A method for temporal and spatial alignment of multi-source satellite upper atmospheric oxygen and nitrogen concentration observations

Abstract

The present invention discloses a method for spatiotemporal alignment of multi-source satellite upper atmospheric oxygen and nitrogen concentration observations, comprising collecting TIMED observation data and FY3E observation data of an area to be observed; performing coordinate conversion on the TIMED observation data and the FY3E observation data, and performing coordinate expansion on the TIMED observation data and the FY3E observation data after the coordinate conversion; setting a time window within a specified time period, using the observation time of the FY3E observation data after the coordinate expansion as a matching target, screening the corresponding TIMED observation data, setting a spatial distance window, using the FY3E observation data as a matching target, screening the corresponding TIMED observation data to obtain TIMED samples; performing coordinate inverse mapping and deduplication screening on the TIMED samples with longitudes outside [-180°, 180°] to obtain nitrogen concentration matching samples and oxygen concentration matching samples. This method can not only improve the accuracy of spatiotemporal alignment of oxygen and nitrogen concentration observations, but also has good interpretability and can be directly applied to the spatiotemporal alignment system of oxygen and nitrogen concentration observations.

Description

Technical Field

[0001] The present invention relates to the field of space weather, and in particular to a method for spatiotemporal alignment of multi-source satellite upper atmosphere oxygen and nitrogen concentration observations. Background Art

[0002] At present, the vast majority of comparative analyses of atmospheric composition from multi-source observations are mainly focused on the middle and lower atmosphere, while research on the upper atmosphere is relatively scarce. Although some studies have compared the observational data from GOLD (Global-scale Observations of the Limb and Disk) and TIMED (Thermosphere Ionosphere Mesosphere Energetics and Dynamics), since GOLD is a synchronous satellite, its observation range is limited and cannot cover the entire globe. Therefore, there is still a gap in the study of other regions, such as the upper atmosphere of China. The evolution of the upper atmosphere is significantly affected by solar activity, which has a distinct 11-year cycle. Since entering the 25th solar cycle in 2019, it is currently at its peak of solar activity, with frequent flares and solar ejections. These phenomena may affect the state and composition of the Earth's upper atmosphere.

[0003] In recent years, the increasing frequency of extreme weather events, such as droughts, floods, and earthquakes, has sparked concern about their relationship to abnormal circulation in the upper atmosphere. Whether the increase in these disasters is linked to abnormal circulation in the upper atmosphere requires further research. In the past, due to the relatively thin upper atmosphere, its impact on meteorological systems received insufficient attention, resulting in relatively few observations and studies. However, with the continuous advancement of research on the upper atmosphere, the scientific community has gradually recognized its important role in the lower atmosphere and the energy cycle outside the Earth.

[0004] Observational data on the upper atmosphere are becoming increasingly abundant, and a growing number of studies are leveraging this data to expand and refine theories of atmospheric circulation mechanisms. For example, observations of upper atmospheric parameters such as composition, temperature, and wind speed can provide a better understanding of the interaction between the upper and lower atmospheres and their impact on global climate. Furthermore, spatiotemporal alignment algorithms play a crucial role in this process. They not only help accumulate long-term datasets but also reveal the vertical circulation mechanisms of the atmosphere, thereby providing a more accurate foundation for climate models and predictions. With the advancement of data analysis techniques, the application of spatiotemporal alignment algorithms is expanding, enabling the effective integration of observational data from various satellites. These techniques enable researchers to better analyze the dynamics of the upper atmosphere and its relationship with surface meteorological phenomena, thereby providing a more reliable scientific basis for climate change predictions and response strategies. These studies not only advance science but also provide crucial data support for policymakers in addressing the increasingly severe climate challenges. Summary of the Invention

[0005] The purpose of the present invention is to provide a method for temporal and spatial alignment of multi-source satellite upper atmosphere oxygen and nitrogen concentration observations.

