource assessment and climate change research.

[0003] Due to the limitations of natural conditions such as complex terrain, high altitude and harsh climate, ice lakes are often located in remote and rarely visited areas, and traditional ground monitoring methods are difficult to carry out stably for a long time. The installation and maintenance of water level meters and flow meters are difficult and costly, and the coverage is limited, resulting in a large number of ice lakes being in a "data lacking" state for a long time, and the morphological information is seriously lacking. Especially in the risk prevention and control of GLOF with strong suddenness and destructive power, the lack of continuous and uniform monitoring data has become a key bottleneck restricting the construction of disaster assessment and early warning system. In recent years, remote sensing technology has become an important supplement to ice lake monitoring due to its advantages of large-scale, periodic and non-contact observation.

[0004] Remote sensing monitoring of ice lake morphological information mainly focuses on three aspects: boundary extraction, water level inversion and water volume estimation. However, the existing methods have obvious shortcomings. In terms of boundary extraction, at present, it mainly depends on medium-high resolution optical images to identify the water body edge through threshold segmentation and spectral feature analysis, but when disturbed by ice and snow and complex terrain, it is easy to misjudge or miss, especially in high albedo environment, it is often difficult to accurately determine the real lake boundary. In terms of water level inversion, the mainstream method relies on digital elevation model (DEM) to calculate the water surface elevation, but the existing DEM data has limited resolution and accuracy, which makes the water level extraction easily affected by terrain undulation and noise, resulting in large water level value error. In addition, although some studies have introduced satellite altimetry data, such as Jason-1 / 2 / 3 series and SWOT satellite, the water surface reflection signal is obtained by radar altimeter, but these data still face problems such as waveform interference and preset tracking door offset in the processing process, and the traditional redefinition method cannot effectively overcome this challenge, thereby affecting the final water level accuracy, which cannot meet the fine needs of ice lake in disaster warning and water resource management. In terms of water volume estimation, most methods use a simplified water depth model to estimate the water volume by simply multiplying the lake area and water depth, but this method fails to fully consider the complex geometry and lake bottom topography changes of ice lake, often ignoring factors such as edge slope and local depression, thereby limiting the reliability and accuracy of the estimated results.

[0005] In summary, most studies only focus on a single element, and lack of collaborative monitoring method of elements, which leads to the inability to achieve full-time and systematic monitoring of ice lake. The existing monitoring method has great limitations in remote sensing monitoring of ice lake in high-cold and data-scarce areas, and an overall solution that comprehensively utilizes multi-source data, combines machine learning and advanced signal processing technology is urgently needed to improve the accuracy and practicality of ice lake morphological information extraction. SUMMARY

[0006] Based on the above technical defects, the present application provides a satellite remote sensing based ice lake morphological information monitoring method in data-scarce areas, which realizes full-time and systematic monitoring of ice lake through a multi-element collaborative monitoring method, so as to improve the accuracy and practicality of ice lake morphological information extraction. The purpose of the present application is realized by the following technical scheme:

[0007] A satellite remote sensing based ice lake morphological information monitoring method in data-scarce areas, comprising the following steps:

[0008] 1) Altimeter satellite data acquisition and preprocessing: altimeter satellite data preprocessing includes denoising and spatio-temporal registration,

[0009] 2) Ice lake water level extraction based on adaptive threshold waveform redefinition method ATR preprocessed altimetric satellite data: Apply adaptive difference sequence to identify the leading edge of the waveform, determine the threshold for detecting the leading edge of the waveform; further expand the range of sub-waveform by weighted redefinition gate; calculate the adaptive threshold to determine whether there is an effective redefinition gate position in the waveform; obtain the distance correction value based on the time of the redefinition gate position to correct the water level measurement formula, and obtain the final water level;

[0010] 3) Multi-source optical satellite data acquisition and preprocessing: The preprocessing of multi-source optical satellite data includes atmospheric correction and geometric correction;

[0011] 4) Water body index calculation based on preprocessed multi-source optical satellite data: The water body index calculated based on preprocessed multi-source optical satellite data includes normalized difference water index (NDWI), modified normalized difference water index (MNDWI), and automatic water extraction index (AWEI), which is divided into shadow formula (AWEI sh ) and non-shadow formula (AWEIn sh );

[0012] 5) Ice lake contour extraction based on machine learning method: Random forest classification method is used to accurately identify the ice lake contour by taking four water body indexes NDWI, MNDWI, AWEI sh , and AWEI nsh information as input.

