Ice surface melt water depth inversion method and system

Abstract

The invention discloses an ice surface melt depth inversion method and system, and the method comprises the steps: obtaining an ICESat-2ATL03 photon data set, and carrying out the preliminary denoising of the ICESat-2ATL03 photon data set, and obtaining target photon data; carrying out water melting surface and water melting bottom photon separation to obtain water melting surface photons and water melting bottom photons; carrying out de-noising processing on the photons on the melt water surface and carrying out ice surface melt water retrieval; de-noising processing is carried out on the melt water bottom photons, and ice surface melt water retrieval is carried out; carrying out shallow water melting and snow mud removal treatment on the ice surface water melting window; multispectral image data are obtained, machine learning model training data are generated in combination with the melt depth, melt wave band reflectivity is input based on the multispectral image data, and a final depth result is generated. According to the embodiment of the invention, the ice surface melt depth can be inversed in a large area with high time resolution and high precision, the ice surface melt depth information with good space coverage and high precision can be obtained, and the method can be widely applied to the technical field of remote sensing geoscience application.

Description

Technical Field

[0001] The invention relates to the technical field of remote sensing geoscience application, and in particular to an ice surface meltwater depth inversion method and system. Background Art

[0002] Meltwater is an important phenomenon on the ice sheet. It mostly occurs in summer and plays a vital role in the material balance of the ice sheet surface. Understanding the volume and spatial distribution of meltwater is of great significance for studying the material loss of the ice sheet and the surface material balance. The basis for inversion of meltwater volume is to be able to estimate the depth of meltwater over a large area with high accuracy.

[0003] In addition to field measurements, related studies on the Greenland ice meltwater depth detection methods also include meltwater depth estimation based on satellite multispectral images and lidar data. Although field measured data has the highest credibility and accuracy, it is currently relatively scarce due to cost, environmental and other factors. In addition, field personnel generally only conduct small-scale bathymetry of meltwater on a certain ice surface in Greenland, and it is difficult to carry out meltwater depth monitoring on multiple ice surfaces over a large area. Therefore, field measured data is generally only used for method verification. Multispectral images themselves have a high temporal resolution and are convenient for large-scale meltwater depth monitoring, but the current image-based methods are still greatly limited in accuracy. Meltwater depth estimation based on lidar data has obvious advantages in terms of accuracy, but the shortcomings of the data itself, such as low temporal resolution and limited detection of ice meltwater depth, still limit its use in ice meltwater depth measurement. Summary of the invention

[0004] The main purpose of the embodiments of the present invention is to propose an ice surface meltwater depth inversion method and system, which can invert the ice surface meltwater depth over a large area, with high temporal resolution and high precision, and can obtain ice surface meltwater depth information with good spatial coverage and high precision.

[0005] To achieve the above object, an embodiment of the present invention provides a method for inverting ice meltwater depth, comprising the following steps:

[0006] Obtain the ICESat-2 ATL03 photon dataset and perform preliminary denoising on it to obtain the target photon data;

[0007] The target photon data after preliminary denoising are subjected to meltwater surface and meltwater bottom photon separation to obtain meltwater surface photons and meltwater bottom photons;

[0008] De-noising the meltwater surface photons and performing a first ice surface meltwater retrieval;

[0009] De-noising the meltwater bottom photons and performing a second ice surface meltwater search;

[0010] According to the result of the first ice surface meltwater search and the result of the second ice surface meltwater search, the ice surface meltwater window is processed to remove shallow meltwater and slush;

[0011] Acquire multispectral image data and generate machine learning model training data in combination with meltwater depth. Input the meltwater band reflectance into the trained machine learning model based on the multispectral image data to generate the final depth result.

[0012] In some embodiments, the step of acquiring an ICESat-2 ATL03 photon data set and performing preliminary denoising on the data to obtain target photon data comprises the following steps:

[0013] Download the ICESat-2 ATL03 photon dataset for the target location;

[0014] The collected photon data set is subjected to preliminary denoising using the DRAGANN algorithm.

[0015] In some embodiments, the step of performing meltwater surface and meltwater bottom photon separation on the target photon data after preliminary denoising to obtain meltwater surface photons and meltwater bottom photons includes the following steps:

[0016] The photons after preliminary denoising are divided into 40m processing windows along the track with a step size of 20m;

[0017] Within each processing window, photon kernel density estimation is performed along the vertical height; the calculation formula for the photon kernel density estimate is: in, is the probability density estimate at the photon point x, n is the number of samples, h is the smoothing parameter, called bandwidth, which controls the width of the kernel function; K is the kernel function; the expression of the kernel function K is: Among them, u represents is the data point x i The normalized distance from point x;

[0018] Find all points whose corresponding photon kernel density estimation values ​​are the minimum values ​​in the photon kernel density estimation curve, take their elevation values, and use the maximum value among these elevation values ​​as the threshold;

[0019] Among them, photons with elevation less than the threshold are considered to be meltwater bottom photons, and photons with elevation greater than the threshold are considered to be meltwater surface photons.