[0006] To achieve the above object, the present invention is implemented according to the following technical solutions:

[0007] The present invention comprises the following steps:

[0008] Collecting TIMED observation data and FY3E observation data of the area to be observed, and preprocessing the TIMED observation data and the FY3E observation data;

[0009] performing coordinate conversion on the TIMED observation data and the FY3E observation data, and performing coordinate expansion on the TIMED observation data and the FY3E observation data after the coordinate conversion;

[0010] A time window is set within a specified time period, and the observation time of satellite FY3E is used as a matching target to filter the TIMED observation data. A spatial distance window is set, and the observation time of satellite FY3E is used as a matching target to filter the TIMED observation data. The filtering principle is as follows:

[0011] For sample FY3E , calculate the spatial distance:

[0012] D FY3E,TIMED =(P(Lat FY3E ,Lon FY3E )-P(Lat TIMED ,Lon TIMED )) 2*0.5

[0013] The spatial distance between the FY3E observation data and the TIMED observation data is D FY3E,TIMED , the longitude of TIMED observation data is Lat TIMED , the longitude coordinate of the FY3E observation data is Lat FY3E , the latitude of TIMED observation data is Lon TIMED , the latitude coordinate of FY3E observation data is Lon FY3E The latitude and longitude coordinates of the FY3E observation data are P(Lat FY3E ,Lon FY3E ), the latitude and longitude coordinates of the TIMED observation data are P(Lat TIMED ,Lon TIMED );

[0014] On the basis of the time window screening, the spatial distance window D is set. Δ , filter sample space distance D <D Δ All samples of , and then retrieve the sample with the smallest D TIMED , get the group matching sample (Sample FY3E ,Sample TIMED );

[0015] The TIMED samples with longitudes outside [-180°, 180°] are subjected to coordinate inverse mapping and deduplication screening to obtain nitrogen concentration matching samples and oxygen concentration matching samples.

[0016] Furthermore, the coordinate conversion method includes:

[0017] The longitude range of satellite FY3E is [-180°, 180°], and the longitude range of satellite IMED is [0, 360°]. Taking FY3E observation data as the sample matching target, first perform longitude conversion on TIMED observation data. The expression is:

[0018] Lon TIMED_new =Lon TIMED_old -180°

[0019] The longitude of the TIMED observation data is Lon TIMED_old , the longitude of the TIMED observation data after longitude conversion is Lon TIMED_new .

[0020] Furthermore, the coordinate expansion method includes:

[0021] The TIMED observation at (0°, 180°] is mapped to (-360°, -180°], and the TIMED observation data and FY3E observation data at (-180°, 0°] are mapped to (180°, 360°], finally obtaining the sample coordinate range of [-360°, 360°].

[0022] Furthermore, the screening method includes matching and selecting, matching the observation data, selecting the matched observation time based on the screening principle, and obtaining the screening result.

[0023] Furthermore, the matching method includes:

[0024] S1 Observation time matching based on time window

[0025] Sort the TIMED observation data in chronological order, input the sorted TIMED observation data into the time matching model, and use the dimensionality reduction algorithm to reduce the TIMED observation data to a one-dimensional convolutional neural network with time series as the main axis to obtain the one-dimensional convolution output of the time series. The expression is:

[0026]

[0027] The size of the convolution kernel is N, and the time sequence weight of the a-th convolution kernel is φ a , the bias term is z, and the input value of the TIMED observation data of the a+sth convolution kernel u-1th time window is The output of the a-th one-dimensional convolution of the u-1-th time window time series is

[0028] The one-dimensional convolution output of the time series and the observation time are incorporated into the matching dataset to form the attention score identification coefficient, which is expressed as:

[0029]

[0030] The attention score identification coefficient of the x-th observation time is The attention score function at the x-th observation time is μ x , the observation time is c, the one-dimensional convolution output of the time series is B, and the attention score is marked as w(B,c);

[0031] Calculate the attention score for the observation time:

[0032]

[0033] The attention score is The xth observation time is c x , the number of TIMED observation data is M;

[0034] The regularization parameter is used to improve the loss function of the time matching model, which is expressed as:

[0035]

[0036] The improved loss function is The mean square error of the loss function is The regularization strength is The regularization parameter is The order of the norm is p, and the regularization term is F(φ);

[0037] Output the TIMED observation data with the highest attention score as the matching result;

[0038] S2 Observation Sample Matching Based on Spatial Distance Window

[0039] The TIMED observation data were selected by matching the observation time of satellite FY3E;

[0040] The TIMED observation data are regionally classified according to space to obtain the observation area dataset. The spatial distance between the FY3E observation data and the TIMED observation data is calculated, and the one with the smallest distance is output as the matching result.