[0013] 6) Water level-area-water volume relationship construction: Water level data is obtained from step 2), and contour change is obtained from step 5) and further calculated area change, water level-area relationship is established first, assuming the lake is conical or circular, area-water volume relationship is constructed, and finally water level-area-water volume relationship is obtained.

[0014] Further scheme, the altimetric satellite data Jason series satellite used in step 1) is a joint development of the United States National Aeronautics and Space Administration (NASA) and the French National Space Research Center (CNES), mainly used for global ocean surface height monitoring. The water surface height data provided by this series of satellites can be obtained through NASA's ocean atmosphere data system (PO.DAAC), covering the area where the ice lake is located, providing important data support for ice lake water level monitoring.

[0015] After obtaining Jason-1 / 2 / 3 altimetry satellite data, the original data is pre-processed in two steps: noise removal and spatio-temporal registration. Gaussian filtering is used to remove noise from the obtained data, eliminating abnormal signals caused by ice, ice floes and other environmental factors. To ensure the spatio-temporal consistency of the data, the altimetry data needs to be registered with other remote sensing images (such as high-resolution optical images) to ensure the consistency of spatial coordinates and time reference. This step uses Google Earth high-resolution remote sensing images for registration.

[0016] In further solutions, in step 2), in ice lake water level extraction, the traditional waveform retracking method may be affected by high noise, unstable ice surface reflection, waves and other environmental factors. In this step, an adaptive threshold waveform retracking method (ATR) is introduced. The core idea of this method is to accurately identify the waveform retracking gate position by introducing an adaptive threshold and a time window, and to optimize the accuracy of water level extraction.

[0017] First, the wavelet is extracted from the echo power sequence, and the adaptive difference sequence (ADS) is applied to more accurately identify the leading edge of the wavelet. The calculation formula of the adaptive difference sequence (ADS) is as follows:

[0018]

[0019] Where P(i) and P(j) are the echo intensity values of the i-th and j-th sampling points of the echo power sequence; D ADS (i) is the adaptive difference sequence, which is used to represent the local difference of the i-th point; k is the local window size, which is used to calculate the local difference; j represents the index of all sampling points within the range of i-k to i+k. By introducing the adaptive difference sequence, the rapidly changing part of the wavelet can be accurately captured, and the wavelet leading edge can be identified. When the local difference exceeds a certain threshold, it is considered to be a potential wavelet leading edge.

[0020] After extracting the wavelet, the threshold is determined, which is used to detect the wavelet leading edge. The calculation formula is as follows:

[0021] T ADS (i) = a · σ(i)

[0022] Where a is a constant used to control the sensitivity of the threshold; σ(i) is the standard deviation within the window of the i-th sampling point, which reflects the range of wavelet changes. When the difference sequence D ADS (i) is greater than the threshold T ADS(i) is considered a potential leading edge of the waveform. This method can be used to identify waveform change points in environments with weak signals or strong noise, thereby extracting possible sub-waveform ranges.

[0023] Next, the range of the sub-waveform is further expanded by using a weighting gate. The formula for calculating the weighted power difference is as follows:

[0024]

[0025] Where P(i) is the echo power at the current sampling point, and max(P[i,j]) and min(P[i,j]) are the minimum and maximum echo power within the sub-waveform range, respectively. This formula enhances the signal in the leading edge by standardizing the power difference of the waveform, thereby improving the ability to identify low-signal waveforms.

[0026] Next, an adaptive threshold is calculated to determine if a valid resetting gate position exists in the waveform. The determination of the resetting gate position depends on the fluctuation of the echo signal and environmental noise, as shown in the following formula:

[0027] T r =β·σ window (i)+γ·Noise

[0028] Wherein: T r For adaptive threshold; σ window (i) is the standard deviation of the current window, representing the fluctuation of the echo signal in this area; β and γ are adjustment parameters; Noise is the noise level, representing the influence of environmental noise. When the signal exceeds the threshold, this point is determined as the reset gate position.