[0020] In some embodiments, the denoising of the meltwater surface photons and the first ice surface meltwater retrieval include the following steps:

[0021] The 3δ iteration method is used to remove the noise photons from the meltwater surface photons;

[0022] When the standard deviation of the photon elevation of the meltwater surface within the processing window is less than 0.12 m, the window is considered to be a meltwater surface window;

[0023] If the distance between two meltwater windows is within 40 meters, group them together;

[0024] Each group of meltwater surface windows is filled, and the meltwater surface window group containing more than 5 windows is selected as the result of the first retrieval of ice surface meltwater.

[0025] In some embodiments, the denoising of the meltwater bottom photons and the second ice surface meltwater retrieval include the following steps:

[0026] First, a Savitzky-Gol ay filter is used to generate a fitting curve for the bottom photons, and then the fitting value is subtracted from the height of the photons to eliminate the trend of the bottom photons; where the expression of the Savitzky-Gol ay filter is: Among them, y i ′ is the elevation value after filtering; y i+j is the elevation value in the original photon; c j is the weight coefficient of the filter and m is the number of points on both sides of the window;

[0027] All bottom photons within 1 m below the meltwater surface elevation were removed to avoid the influence of multi-surface reflection on the final extracted meltwater depth;

[0028] All photons below the median elevation of the bottom photons were removed in each meltwater surface window. Then, 3δ iterations were performed to remove noise photons in the vertical direction and denoised by an elliptical filter kernel in the horizontal direction.

[0029] If the number of bottom photons in a meltwater surface window is greater than 10, the window is identified as an ice surface meltwater window;

[0030] These ice surface meltwater windows are grouped according to the same criteria as the meltwater surface windows obtained in the first ice surface meltwater retrieval.

[0031] In some embodiments, the step of removing shallow meltwater and slush from the ice surface meltwater window according to the result of the first ice surface meltwater search and the result of the second ice surface meltwater search comprises the following steps:

[0032] The photons whose first elevation value on both sides of each group of ice surface meltwater windows exceeds the range of three times the standard deviation of the surface elevation value are retained and regarded as the edge of ice surface meltwater;

[0033] Connect the edge of the meltwater bottom to the meltwater surface and generate the final meltwater bottom through Savitzky-Gol ay filtering again;

[0034] A refractive index of ~1.33 was used for the height of meltwater bottom photons to eliminate the refraction effect of meltwater;

[0035] Remove meltwater ponds with an average depth of less than 1.15m.

[0036] In some embodiments, the method of acquiring multispectral image data and generating machine learning model training data in combination with meltwater depth, inputting meltwater band reflectance into the trained machine learning model based on the multispectral image data, and generating a final depth result includes the following steps:

[0037] Obtain multispectral images that correspond to ATL03 images in time and space, including Landsat8 L1 / L2 and Sentine l2L1C / L2A, a total of four images. The time difference between these images and ATL03 data is no more than 4 days;

[0038] The median value of the depth sampling point within an image pixel is used to generate training data with its multi-band reflectance. For Landsat-8, the median depth is paired with the reflectance value of 7 bands. For Sentine l-2, the low-resolution bands are first uniformly sampled to 10m, and the median depth is paired with the reflectance value of 9 bands.

[0039] Manually debug the training parameters of the two machine learning algorithms, XGBoost and LightGBM, input the training data, and derive the meltwater depth inversion machine learning model. Then, based on the multispectral image, the meltwater band reflectivity is input into the model to generate the final depth result.

[0040] Another aspect of the embodiment of the present invention further provides an ice surface meltwater depth inversion system, comprising:

[0041] The first module is used to obtain the ICESat-2 ATL03 photon data set and perform preliminary denoising on it to obtain the target photon data;

[0042] The second module is used to separate the meltwater surface and meltwater bottom photons from the target photon data after preliminary denoising to obtain meltwater surface photons and meltwater bottom photons;

[0043] The third module is used for denoising the meltwater surface photons and performing the first ice surface meltwater retrieval;

[0044] The fourth module is used to perform denoising on the meltwater bottom photons and perform a second ice surface meltwater retrieval;

[0045] A fifth module is used to remove shallow meltwater and slush from the ice surface meltwater window according to the result of the first ice surface meltwater search and the result of the second ice surface meltwater search;

[0046] The sixth module is used to obtain multispectral image data and generate machine learning model training data in combination with meltwater depth. The meltwater band reflectance is input into the trained machine learning model based on the multispectral image data to generate the final depth result.

[0047] To achieve the above objective, another aspect of an embodiment of the present invention provides an electronic device, the electronic device comprising a memory and a processor, the memory storing a computer program, and the processor implementing the above-mentioned method when executing the computer program.

[0048] To achieve the above objective, another aspect of an embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the above-mentioned method is implemented.