[0041] Furthermore, the selection method includes:

[0042] A. Filtering TIMED observation data based on the selection of time windows

[0043] Extract TIMED observation data through the time window and convert the time format of IMED observation data according to FY3E observation data;

[0044] Filter the observation time of the corresponding satellite TIMED. The expression is:

[0045] Datetime TIMED =Datetime FY3E -T Δ

[0046] The observation time of satellite TIMED is Datetime TIMED , the time window is T Δ , the observation time of satellite FY3E is Datetime FY3E ;

[0047] Extract the TIMED observation data within the observation time as the time window screening result;

[0048] B. Selection and screening of TIMED observation data based on spatial distance

[0049] The correlation coefficients between the FY3E observation data and the IMED observation data were calculated based on the TIMED observation data after selecting the time window by spatial distance, and the IMED observation data with a correlation coefficient less than 0.216 were eliminated;

[0050] Calculate the information coefficient of IMED observation data, and calculate the distance metric based on the IMED observation data with the maximum information coefficient. The expression is:

[0051]

[0052] The longitude coordinate of the i-th FY3E observation data is Lat FY3E,i , the latitude of the jth TIMED observation data is Lon TIMED,j , the latitude coordinate of the i-th FY3E observation data is Lon FY3E,i , the longitude coordinate of the jth TIMED observation data is Lat TIMED,j , the distance measure between FY3E observation data and TIMED observation data is K i,j , the Harvard-Sain coefficient is ρ;

[0053] Eliminate the TIMED observation data with the smallest distance metric, repeat the iteration until all time windows are traversed, and output the remaining TIMED observation data as the spatiotemporal distance screening result.

[0054] In a second aspect, an embodiment of the present application further provides an electronic device, including:

[0055] A processor; and a memory arranged to store computer executable instructions, which when executed cause the processor to perform the method steps described in the first aspect.

[0056] In a third aspect, an embodiment of the present application further provides a computer-readable storage medium, which stores one or more programs. When the one or more programs are executed by an electronic device including multiple applications, the electronic device executes the method steps described in the first aspect.

[0057] The beneficial effects of the present invention are:

[0058] The present invention is a method for spatiotemporal alignment of multi-source satellite upper atmospheric oxygen and nitrogen concentration observations. Compared with the prior art, the present invention has the following technical effects:

[0059] The present invention can improve the accuracy of the spatiotemporal alignment of multi-source satellite upper-atmosphere oxygen and nitrogen concentration observations through preprocessing, coordinate conversion, coordinate expansion, data screening, coordinate inverse mapping and deduplication steps, thereby improving the precision of the spatiotemporal alignment of multi-source satellite upper-atmosphere oxygen and nitrogen concentration observations, optimizing the spatiotemporal alignment of multi-source satellite upper-atmosphere oxygen and nitrogen concentration observations, greatly saving resources, and improving work efficiency. It can realize the intelligent spatiotemporal alignment of multi-source satellite upper-atmosphere oxygen and nitrogen concentration observations, and perform data matching and screening on the spatiotemporal alignment of multi-source satellite upper-atmosphere oxygen and nitrogen concentration observations in real time. It is of great significance to the spatiotemporal alignment of multi-source satellite upper-atmosphere oxygen and nitrogen concentration observations, can adapt to the spatiotemporal alignment of multi-source satellite upper-atmosphere oxygen and nitrogen concentration observations of different standards, and has certain universality. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 A flowchart of the steps of the method for spatiotemporal alignment of multi-source satellite upper atmosphere oxygen and nitrogen concentration observations according to the present invention;

[0061] Figure 2 This is a schematic diagram of the structure of an electronic device in an embodiment of this specification. DETAILED DESCRIPTION

[0062] The present invention will be further described below through specific examples. The illustrative examples and descriptions of the present invention are used to explain the present invention but are not intended to limit the present invention.