[0029] The final distance correction formula corrects for water level measurement errors and yields the final water level. The formula is as follows:

[0030] ΔR=2c·(t GR -t GN )

[0031] H = R - ΔR

[0032] Where: c is the speed of light, t GR t is the time for repositioning the gate. GN The preset tracking gate time is ΔR, the distance correction value is ΔR, the initially measured distance is ΔR, and the final water level is H.

[0033] Further, the multi-source optical satellite data used in step 3) includes data such as Sentinel-2 and Landsat series. Sentinel-2 data is provided by the European Space Agency (ESA) and can be accessed through the Copernicus Open Access Platform, with a spatial resolution of 10 m; Landsat series is provided by the United States Geological Survey (USGS) and downloaded through the EarthExplorer platform, with a spatial resolution of 30 m.

[0034] To ensure image quality, standardization preprocessing was carried out on the acquired remote sensing data. Atmospheric correction used Sen2Cor (for Sentinel-2) and LEDAPS (for Landsat series) methods based on radiation transfer model to eliminate the interference of atmospheric aerosol, water vapor and other factors on surface reflectance; Geometric correction was then performed with Google Earth high-resolution remote sensing image as reference for accurate registration to ensure the consistency of multi-source data in spatial positioning.

[0035] Further, the calculation method in step 4) is shown in the following formula:

[0036]

[0037]

[0038] AWEI sh =4(Green-SWIR1) / (0.25NIR+2.75SWIR2)

[0039] AWEI nsh =Blue+2.5Green-1.5(NIR+SWIR1)-0.25SWIR2

[0040] Where: Green represents the green band (520-600 nm) of the multi-spectral image, NIR represents the near-infrared band (760-900 nm), SWIR represents the short-wave infrared band (1550-1750 nm), which contains two bands SWIR1 and SWIR2, with wavelength ranges of about 1550-1750 nm and 2080-2350 nm, respectively.

[0041] Further, in step 5), random forest is an ensemble learning method that builds multiple decision trees for classification and regression. Each tree is trained by randomly selecting a subset of data and a subset of features, and the final classification result is determined by a voting mechanism. This method can effectively handle high-dimensional data, reduce overfitting, and has strong robustness.

[0042] In ice lake contour extraction, random forests can integrate four water body indices (NDWI, MNDWI, AWEI). sh AWEI nsh This information accurately identifies glacial lake boundaries, adapts to different environmental changes, and provides high-precision classification results. The prediction of each decision tree can be expressed as a function based on four water body index features, as shown below.

[0043]

[0044] Where: X j =[NDWI, MNDWI, AWEI sh AWEI nsh ] represents the four water body index features input, R i It is the i-th leaf node region, I(X) j ∈R i ) is an indicator function, representing the eigenvector X. j Whether it belongs to the i-th leaf node, y i This is the class label for the leaf nodes. This formula indicates that each tree is based on four water quality indices (NDWI, MNDWI, AWEI). sh AWEI nsh The feature space of the samples is divided using the input feature X, and the samples are classified into corresponding leaf node regions for classification prediction. The output of each tree is based on the input feature X. j The classification results.

[0045] Finally, the overall classification result of the random forest is determined by voting among multiple decision trees:

[0046] y RF =model(f1(X),f2(X),...,f N (X))

[0047] Where: y RF The classification result is represented by `model(...)`, which is the trained random forest classification model; f i (X) represents the classification result for the i-th class. The final classification of the ice lake's outline is determined by voting from all the trees.

[0048] In a further scheme, in step 6), the water level data are obtained from step 2), and the contour change is obtained from step 5), and the contour change generally represents the water area change, so that the water level-area relationship can be established, and first, a suitable time period is selected, and the time difference between the optical image and the altimetric satellite water level data is ensured to be as small as possible (generally controlled within 10 days), the extracted local contour change is reflected to the water area, and is matched with the altimetric satellite water level data in the corresponding time period. Since the altimetric satellite data provide the water surface height, and the lake water level change is generally uniform, the contour change and the water level change present a direct relationship. Through regression analysis, the mathematical relationship between the area change and the water level is established. Generally, the relationship presents a linear characteristic, that is, with the increase of the water level, the water area changes linearly. Through the linear regression model, the regression equation most suitable for the data can be found, so as to convert the area change into the corresponding water level value.