[0049] The embodiment of the present invention also discloses a computer program product or a computer program, which includes a computer instruction stored in a computer-readable storage medium. A processor of a computer device can read the computer instruction from the computer-readable storage medium, and the processor executes the computer instruction, so that the computer device executes the above method.

[0050] The embodiments of the present invention include at least the following beneficial effects: the present invention provides an ice surface meltwater depth inversion method and system, the scheme obtains the ICESat-2 ATL03 photon data set, and performs preliminary denoising on it to obtain target photon data; the meltwater surface and meltwater bottom photons are separated for the target photon data after preliminary denoising to obtain meltwater surface photons and meltwater bottom photons; the meltwater surface photons are denoised and the first ice surface meltwater search is performed; the meltwater bottom photons are denoised and the second ice surface meltwater search is performed; according to the results of the first ice surface meltwater search and the second ice surface meltwater search, the ice surface meltwater window is subjected to shallow meltwater and slush removal processing; multispectral image data is obtained, and machine learning model training data is generated in combination with meltwater depth, and the meltwater band reflectivity is input into the trained machine learning model based on the multispectral image data to generate the final depth result. The embodiments of the present invention can invert the ice surface meltwater depth over a large area, with high temporal resolution and high precision, and can obtain ice surface meltwater depth information with good spatial coverage and high precision. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 It is a schematic diagram of an implementation environment provided by an embodiment of the present invention;

[0052] Figure 2 is a flow chart of the overall steps provided by an embodiment of the present invention;

[0053] Figure 3 It is an implementation process diagram provided by an embodiment of the present invention in a specific application scenario;

[0054] Figure 4 is a schematic diagram of the results of estimating meltwater depth using Landsat8 L1 data using different methods provided in an embodiment of the present invention, and performing a one-dimensional comparison with refraction-corrected ICESat-2 meltwater depth data;

[0055] Figure 5 It is a schematic diagram of two-dimensional comparison between the depth estimated by different methods provided in the embodiment of the present invention on the test ice lake 2 and the Arctic DEM data;

[0056] Figure 6 It is a schematic diagram of the hardware structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0057] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the embodiments of the present invention. They are only examples of devices and methods consistent with some aspects of the embodiments of the present invention as detailed in the attached claims.

[0058] It is understood that the terms "first", "second", "third", "fourth", etc. (if any) in the description of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0059] It should be understood that in the present invention, "at least one (item)" refers to one or more, and "plurality" refers to two or more. "And / or" is used to describe the association relationship of associated objects, indicating that three relationships may exist. For example, "A and / or B" can represent: only A exists, only B exists, and A and B exist at the same time, where A and B can be singular or plural. The character " / " generally indicates that the previous and next associated objects are in an "or" relationship. "At least one of the following items" or similar expressions refers to any combination of these items, including any combination of single items or plural items. For example, at least one of a, b or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, c can be single or multiple.

[0060] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which the present invention belongs. The terms used herein are only for the purpose of describing the embodiments of the present invention and are not intended to limit the present invention.

[0061] The ice meltwater depth inversion method and system provided in the embodiments of the present invention relate to the field of remote sensing geoscience applications. The ice meltwater depth inversion method provided in the embodiments of the present invention can be applied to a terminal, can be applied to a server, and can also be software running in a terminal or a server. In some embodiments, the terminal can be a smart phone, a tablet computer, a laptop computer, a desktop computer, a smart speaker, a smart watch, and a car terminal, etc., but is not limited to this; the server side can be configured as an independent physical server, or a server cluster or distributed system composed of multiple physical servers, and can also be configured as a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms. The server can also be a node server in a blockchain network; the software can be an application that implements the ice meltwater depth inversion method, etc., but is not limited to the above forms.

[0062] The present invention can be used in many general or special computer system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronic devices, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, etc. The present invention can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. The present invention can also be practiced in distributed computing environments, in which tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules can be located in local and remote computer storage media including storage devices.

[0063] like Figure 1 FIG. 1 is a schematic diagram of an implementation environment provided by an embodiment of the present invention. Figure 1 The implementation environment includes at least one terminal 102 and a server 101. The terminal 102 and the server 101 can be connected to a network wirelessly or wired to complete data transmission and exchange.

[0064] Server 101 can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers. It can also be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN (Content Delivery Network), as well as big data and artificial intelligence platforms.

[0065] In addition, the server 101 can also be a node server in the blockchain network. Blockchain is a new application model of computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanism, and encryption algorithm.

[0066] The terminal 102 may be a smart phone, a tablet computer, a laptop computer, a desktop computer, a smart speaker, a smart watch, etc. The terminal 102 may also be a vehicle-mounted terminal of various device types described above, but is not limited thereto. The terminal 102 and the server 101 may be directly or indirectly connected via wired or wireless communication, which is not limited in this embodiment of the present invention.

[0067] Based on the example Figure 1In the implementation environment shown, an embodiment of the present invention provides an ice surface meltwater depth inversion method. The following is explained using the ice surface meltwater depth inversion method applied to the server 101 as an example. It can be understood that the method can also be applied to the terminal 102.