[0063] The method for spatiotemporal alignment of multi-source satellite upper atmosphere oxygen and nitrogen concentration observations of the present invention comprises the following steps:

[0064] like Figure 1 As shown, in this embodiment, the following steps are included:

[0065] Collecting TIMED observation data and FY3E observation data of the area to be observed, and preprocessing the TIMED observation data and the FY3E observation data;

[0066] performing coordinate conversion on the TIMED observation data and the FY3E observation data, and performing coordinate expansion on the TIMED observation data and the FY3E observation data after the coordinate conversion;

[0067] A time window is set within a specified time period, and the observation time of satellite FY3E is used as a matching target to filter the TIMED observation data. A spatial distance window is set, and the observation time of satellite FY3E is used as a matching target to filter the TIMED observation data. The filtering principle is as follows:

[0068] For sample FY3E , calculate the spatial distance:

[0069] D FY3E,TIMED =(P(Lat FY3E ,Lon FY3E )-P(Lat TIMED ,Lon TIMED )) 2*0.5

[0070] The spatial distance between the FY3E observation data and the TIMED observation data is D FY3E,TIMED , the longitude of TIMED observation data is Lat TIMED , the longitude coordinate of the FY3E observation data is Lat FY3E , the latitude of TIMED observation data is Lon TIMED , the latitude coordinate of FY3E observation data is Lon FY3E The latitude and longitude coordinates of the FY3E observation data are P(Lat FY3E ,Lon FY3E ), the latitude and longitude coordinates of the TIMED observation data are P(Lat TIMED ,Lon TIMED );

[0071] On the basis of the time window screening, the spatial distance window D is set. Δ , filter sample space distance D <D Δ All samples of , and then retrieve the sample with the smallest D TIMED , get the group matching sample (Sample FY3E ,Sample TIMED );

[0072] The TIMED samples with longitudes outside [-180°, 180°] are subjected to coordinate inverse mapping and deduplication screening to obtain nitrogen concentration matching samples and oxygen concentration matching samples.

[0073] In this embodiment, the coordinate conversion method includes:

[0074] The longitude range of satellite FY3E is [-180°, 180°], and the longitude range of satellite IMED is [0, 360°]. Taking FY3E observation data as the sample matching target, first perform longitude conversion on TIMED observation data. The expression is:

[0075] Lon TIMED_new =Lon TIMED_old -180°

[0076] The longitude of the TIMED observation data is Lon TIMED_old , the longitude of the TIMED observation data after longitude conversion is Lon TIMED_new .

[0077] In this embodiment, the coordinate expansion method includes:

[0078] The TIMED observation at (0°, 180°] is mapped to (-360°, -180°], and the TIMED observation data and FY3E observation data at (-180°, 0°] are mapped to (180°, 360°], finally obtaining the sample coordinate range of [-360°, 360°].

[0079] In this embodiment, the screening method includes matching and selecting, matching the observation data, and selecting the matched observation time based on the screening principle to obtain the screening result.

[0080] In this embodiment, the matching method includes:

[0081] S1 Observation time matching based on time window

[0082] Sort the TIMED observation data in chronological order, input the sorted TIMED observation data into the time matching model, and use the dimensionality reduction algorithm to reduce the TIMED observation data to a one-dimensional convolutional neural network with time series as the main axis to obtain the one-dimensional convolution output of the time series. The expression is:

[0083]

[0084] The size of the convolution kernel is N, and the time sequence weight of the a-th convolution kernel is φ a , the bias term is z, and the input value of the TIMED observation data of the a+sth convolution kernel u-1th time window is The output of the a-th one-dimensional convolution of the u-1-th time window time series is

[0085] The one-dimensional convolution output of the time series and the observation time are incorporated into the matching dataset to form the attention score identification coefficient, which is expressed as:

[0086]

[0087] The attention score identification coefficient at the xth observation time is The attention score function at the x-th observation time is μ x , the observation time is c, the one-dimensional convolution output of the time series is B, and the attention score is marked as w(B,c);

[0088] Calculate the attention score for the observation time:

[0089]

[0090] The attention score is The xth observation time is c x , the number of TIMED observation data is M;

[0091] The regularization parameter is used to improve the loss function of the time matching model, which is expressed as:

[0092]

[0093] The improved loss function is The mean square error of the loss function is The regularization strength is The regularization parameter is The order of the norm is p, and the regularization term is F(φ);

[0094] Output the TIMED observation data with the highest attention score as the matching result;

[0095] S2 Observation Sample Matching Based on Spatial Distance Window

[0096] The TIMED observation data were selected by matching the observation time of satellite FY3E;

[0097] The TIMED observation data are regionally classified according to space to obtain the observation area dataset. The spatial distance between the FY3E observation data and the TIMED observation data is calculated, and the one with the smallest distance is output as the matching result.

[0098] In this embodiment, the selection method includes:

[0099] A. Filtering TIMED observation data based on the selection of time windows

[0100] Extract TIMED observation data through the time window and convert the time format of IMED observation data according to FY3E observation data;

[0101] Filter the observation time of the corresponding satellite TIMED. The expression is:

[0102] Datetime TIMED =Datetime FY3E -T Δ

[0103] The observation time of satellite TIMED is Datetime TIMED , the time window is T Δ , the observation time of satellite FY3E is Datetime FY3E ;

[0104] Extract the TIMED observation data within the observation time as the time window screening result;

[0105] B. Selection and screening of TIMED observation data based on spatial distance

[0106] The correlation coefficients between the FY3E observation data and the IMED observation data were calculated based on the TIMED observation data after selecting the time window by spatial distance, and the IMED observation data with a correlation coefficient less than 0.216 were eliminated;

[0107] Calculate the information coefficient of IMED observation data, and calculate the distance metric based on the IMED observation data with the maximum information coefficient. The expression is:

[0108]

[0109] The longitude coordinate of the i-th FY3E observation data is Lat FY3E,i , the latitude of the jth TIMED observation data is Lon TIMED,j , the latitude coordinate of the i-th FY3E observation data is Lon FY3E,i , the longitude coordinate of the jth TIMED observation data is Lat TIMED,j , the distance measure between FY3E observation data and TIMED observation data is K i,j , the Harvard-Sain coefficient is ρ;

[0110] Eliminate the TIMED observation data with the smallest distance metric, repeat the iteration until all time windows are traversed, and output the remaining TIMED observation data as the spatiotemporal distance screening result.

[0111] Figure 2 This is a schematic diagram of the structure of an electronic device according to an embodiment of the present application. Figure 2 At the hardware level, the electronic device includes a processor and, optionally, an internal bus, a network interface, and memory. The memory may include internal memory, such as high-speed random-access memory (RAM), or non-volatile memory, such as at least one disk drive. Of course, the electronic device may also include other hardware required for its services.

[0112] The processor, network interface, and memory can be interconnected via an internal bus, which can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, or an EISA (Extended Industry Standard Architecture) bus. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 2 Only one bidirectional arrow is used in the diagram, but this does not mean that there is only one bus or one type of bus.

[0113] The memory is used to store programs. Specifically, the program may include program code, which includes computer operating instructions. The memory may include internal memory and non-volatile memory, and provides instructions and data to the processor.

[0114] The processor reads the corresponding computer program from the non-volatile memory into the internal memory and then executes it, logically forming a device for spatiotemporal alignment of multi-source satellite upper-atmosphere oxygen and nitrogen concentration observations. The processor executes the program stored in the memory and is specifically configured to perform any of the aforementioned methods for spatiotemporal alignment of multi-source satellite upper-atmosphere oxygen and nitrogen concentration observations.

[0115] The above application Figure 1The method for spatiotemporal alignment of multi-source satellite upper atmospheric oxygen and nitrogen concentration observations disclosed in the illustrated embodiment can be applied to a processor or implemented by a processor. The processor may be an integrated circuit chip with signal processing capabilities. During implementation, each step of the above method can be completed by hardware integrated logic circuits in the processor or by software instructions. The above processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the various methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in conjunction with the embodiments of this application can be directly implemented and executed by a hardware decoding processor, or by a combination of hardware and software modules in the decoding processor. The software module can be located in a storage medium well-known in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, etc. The storage medium is located in the memory, and the processor reads the information in the memory and, in conjunction with its hardware, completes the steps of the above method.