[0049] In the construction of the area-water volume relationship, it is generally assumed that the lake is conical or circular table-shaped, so that the water level and the area reflected by the contour generally present a quadratic function relationship. Through regression analysis, the mathematical relationship between the water level and the area can be derived from the actually measured lake area and water level data. Further, by using the relationship, the relationship between the water level and the lake water volume is obtained through integration calculation.

[0050] Advantages of the present application:

[0051] The present application takes the contour, water level and water storage capacity and other morphological information of the ice lake as the research object, and provides a satellite remote sensing based ice lake morphological information monitoring method for a lack of data area based on multi-source remote sensing satellites, altimetric satellites and the like, combined with machine learning and waveform redefinition, so as to realize remote sensing monitoring of the ice lake morphological information in a lack of data area, and is suitable for ice lake dynamic monitoring, risk analysis and the like.

[0052] The present application realizes multi-temporal and systematic monitoring of the ice lake by using the multi-element collaborative monitoring method, so as to improve the precision and practicability of the ice lake morphological information extraction. BRIEF DESCRIPTION OF DRAWINGS

[0053] The present application will be further described below in combination with the drawings and embodiments.

[0054] Figure 1 FIG. 1 is a technical flowchart of the satellite remote sensing based ice lake morphological information monitoring method for a lack of data area according to the embodiment 1.

[0055] Figure 2 FIG. 4 is the water level extraction precision verification result of the Selinco lake in the Selinco river basin in the embodiment 1.

[0056] Figure 3 FIG. 6 is the water body index calculation result of the typical lake in the Selinco river basin in the embodiment 1.

[0057] Figure 4 The profile extraction results of typical lakes in the Selin Co basin in Example 1.

[0058] Figure 5 The water level-area-water volume relationship of typical lakes in the Selin Co basin in Example 1. DETAILED DESCRIPTION

[0059] Example 1:

[0060] A satellite remote sensing-based ice lake shape information monitoring method for a lack of data area, the monitoring area in the present embodiment is the Selin Co basin, and the following are the specific steps of Example 1:

[0061] 1) Get and preprocess altimetry satellite data: altimetry satellite data preprocessing includes denoising and spatio-temporal registration,

[0062] In the present embodiment, after obtaining the Jason-1 / 2 / 3 altimetry satellite data of the Selin Co basin from 2010 to 2024 through the NASA's PO.DAAC (Physical Oceanographic Data Assimilation and Analysis Center), the original data is preprocessed by denoising and spatio-temporal registration.

[0063] After obtaining the Jason-1 / 2 / 3 altimetry satellite data, the original data is preprocessed by denoising and spatio-temporal registration. The obtained data is denoised by using Gaussian filtering to eliminate abnormal signals caused by ice, floating ice and other environmental factors. In order to ensure the spatio-temporal consistency of the data, the altimetry data needs to be registered with other remote sensing images (such as high-resolution optical images) to ensure the consistency of spatial coordinates and time reference. In this step, Google Earth high-resolution remote sensing images are used for registration.

[0064] 2) Ice lake water level extraction based on the altimetry satellite data preprocessed by the adaptive threshold waveform redefinition method ATR: the leading edge of the waveform is identified by using the adaptive difference sequence, and the threshold for detecting the leading edge of the waveform is determined; the range of the sub-waveform is further expanded by using the weighted redefinition gate; the adaptive threshold is calculated to determine whether there is an effective redefinition gate position in the waveform; the distance correction value is obtained based on the time of the redefinition gate position to correct the water level measurement formula, and the final water level is obtained;

[0065] In the present embodiment, the water level data is extracted based on the adaptive threshold waveform redefinition method (ATR) for the results of step 1). Taking the Selin Co water level data observed by the Jason-3 altimetry satellite as an example, through the actual water level verification of the Selin Co in 2021, 2022 and 2023, after removing the relative error between the altimetry satellite and the measured water level, the water level accuracy of the Jason-3 altimetry satellite can reach 7 cm (RMSE), and the R 2 reaches 0.91.

[0066] 3) Multi-source optical satellite data acquisition and preprocessing: The preprocessing of multi-source optical satellite data includes atmospheric correction and geometric correction.

[0067] In the embodiment, the high-resolution optical satellite data used includes data such as Sentinel-2 and Landsat series. Sentinel-2 data is obtained through the Copernicus Open Access Platform, and Landsat series is obtained through the EarthExplorer platform.