[0068] Reference Figure 2 , Figure 2 The flowchart of the ice surface meltwater depth inversion method applied to the server provided in the embodiment of the present invention, the execution subject of the method can be any of the aforementioned computer devices (including servers or terminals). Figure 2 , the method may include the following steps:

[0069] Obtain the ICESat-2 ATL03 photon dataset and perform preliminary denoising on it to obtain the target photon data;

[0070] The target photon data after preliminary denoising are subjected to meltwater surface and meltwater bottom photon separation to obtain meltwater surface photons and meltwater bottom photons;

[0071] De-noising the meltwater surface photons and performing a first ice surface meltwater retrieval;

[0072] De-noising the meltwater bottom photons and performing a second ice surface meltwater search;

[0073] According to the result of the first ice surface meltwater search and the result of the second ice surface meltwater search, the ice surface meltwater window is processed to remove shallow meltwater and slush;

[0074] Acquire multispectral image data and generate machine learning model training data in combination with meltwater depth. Input the meltwater band reflectance into the trained machine learning model based on the multispectral image data to generate the final depth result.

[0075] In some embodiments, the step of acquiring the ICESat-2 ATL03 photon data set and performing preliminary denoising on the data to obtain target photon data comprises the following steps:

[0076] Download the ICESat-2 ATL03 photon dataset for the target location;

[0077] The collected photon data set is subjected to preliminary denoising using the DRAGANN algorithm.

[0078] In some embodiments, the step of performing meltwater surface and meltwater bottom photon separation on the target photon data after preliminary denoising to obtain meltwater surface photons and meltwater bottom photons includes the following steps:

[0079] The photons after preliminary denoising are divided into 40m processing windows along the track with a step size of 20m;

[0080] Within each processing window, photon kernel density estimation is performed along the vertical height; the calculation formula for the photon kernel density estimate is: in, is the probability density estimate at the photon point x, n is the number of samples, h is the smoothing parameter, called bandwidth, which controls the width of the kernel function; K is the kernel function; the expression of the kernel function K is: Among them, u represents is the data point x i The normalized distance from point x;

[0081] Find all points whose corresponding photon kernel density estimation values ​​are the minimum values ​​in the photon kernel density estimation curve, take their elevation values, and use the maximum value among these elevation values ​​as the threshold;

[0082] Among them, photons with elevation less than the threshold are considered to be meltwater bottom photons, and photons with elevation greater than the threshold are considered to be meltwater surface photons.

[0083] In some embodiments, the denoising of the meltwater surface photons and the first ice surface meltwater retrieval include the following steps:

[0084] The 3δ iteration method is used to remove the noise photons from the meltwater surface photons;

[0085] When the standard deviation of the photon elevation of the meltwater surface within the processing window is less than 0.12 m, the window is considered to be a meltwater surface window;

[0086] If the distance between two meltwater windows is within 40 meters, group them together;

[0087] Each group of meltwater surface windows is filled, and the meltwater surface window group containing more than 5 windows is selected as the result of the first retrieval of ice surface meltwater.

[0088] In some embodiments, the denoising of the meltwater bottom photons and the second ice surface meltwater retrieval include the following steps:

[0089] First, a Savitzky-Gol ay filter is used to generate a fitting curve for the bottom photons, and then the fitting value is subtracted from the height of the photons to eliminate the trend of the bottom photons; where the expression of the Savitzky-Gol ay filter is: Among them, y i ′ is the elevation value after filtering; y i+j is the elevation value in the original photon; c j is the weight coefficient of the filter and m is the number of points on both sides of the window;

[0090] All bottom photons within 1 m below the meltwater surface elevation were removed to avoid the influence of multi-surface reflection on the final extracted meltwater depth;

[0091] All photons below the median elevation of the bottom photons were removed in each meltwater surface window. Then, 3δ iterations were performed to remove noise photons in the vertical direction and denoised by an elliptical filter kernel in the horizontal direction.

[0092] If the number of bottom photons in a meltwater surface window is greater than 10, the window is identified as an ice surface meltwater window;

[0093] These ice surface meltwater windows are grouped according to the same criteria as the meltwater surface windows obtained in the first ice surface meltwater retrieval.

[0094] In some embodiments, the step of removing shallow meltwater and slush from the ice surface meltwater window according to the result of the first ice surface meltwater search and the result of the second ice surface meltwater search comprises the following steps:

[0095] The photons whose first elevation value on both sides of each group of ice surface meltwater windows exceeds the range of three times the standard deviation of the surface elevation value are retained and regarded as the edge of ice surface meltwater;

[0096] Connect the meltwater bottom edge to the meltwater surface and generate the final meltwater bottom through Savitzky-Golay filtering again;

[0097] A refractive index of ~1.33 was used for the height of meltwater bottom photons to eliminate the refraction effect of meltwater;

[0098] Remove meltwater ponds with an average depth of less than 1.15m.