[0116] The electronic device may also perform Figure 1 A method for temporal and spatial alignment of multi-source satellite upper atmospheric oxygen and nitrogen concentration observations, and implementation Figure 1 The functions of the illustrated embodiment will not be described in detail in the embodiments of the present application.

[0117] An embodiment of the present application also proposes a computer-readable storage medium, which stores one or more programs, and the one or more programs include instructions. When the instructions are executed by an electronic device including multiple applications, the instructions perform the aforementioned method of spatiotemporal alignment of upper atmospheric oxygen and nitrogen concentration observations by any multi-source satellite.

[0118] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0119] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the steps in the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0120] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0121] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0122] In a typical configuration, a computing device includes one or more processors (CPUs), input / output interfaces, network interfaces, and memory.

[0123] Memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.

[0124] Computer-readable media includes permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. The information can be computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include transitory computer-readable media (transitory media), such as modulated data signals and carrier waves.

[0125] It should also be noted that the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, commodity, or apparatus that includes a series of elements includes not only those elements but also other elements not explicitly listed, or includes elements inherent to such process, method, commodity, or apparatus. In the absence of further limitations, an element defined by the phrase "comprises a ..." does not exclude the presence of other identical elements in the process, method, commodity, or apparatus that includes the element.

[0126] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0127] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for spatiotemporal alignment of multi-source satellite upper atmosphere oxygen and nitrogen concentration observations, characterized in that: The following steps are involved: Collecting TIMED observation data and FY3E observation data of the area to be observed, and preprocessing the TIMED observation data and the FY3E observation data; performing coordinate conversion on the TIMED observation data and the FY3E observation data, and performing coordinate expansion on the TIMED observation data and the FY3E observation data after the coordinate conversion; A time window is set within a specified time period, and the observation time of satellite FY3E is used as a matching target to filter the TIMED observation data. A spatial distance window is set, and the observation time of satellite FY3E is used as a matching target to filter the TIMED observation data. The filtering principle is as follows: For sample FY3E , calculate the spatial distance: D FY3E,TIMED =(P(Lat FY3E ,Lon FY3E )-P(Lat TIMED ,Lon TIMED )) 2*0.5 The spatial distance between the FY3E observation data and the TIMED observation data is D FY3E,TIMED , the longitude of TIMED observation data is Lat TIMED , the longitude coordinate of the FY3E observation data is Lat FY3E , the latitude of TIMED observation data is Lon TIMED , the latitude coordinate of FY3E observation data is Lon FY3E The latitude and longitude coordinates of the FY3E observation data are P(Lat FY3E ,Lon FY3E ), the latitude and longitude coordinates of the TIMED observation data are P(Lat TIMED ,Lon TIMED ); The matching method includes: S1 Observation time matching based on time window Sort the TIMED observation data in chronological order, input the sorted TIMED observation data into the time matching model, and use the dimensionality reduction algorithm to reduce the TIMED observation data to a one-dimensional convolutional neural network with the time series as the main axis to obtain the one-dimensional convolution output of the time series; The one-dimensional convolution output of the time series and the observation time are incorporated into the matching dataset to form the attention score identification coefficient; Calculate the attention score of the observation time, use the regularization parameter to improve the loss function of the time matching model, and output the TIMED observation data with the highest attention score as the matching result; S2 Observation Sample Matching Based on Spatial Distance Window The TIMED observation data were selected by matching the observation time of satellite FY3E; The TIMED observation data are regionally classified according to space to obtain the observation area dataset. The spatial distance between the FY3E observation data and the TIMED observation data is calculated, and the one with the smallest distance is output as the matching result. On the basis of time window screening, set the spatial distance window D Δ , filter sample space distance D<D Δ All samples of , and then retrieve the sample with the smallest D TIMED , get the group matching sample (Sample FY3E ,Sample TIMED ); The TIMED samples with longitudes outside [-180°, 180°] were subjected to coordinate inverse mapping and deduplication screening to obtain nitrogen concentration matching samples and oxygen concentration matching samples.