[0068] The downloaded Sentinel-2 Level-1C image is first subjected to atmospheric correction using the Sen2Cor tool to convert it into a surface reflectance product (Level-2A). Sen2Cor is based on a physical radiation transfer model and can effectively remove the interference of atmospheric aerosols and water vapor, improving the accuracy of spectral indicators such as the CI index. Landsat series images are converted into surface reflectance products (Level-2) through the LaSRC atmospheric correction method, effectively removing atmospheric interference and improving the accuracy of spectral indices. Both types of optical images use Google Earth high-resolution remote sensing images as a reference for geometric correction.

[0069] 4) Water index calculation: Four water indices are calculated, including the normalized difference water index (NDWI), the modified normalized difference water index (MNDWI), the water automatic extraction index AWEI, which is divided into shadow formula (AWEI sh ) and non-shadow formula (AWEI nsh ) two kinds. The calculation formula is as follows:

[0070]

[0071] AWEI sh = 4(Green-SWIR1) / (0.25NIR+2.75SWIR2)

[0072] AWEI nsh = Blue+2.5Green-1.5(NIR+SWIR1)-0.25SWIR2

[0073] Green represents the green band (520-600 nm) of the multispectral image, corresponding to the B3 band of Sentinel-2 and the B3 band of the Landsat series of satellites (taking Landsat 8 as an example); NIR represents the near-infrared band (760-900 nm), corresponding to the B8 band of Sentinel-2 and the B5 band of the Landsat series of satellites (taking Landsat 8 as an example); SWIR represents the short-wave infrared band (1550-1750 nm), including two bands SWIR1 and SWIR2 corresponding to the B11 and B12 bands of Sentinel-2 and the B6 and B7 bands of the Landsat series of satellites (taking Landsat 8 as an example), and the extraction results are shown in the following Figure 3 .

[0074] 5) Ice lake contour extraction based on machine learning method: With the help of random forest classification method, four water body indexes (NDWI, MNDWI, AWEI sh , AWEI nsh ) information as input, the ice lake contour is accurately identified.

[0075] With four water body indexes as input, more than 400 sampling points were uniformly arranged in Selin Co Basin, and 140,000 training samples were collected from Landsat 8 / 9 and Sentinel-2 images in 2021 and 2022. The size of the random forest decision tree is 50, the input information is AWEI sh , AWEI nsh , MNDWI and NDWI, and the target is to divide each pixel into two categories of water body and non-water body, and the water body pixel also includes frozen lake surface. The classification accuracy (test set) of the trained model can reach 0.97, and the Kappa coefficient can reach 0.93, which can meet the needs of extracting the range of lake water surface and the range of shoreline change, and the typical ice lake contour extraction is shown in the following Figure 4 .

[0076] 6) Construction of water level-area-water volume relationship

[0077] The water level-area-water volume relationship of 11 main lakes in Selin Co Basin is shown in the following Figure 5 . Among the 11 main lakes in Selin Co Basin, two main lake inflow systems are formed, the main one is through the Zhagen Zangbo River system connected lakes, namely Chazang Co-Yege Co-Mudiladuo Co-Gerencuo Co-Jiaxia Zangbo River-Ziguicuo Co-Sirong Zangbo River-Wuru Co-Qaigui Co-Zhagen Zangbo River-Selin Co, and the second one is through the Ali Zangbo River connected lakes, namely Mugecuo Co-Yongzhu Zangbo River-Cuo'e-Ali Zangbo River-Selin Co. This embodiment focuses on extracting the lake water regime elements of Gerencuo, Ziguicuo, Wuru, Qaigui, Mugecuo and Cuo'e.

[0078] It should be pointed out finally that the above merely serves to illustrate the technical solutions of the present application but not to limit the present application. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or equivalently replaced without departing from the spirit and scope of the present application.