[0099] In some embodiments, the method of acquiring multispectral image data and generating machine learning model training data in combination with meltwater depth, inputting meltwater band reflectance into the trained machine learning model based on the multispectral image data, and generating a final depth result includes the following steps:

[0100] Obtain multispectral images that correspond to ATL03 images in time and space, including Landsat8 L1 / L2 and Sentinel2L1C / L2A, a total of four images. The time difference between these images and ATL03 data is no more than 4 days;

[0101] The median value of the depth sampling point within an image pixel is used to generate training data with its multi-band reflectance. For Landsat-8, the median depth is paired with the reflectance value of 7 bands. For Sentinel-2, the low-resolution bands are first uniformly sampled to 10m, and the median depth is paired with the reflectance value of 9 bands.

[0102] Manually debug the training parameters of the two machine learning algorithms, XGBoost and LightGBM, input the training data, and derive the meltwater depth inversion machine learning model. Then, based on the multispectral image, the meltwater band reflectivity is input into the model to generate the final depth result.

[0103] The following describes the specific implementation process of the embodiment of the present invention in detail by taking a specific application scenario as an example:

[0104] In order to monitor the volume and spatial distribution of meltwater on the ice surface, the present invention combines ICESat-2 ATL03 photon data and multispectral images through a machine learning model, and proposes a method that can invert the depth of meltwater on the ice surface over a large area, with high temporal resolution and high precision.

[0105] The purpose of the present invention is to provide a method for inverting the ice meltwater depth over a large area, with high temporal resolution and high precision. With the support of machine learning technology, ice meltwater depth information with good spatial coverage and high precision can be obtained from multispectral image data such as Landsat8 and Sentine l2.

[0106] Specifically, the ice meltwater depth inversion method based on ICESat-2 and multispectral image data sets of the present invention is as follows: Figure 3 As shown, the following steps may be included:

[0107] Step 1: Obtain a total of 15 ICESat-2 ATL03 photon datasets, which are spatially concentrated in the densely melted area on the southwestern Greenland Ice Sheet in the summer of 2019-2022 (June-August); Use the DRAGANN algorithm to perform preliminary denoising on the acquired ATL03 photon dataset to facilitate subsequent ice surface meltwater retrieval and meltwater depth extraction.

[0108] Step 2: Divide the photons after preliminary denoising into 40m processing windows along the track with a step size of 20m. In each processing window, perform photon kernel density estimation (KDE) along the vertical height with a bin width of 0.1m. Then, we find all points whose corresponding KDE values ​​are the minimum values ​​in the KDE curve and take their elevation values. We use the maximum value of these elevation values ​​as the threshold. Photons with elevations less than the threshold are considered to be meltwater bottom photons, and photons with elevations greater than the threshold are considered to be meltwater surface photons. The corresponding formula for photon KDE is as follows:

[0109]

[0110] in, is the probability density estimate at the photon point x, n is the number of samples, h is the smoothing parameter, called bandwidth, which controls the width of the kernel function. K is the kernel function, and the kernel function in this experiment is a Gaussian kernel, with the following formula:

[0111]

[0112] Step 3: Use the 3δ iteration method to remove noise photons from meltwater surface photons. When the standard deviation of the surface photon elevation in the processing window is less than 0.12m, the window is considered to be a meltwater surface window. If the distance between two meltwater surface windows is within 40 meters, they are grouped together. Subsequently, fill each group of meltwater surface windows and select the meltwater surface window group containing more than 5 windows as the result of the first retrieval of ice surface meltwater.

[0113] Step 4: First, use the Savitzky-Go l ay filter to generate a fitting curve for the bottom photons, and then subtract the fitting value from the photon height to eliminate the trend of the bottom photons. Subsequently, we remove all bottom photons within 1 m below the meltwater surface elevation to avoid the influence of multi-surface reflection on the final extracted meltwater depth. All photons below the median elevation of the bottom photons are removed in each meltwater surface window. Then, 3δ iterations are performed in the vertical direction to remove noise photons, and denoising is performed in the horizontal direction using an elliptical filter kernel. If the number of bottom photons in a meltwater surface window is greater than 10, the window is identified as an ice surface meltwater window. Finally, these ice surface meltwater windows will be grouped according to the same criteria as the meltwater surface windows in step (3). The Savitzky-Go l ay filter formula is as follows:

[0114]

[0115] y i ′ is the filtered elevation value, y i+j is the elevation value in the original photon, c j is the weight coefficient of the filter, m is the number of points on both sides of the window, assuming that the input window length is N, that is, N=2m+1.