2. The method for spatiotemporal alignment of multi-source satellite upper atmospheric oxygen and nitrogen concentration observations according to claim 1, characterized in that: The coordinate conversion method comprises: The longitude range of satellite FY3E is [-180°, 180°], and the longitude range of satellite IMED is [0, 360°]. Taking FY3E observation data as the sample matching target, first perform longitude conversion on TIMED observation data. The expression is: Lon TIMED_new =Lon TIMED_old -180° The longitude of the TIMED observation data is Lon TIMED_old , the longitude of the TIMED observation data after longitude conversion is Lon TIMED_new .

3. The method for spatiotemporal alignment of multi-source satellite upper atmosphere oxygen and nitrogen concentration observations according to claim 1, characterized in that: The coordinate expansion method includes: The TIMED observation at (0°, 180°] is mapped to (-360°, -180°], and the TIMED observation data and FY3E observation data at (-180°, 0°] are mapped to (180°, 360°], finally obtaining the sample coordinate range of [-360°, 360°].

4. The method for spatiotemporal alignment of multi-source satellite upper atmosphere oxygen and nitrogen concentration observations according to claim 1, characterized in that: The screening method includes matching and selecting, matching the observation data, selecting the matched observation time based on the screening principle, and obtaining the screening result.

5. The method for spatiotemporal alignment of multi-source satellite upper atmosphere oxygen and nitrogen concentration observations according to claim 4, characterized in that: The matching method includes: The one-dimensional convolution output of the time series is expressed as: The size of the convolution kernel is N, and the time sequence weight of the a-th convolution kernel is φ a , the bias term is z, and the input value of the TIMED observation data of the a+sth convolution kernel u-1th time window is The output of the a-th one-dimensional convolution of the u-1-th time window time series is The attention score identification coefficient is expressed as: The attention score identification coefficient of the x-th observation time is θ x , the attention score function at the x-th observation time is μ x , the observation time is c, the one-dimensional convolution output of the time series is B, and the attention score is marked as w(B,c); The attention score of the observation time is expressed as: The attention score is The xth observation time is c x , the number of TIMED observation data is M; The loss function is expressed as: The improved loss function is The mean square error of the loss function is The regularization strength is l, and the regularization parameter is The order of the norm is p, and the regularization term is F(φ).

6. The method for spatiotemporal alignment of multi-source satellite upper atmosphere oxygen and nitrogen concentration observations according to claim 4, characterized in that: The selection method comprises: A. Filtering TIMED observation data based on the selection of time windows Extract TIMED observation data through the time window and convert the time format of IMED observation data according to FY3E observation data; Filter the observation time of the corresponding satellite TIMED. The expression is: Datetime TIMED =Datetime FY3E -T Δ The observation time of satellite TIMED is Datetime TIMED , the time window is T Δ , the observation time of satellite FY3E is Datetime FY3E ; Extract the TIMED observation data within the observation time as the time window screening result; B. Selection and screening of TIMED observation data based on spatial distance The correlation coefficients between the FY3E observation data and the IMED observation data were calculated based on the TIMED observation data after selecting the time window by spatial distance, and the IMED observation data with a correlation coefficient less than 0.216 were eliminated; Calculate the information coefficient of IMED observation data, and calculate the distance metric based on the IMED observation data with the maximum information coefficient. The expression is: The longitude coordinate of the i-th FY3E observation data is Lat FY3E,i , the latitude of the jth TIMED observation data is Lon TIMED,j , the latitude coordinate of the i-th FY3E observation data is Lon FY3E,i , the longitude coordinate of the jth TIMED observation data is Lat TIMED,j , the distance measure between FY3E observation data and TIMED observation data is K i,j , the Harvard-Sain coefficient is ρ; Eliminate the TIMED observation data with the smallest distance metric, repeat the iteration until all time windows are traversed, and output the remaining TIMED observation data as the spatiotemporal distance screening result.

7. An electronic device comprising: processor; as well as A memory arranged to store computer executable instructions, which, when executed, cause the processor to perform the method according to any one of claims 1 to 5.

8. A computer-readable storage medium storing one or more programs, wherein when the one or more programs are executed by an electronic device including a plurality of application programs, the electronic device executes the method according to any one of claims 1 to 5.

Images