Claims

1. A method for monitoring ice lake shape information in an area with insufficient data based on satellite remote sensing, characterized in that: Comprising the following steps: 1) Altimeter satellite data acquisition and preprocessing: altimeter satellite data preprocessing includes denoising and space-time registration; 2) Ice lake water level extraction based on adaptive threshold waveform redefinition method ATR of pretreated altimeter satellite data: the leading edge of the waveform is identified by applying adaptive difference sequence, and the threshold for detecting the leading edge of the waveform is determined; the range of the sub-waveform is further expanded by weighted redefinition gate; the adaptive threshold is calculated to determine whether there is an effective redefinition gate position in the waveform; the distance correction value is obtained based on the time of the redefinition gate position to correct the water level measurement formula, and the final water level is obtained; the adaptive difference sequence ADS calculation formula is as follows: wherein: P(i) and P(j) are echo intensity values of the i-th and j-th sampling points of the echo power sequence; D ADS (i) is an adaptive difference sequence, which is used to represent the local difference of the i-th point; k is the size of the local window, which is used to calculate the local difference, and j represents the index of all sampling points within the range from i-k to i+k; the range of the sub-waveform is further expanded by weighting the gate, wherein the calculation formula of the weighted power difference used is as follows: Wherein: P(i) is the echo power of the current sampling point, max(P[i,j]) and min(P[i,j]) are the minimum and maximum echo power in the sub-waveform range respectively; The adaptive threshold is calculated to determine whether there is an effective redefinition gate position in the waveform, and the determination of the redefinition gate position depends on the fluctuation of the echo signal and the environmental noise, and the formula is as follows: T r = β · σ window (i) + γ · Noise Wherein: T r is an adaptive threshold; σ window (i) is the standard deviation of the current window, indicating the fluctuation of the echo signal in the window area; β and γ are adjustment parameters; Noise is the noise level, indicating the influence of environmental noise; when the signal exceeds the threshold, the sampling point is determined as the re-gating position; 3) Multi-source optical satellite data acquisition and preprocessing: The preprocessing of multi-source optical satellite data includes atmospheric correction and geometric correction; 4) Water body index calculation based on pre-processed multi-source optical satellite data: Calculated water body indexes include normalized difference water index NDWI, improved normalized difference water index MNDWI and water body automatic extraction index AWEI, which is divided into shadow formula AWEI sh and non-shadow formula AWEI nsh two; 5) Ice lake contour extraction based on machine learning method: Random forest classification method is adopted, and four water body indexes NDWI, MNDWI, AWEI sh , AWEI nsh are taken as input to accurately identify the ice lake contour; 6) Water level-area-water volume relationship construction: the water level data is obtained from step 2), and the contour change is obtained from step 5), first establish the water level-area relationship, assuming that the lake is conical or circular table shape, construct the area-water volume relationship, and finally obtain the water level-area-water volume relationship.

2. The satellite remote sensing based method for monitoring ice lake morphological information in an area with insufficient data according to claim 1, characterized in that: In step 1), the acquired altimeter satellite data is denoised by using Gaussian filter; and the satellite altimeter data is space-time registered with Google Earth high-resolution remote sensing image.

3. The method for monitoring glacial lake morphology information in data-scarce areas based on satellite remote sensing according to claim 1, characterized in that: The atmospheric correction in step 3) adopts Sen2Cor and / or LEDAPS method based on radiation transfer model; the geometric correction is accurately registered with Google Earth high-resolution remote sensing image as reference, to ensure the consistency of multi-source data in spatial positioning.

4. The method according to claim 1, wherein the method comprises the following steps: 1) obtaining the satellite remote sensing data of the ice lake in the area with insufficient data; 2) determining the ice lake shape information of the ice lake in the area with insufficient data according to the satellite remote sensing data of the ice lake in the area with insufficient data. The calculation formula of each water body index in step 5) is as follows: AWEI sh = 4(Green-SWIR1) / (0.25NIR+2.75SWIR2) AWEI nsh = Blue + 2.5 Green - 1.5 (NIR + SWIR1) - 0.25 SWIR2 Wherein: Green represents the green band of multispectral image, NIR represents near-infrared band, Blue represents blue band, SWIR represents short-wave infrared band, which includes two bands SWIR1 and SWIR2, and the wavelength ranges are 1550-1750 nm and 2080-2350 nm respectively.

5. The method for monitoring glacial lake morphology information in data-scarce areas based on satellite remote sensing according to claim 1, characterized in that: The prediction expression of each decision tree in the random forest classification method in step 5) is a function based on four water body index characteristics, and the specific formula is as follows: where: X j = [NDWI, MNDWI, AWEI sh , AWEI nsh ] is the input four water body index features, R i is the ith leaf node region, I(X j ∈ R i ) is an indicator function, indicating whether the feature vector X j belongs to the ith leaf node, y i is the class label of the leaf node.

Images