[0116] Step 5: The first photon whose elevation value exceeds three times the standard deviation of the surface elevation value on both sides of each group of ice surface meltwater windows is retained and regarded as the edge of the ice surface meltwater. The edge of the meltwater bottom is then connected to the meltwater surface, and the final meltwater bottom is generated again through Savitzky-Golay filtering. Then, a refractive index of ~1.33 is used for the height of the meltwater bottom photons to eliminate the refraction effect of the meltwater. Shallow meltwater (or slush area) has too much noise. In order to improve the confidence of the meltwater pond depth estimation, meltwater ponds with an average depth of less than 1.15m are removed. Finally, we output the position of the ice surface meltwater, the signal photons of the meltwater surface and meltwater bottom (including position and elevation values), the meltwater surface elevation, and the generated meltwater bottom.

[0117] Step 6: Obtain multispectral images that correspond to the ATL03 images in time and space, including four images of Landsat8 L1 / L2 and Sentinel2 L1C / L2A. The time difference between these images and the ATL03 data is no more than 4 days. Use the median value of the depth sampling point within an image pixel to generate training data with its multi-band reflectivity. For Landsat-8, the median depth is paired with the reflectivity values ​​of 7 bands (bands 1, 2, 3, 4, 5, 6, and 7). For Sentinel-2, first, the low-resolution bands are uniformly sampled to 10m, and the median depth is paired with the reflectivity values ​​of 9 bands (bands 2, 3, 4, 5, 6, 7, 8, 11, and 12). Finally, manually debug the training parameters of the two machine learning algorithms XGBoost and LightGBM, input the training data, and obtain the meltwater depth inversion machine learning model. Then, based on the multispectral image, the meltwater band reflectivity is input into the model to generate the final depth result.

[0118] Figure 4 The content describes the one-dimensional results of estimating meltwater depth using Landsat8 L1 data by different methods, including the method of the present invention, and compares them with the refraction-corrected ICESat-2 meltwater depth data in one dimension. Figure 4 The results show that the proposed method achieves the best accuracy on Landsat-8L1 images. The RMSE calculated by the proposed method and the refraction-corrected ICESat-2 meltwater depth validation data is 1.68m, 0.66m and 0.09m lower than the traditional lake depth inversion methods Red-RTE, Green-RTE and SBR, respectively.

[0119] Figure 5 The content describes the two-dimensional depth results estimated on the test ice lake 2 by different methods provided by the embodiments of the present invention, and compares them with ArcticDEM data in two dimensions. Figure 5The results show that the two-dimensional ice lake depth inverted by the method of the present invention is closer to the verification data ArcticDEM than the traditional algorithm, whether on Landsat-8L1 image or Sentinel-2L1C image.

[0120] Table 1 shows the one-dimensional evaluation results of meltwater depth accuracy of different algorithms on four multispectral images. The results show that, except for L2 data, the machine learning method proposed in the present invention achieves better accuracy in testing ice surface meltwater than the traditional methods RTE and SBR based only on multispectral images. XGBoost has the highest accuracy when applied to the Sentine l-2L1C image, with RMSE, MAE and MedAE of 0.54m, 0.43m and 0.36m respectively. The method of the present invention has certain advantages. The method of the present invention is expected to provide a more effective method for estimating the inversion of ice surface meltwater depth and lay a foundation for subsequent research to accurately quantify the amount of meltwater in large surface areas.

[0121] Table 1 Meltwater depth accuracy evaluation results of different algorithms on four multispectral images

[0122]

[0123] Another aspect of the embodiment of the present invention further provides an ice surface meltwater depth inversion system, comprising:

[0124] The first module is used to obtain the ICESat-2 ATL03 photon data set and perform preliminary denoising on it to obtain the target photon data;

[0125] The second module is used to separate the meltwater surface and meltwater bottom photons from the target photon data after preliminary denoising to obtain meltwater surface photons and meltwater bottom photons;

[0126] The third module is used for denoising the meltwater surface photons and performing the first ice surface meltwater retrieval;

[0127] The fourth module is used to perform denoising on the meltwater bottom photons and perform a second ice surface meltwater retrieval;

[0128] A fifth module is used to remove shallow meltwater and slush from the ice surface meltwater window according to the result of the first ice surface meltwater search and the result of the second ice surface meltwater search;

[0129] The sixth module is used to obtain multispectral image data and generate machine learning model training data in combination with meltwater depth. The meltwater band reflectance is input into the trained machine learning model based on the multispectral image data to generate the final depth result.

[0130] It can be understood that the contents of the above method embodiments are all applicable to the present system embodiments, the functions specifically implemented by the present system embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0131] The embodiment of the present invention further provides an electronic device, the electronic device includes a memory and a processor, the memory stores a computer program, and the processor implements the above-mentioned ice surface meltwater depth inversion method when executing the computer program. The electronic device can be any intelligent terminal including a tablet computer, a car computer, etc.

[0132] It can be understood that the contents of the above method embodiments are all applicable to the present device embodiments, the functions specifically implemented by the present device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0133] See also Figure 6 , Figure 6 The hardware structure of an electronic device of another embodiment is illustrated, and the electronic device includes:

[0134] The processor 601 may be implemented by a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided by the embodiments of the present invention.

[0135] The memory 602 can be implemented in the form of a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 602 can store an operating system and other application programs. When the technical solution provided in the embodiment of this specification is implemented by software or firmware, the relevant program code is stored in the memory 602, and the processor 601 calls and executes the ice surface meltwater depth inversion method of the embodiment of the present invention;

[0136] Input / output interface 603, used to implement information input and output;

[0137] Communication interface 604, used to realize communication interaction between the device and other devices, which can be realized through wired mode (such as USB, network cable, etc.) or wireless mode (such as mobile network, WI FI, Bluetooth, etc.);

[0138] Bus 605 , which transmits information between various components of the device (e.g., processor 601 , memory 602 , input / output interface 603 , and communication interface 604 );

[0139] The processor 601 , the memory 602 , the input / output interface 603 and the communication interface 604 are connected to each other in communication within the device via a bus 605 .

[0140] An embodiment of the present invention further provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the above-mentioned ice surface meltwater depth inversion method is implemented.

[0141] It can be understood that the contents of the above method embodiments are all applicable to the present storage medium embodiments, the functions specifically implemented by the present storage medium embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0142] The memory, as a non-transient computer-readable storage medium, can be used to store non-transient software programs and non-transient computer executable programs. In addition, the memory may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some embodiments, the memory may optionally include a memory remotely disposed relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0143] It should be noted that in various specific embodiments of the present invention, when it comes to the need to perform relevant processing based on data related to user identity or characteristics such as user information, user behavior data, user historical data, and user location information, the user's permission or consent will be obtained first, and the collection, use, and processing of these data will comply with relevant laws, regulations, and standards. In addition, when the embodiment of the present invention needs to obtain the user's sensitive personal information, it will obtain the user's separate permission or consent through a pop-up window or jump to a confirmation page, and after clearly obtaining the user's separate permission or consent, it will obtain the necessary user-related data for the normal operation of the embodiment of the present invention.

[0144] The embodiments described in the embodiments of the present invention are intended to more clearly illustrate the technical solutions of the embodiments of the present invention, and do not constitute a limitation on the technical solutions provided by the embodiments of the present invention. Those skilled in the art can appreciate that with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of the present invention are also applicable to similar technical problems.

[0145] Those skilled in the art will appreciate that the technical solutions shown in the figures do not limit the embodiments of the present invention and may include more or fewer steps than shown in the figures, or a combination of certain steps, or different steps.

[0146] The device embodiments described above are merely illustrative, and the units described as separate components may or may not be physically separated, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0147] Those skilled in the art will appreciate that all or some of the steps in the methods disclosed above, and the functional modules / units in the systems and devices may be implemented as software, firmware, hardware, or a suitable combination thereof.

[0148] In the several embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the above units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0149] The units described above as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0150] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0151] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including multiple instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods of various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc., various media that can store programs.

[0152] The preferred embodiments of the present invention are described above with reference to the accompanying drawings, but the scope of the rights of the embodiments of the present invention is not limited thereby. Any modification, equivalent substitution and improvement made by those skilled in the art without departing from the scope and essence of the embodiments of the present invention shall be within the scope of the rights of the embodiments of the present invention.

Claims

1. A method for inverting the depth of ice meltwater, characterized in that: The following steps are involved: Obtain the ICESat-2 ATL03 photon dataset and perform preliminary denoising on it to obtain the target photon data; The target photon data after preliminary denoising are subjected to meltwater surface and meltwater bottom photon separation to obtain meltwater surface photons and meltwater bottom photons; De-noising the meltwater surface photons and performing a first ice surface meltwater retrieval; De-noising the meltwater bottom photons and performing a second ice surface meltwater search; According to the result of the first ice surface meltwater search and the result of the second ice surface meltwater search, the ice surface meltwater window is processed to remove shallow meltwater and slush; Acquire multispectral image data and generate machine learning model training data in combination with meltwater depth. Input the meltwater band reflectance into the trained machine learning model based on the multispectral image data to generate the final depth result.

2. The ice surface meltwater depth inversion method according to claim 1 is characterized in that: The method of obtaining the ICESat-2ATL03 photon data set and performing preliminary denoising on the data to obtain target photon data includes the following steps: Download the ICESat-2 ATL03 photon dataset for the target location; The collected photon data set is subjected to preliminary denoising using the DRAGANN algorithm.

3. The ice surface meltwater depth inversion method according to claim 1 is characterized in that: The step of performing meltwater surface and meltwater bottom photon separation on the target photon data after preliminary denoising to obtain meltwater surface photons and meltwater bottom photons includes the following steps: The photons after preliminary denoising are divided into 40m processing windows along the track with a step size of 20m; Within each processing window, photon kernel density estimation is performed along the vertical height; the calculation formula for the photon kernel density estimate is: in, is the probability density estimate at the photon point x, n is the number of samples, h is the smoothing parameter, called bandwidth, which controls the width of the kernel function; K is the kernel function; the expression of the kernel function K is: Among them, u represents is the data point x i The normalized distance from point x; Find all points whose corresponding photon kernel density estimation values ​​are the minimum values ​​in the photon kernel density estimation curve, take their elevation values, and use the maximum value among these elevation values ​​as the threshold; Among them, photons with elevation less than the threshold are considered to be meltwater bottom photons, and photons with elevation greater than the threshold are considered to be meltwater surface photons.

4. The ice surface meltwater depth inversion method according to claim 1 is characterized in that: The denoising process of the meltwater surface photons and the first ice surface meltwater retrieval comprises the following steps: The 3δ iteration method is used to remove the noise photons from the meltwater surface photons; When the standard deviation of the photon elevation of the meltwater surface within the processing window is less than 0.12 m, the window is considered to be a meltwater surface window; If the distance between two meltwater windows is within 40 meters, group them together; Each group of meltwater surface windows is filled, and the meltwater surface window group containing more than 5 windows is selected as the result of the first retrieval of ice surface meltwater.

5. The ice meltwater depth inversion method according to claim 1 is characterized in that: The denoising process of the meltwater bottom photons and the second ice surface meltwater retrieval comprises the following steps: First, a Savitzky-Golay filter is used to generate a fitting curve for bottom photons, and then the fitting value is subtracted from the height of the photons to eliminate the trend of bottom photons; where the expression of the Savitzky-Golay filter is: Among them, y i ′ is the elevation value after filtering; y i+j is the elevation value in the original photon; c j is the weight coefficient of the filter and m is the number of points on both sides of the window; All bottom photons within 1 m below the meltwater surface elevation were removed to avoid the influence of multi-surface reflection on the final extracted meltwater depth; All photons below the median elevation of the bottom photons were removed in each meltwater surface window. Then, 3δ iterations were performed to remove noise photons in the vertical direction and denoised by an elliptical filter kernel in the horizontal direction. If the number of bottom photons in a meltwater surface window is greater than 10, the window is identified as an ice surface meltwater window; These ice surface meltwater windows are grouped according to the same criteria as the meltwater surface windows obtained in the first ice surface meltwater retrieval.

6. The ice surface meltwater depth inversion method according to claim 1 is characterized in that: The method of removing shallow meltwater and slush from the ice surface meltwater window according to the result of the first ice surface meltwater search and the result of the second ice surface meltwater search comprises the following steps: The photons whose first elevation value on both sides of each group of ice surface meltwater windows exceeds the range of three times the standard deviation of the surface elevation value are retained and regarded as the edge of ice surface meltwater; Connect the meltwater bottom edge to the meltwater surface and generate the final meltwater bottom through Savitzky-Golay filtering again; A refractive index of ~1.33 was used for the height of meltwater bottom photons to eliminate the refraction effect of meltwater; Remove meltwater ponds with an average depth of less than 1.15m.

7. The ice surface meltwater depth inversion method according to claim 1 is characterized in that: The method of obtaining multispectral image data and generating machine learning model training data in combination with meltwater depth, inputting meltwater band reflectance into the trained machine learning model based on the multispectral image data, and generating a final depth result includes the following steps: Obtain multispectral images that correspond to ATL03 images in time and space, including Landsat8 L1 / L2 and Sentinel2L1C / L2A, a total of four images. The time difference between these images and ATL03 data is no more than 4 days; The median value of the depth sampling point within an image pixel is used to generate training data with its multi-band reflectance. For Landsat-8, the median depth is paired with the reflectance value of 7 bands. For Sentinel-2, the low-resolution bands are first uniformly sampled to 10m, and the median depth is paired with the reflectance value of 9 bands. Manually debug the training parameters of the two machine learning algorithms, XGBoost and LightGBM, input the training data, and derive the meltwater depth inversion machine learning model. Then, based on the multispectral image, the meltwater band reflectivity is input into the model to generate the final depth result.

8. An ice surface meltwater depth inversion system, characterized in that: include: The first module is used to obtain the ICESat-2 ATL03 photon data set and perform preliminary denoising on it to obtain the target photon data; The second module is used to separate the meltwater surface and meltwater bottom photons from the target photon data after preliminary denoising to obtain meltwater surface photons and meltwater bottom photons; The third module is used for denoising the meltwater surface photons and performing the first ice surface meltwater retrieval; The fourth module is used to perform denoising on the meltwater bottom photons and perform a second ice surface meltwater retrieval; A fifth module is used to remove shallow meltwater and slush from the ice surface meltwater window according to the result of the first ice surface meltwater search and the result of the second ice surface meltwater search; The sixth module is used to obtain multispectral image data and generate machine learning model training data in combination with meltwater depth. The meltwater band reflectance is input into the trained machine learning model based on the multispectral image data to generate the final depth result.

9. An electronic device, characterized in that: including a processor and a memory; The memory is used to store programs; The processor executes the program to implement the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The storage medium stores a program, and the program is executed by a processor to implement the method according to any one of claims 1 to 7.

Images