Method for determining atmospheric brown carbon aerosol light absorption characteristics and electronic device

Abstract

The application provides a method for determining the light absorption characteristics of atmospheric brown carbon aerosols and an electronic device, wherein the method comprises: obtaining aerosol observation data and organic aerosol mass spectrum data; determining the brown carbon absorption coefficient corresponding to each wavelength according to the aerosol observation data; determining a plurality of organic aerosol sources and the aerosol concentration corresponding to each organic aerosol source according to the organic aerosol mass spectrum data; training a plurality of brown carbon source decision models according to the brown carbon absorption coefficient corresponding to each wavelength and the aerosol concentration corresponding to each organic aerosol source, and determining at least one brown carbon source corresponding to each wavelength according to each brown carbon source decision model; and determining the absorption spectrum of each brown carbon source according to each brown carbon source corresponding to each wavelength. The application can accurately retrieve the absorption spectrum of each brown carbon source, realize the cross-scale correlation from chemical composition to climate effect, and improve the accuracy and scientificity of the evaluation of the climate impact of brown carbon.

Description

Technical Field

[0001] This application relates to the field of brown carbon analysis technology, and more specifically, to a method and electronic device for determining the light absorption characteristics of atmospheric brown carbon aerosols. Background Technology

[0002] Atmospheric aerosols, as a crucial factor influencing the climate system, significantly impact the radiation balance of the Earth-atmosphere system through direct scattering and absorption of solar radiation, and indirect alteration of cloud microphysical properties. Brown carbon (BrC) is a type of organic aerosol component exhibiting wavelength-dependent light absorption characteristics. Due to its widespread presence in urban and regional atmospheres and its significant spatiotemporal variability, its contribution to atmospheric radiative forcing has garnered increasing attention. Accurately quantifying the light absorption behavior and climate effects of brown carbon has become a key scientific issue in atmospheric environment and climate change research.

[0003] Currently, the observation and estimation of the light absorption characteristics of brown carbon mostly rely on statistical regression methods to establish an empirical relationship between source concentration and brown carbon absorption, thereby enabling the identification and contribution assessment of major brown carbon sources.

[0004] However, traditional technical approaches suffer from drawbacks such as low accuracy and difficulty in capturing the dynamic relationship between various source types and light absorption responses under complex nonlinear environmental processes. Summary of the Invention

[0005] The purpose of this application is to provide a method and electronic device for determining the light absorption characteristics of atmospheric brown carbon aerosol, in order to address the shortcomings of the prior art, and to solve the problems of low accuracy and difficulty in capturing the dynamic relationship between multiple source types and light absorption response under complex nonlinear environmental processes in the prior art.

[0006] To achieve the above objectives, the technical solutions adopted in the embodiments of this application are as follows:

[0007] In a first aspect, one embodiment of this application provides a method for determining the light absorption characteristics of atmospheric brown carbon aerosol, the method comprising:

[0008] Acquire aerosol observation data and organic aerosol mass spectrometry data. The aerosol observation data includes: filter membrane aerosol absorption coefficients corresponding to multiple wavelengths, correction parameters corresponding to multiple wavelengths, and optical attenuation parameters corresponding to multiple wavelengths. The organic aerosol mass spectrometry data includes: aerosol composition at multiple sampling time points.

[0009] Based on the aerosol observation data, the absorption coefficient of brown carbon corresponding to each wavelength was determined;

[0010] Based on the organic aerosol mass spectrometry data, multiple sources of organic aerosols and the corresponding aerosol concentrations of each organic aerosol source were determined.

[0011] Based on the brown carbon absorption coefficient corresponding to each wavelength and the aerosol concentration corresponding to each organic aerosol source, multiple brown carbon source decision models are trained, and based on each brown carbon source decision model, at least one brown carbon source corresponding to each wavelength is determined.

[0012] Based on each of the brown carbon sources corresponding to each wavelength, the absorption spectrum of each brown carbon source is determined, and the absorption spectrum of the brown carbon source is used to characterize the change of the light absorption characteristics of the brown carbon source with wavelength.

[0013] As one possible implementation, it also includes:

[0014] Based on the absorption spectra of each brown carbon source, the direct radiative forcing corresponding to each brown carbon source is determined. The direct radiative forcing includes: direct radiative forcing of brown carbon aerosols at the top of the atmosphere, direct radiative forcing of brown carbon aerosols at the Earth's surface, and direct radiative forcing of brown carbon aerosols in the atmosphere.

[0015] As one possible implementation, determining the brown carbon absorption coefficient corresponding to each wavelength based on the aerosol observation data includes:

[0016] Based on the filter membrane aerosol absorption coefficient corresponding to each wavelength, the correction parameters corresponding to each wavelength, the optical attenuation parameters corresponding to each wavelength, and the preset multiple scattering correction parameters, the atmospheric aerosol absorption coefficient corresponding to each wavelength is calculated.

[0017] Based on the atmospheric aerosol absorption coefficient corresponding to the target wavelength and each wavelength, the black carbon absorption coefficient corresponding to each wavelength is calculated.

[0018] The difference between the atmospheric aerosol absorption coefficient and the black carbon absorption coefficient at each wavelength is calculated to obtain the brown carbon absorption coefficient at each wavelength.

[0019] As one possible implementation, determining multiple organic aerosol sources and the corresponding aerosol concentrations based on the organic aerosol mass spectrometry data includes:

[0020] Based on the organic aerosol mass spectrometry data and the preset positive definite matrix factorization model, the concentration of each organic aerosol component at each time point is obtained, and each organic aerosol component is taken as an organic aerosol source. Based on the concentration of each organic aerosol component at each time point, the aerosol concentration corresponding to each organic aerosol source is obtained.

[0021] As one possible implementation, multiple brown carbon source decision models are trained based on the brown carbon absorption coefficients corresponding to each wavelength and the aerosol concentrations corresponding to each organic aerosol source, including:

[0022] The brown carbon absorption coefficients corresponding to each wavelength and the aerosol concentrations corresponding to each organic aerosol source are time-aligned to obtain the target brown carbon absorption coefficients corresponding to each wavelength and the target aerosol concentrations corresponding to each organic aerosol source.

[0023] By iterating through each wavelength, for the first wavelength reached, the aerosol concentration corresponding to each organic aerosol source is used as the input variable, and the target brown carbon absorption coefficient corresponding to the first wavelength is used as the prediction variable, thereby training a brown carbon source decision model corresponding to the first wavelength.

[0024] As one possible implementation, determining at least one brown carbon source corresponding to each wavelength based on each brown carbon source decision model includes:

[0025] Based on the decision models of multiple brown carbon sources corresponding to each wavelength, the importance coefficient of each brown carbon source corresponding to each wavelength is determined. The importance coefficient is used to characterize the degree of influence of each brown carbon source on the brown carbon absorption coefficient at the current wavelength.

[0026] Based on the importance coefficient of each brown carbon source corresponding to each wavelength, at least one brown carbon source corresponding to each wavelength is determined.

[0027] As one possible implementation, determining the importance coefficient of each brown carbon source corresponding to each wavelength based on multiple brown carbon source decision models for each wavelength includes:

[0028] Based on the decision models of multiple brown carbon sources corresponding to each wavelength, at least one sub-importance coefficient of each brown carbon source corresponding to each wavelength is determined.

[0029] The importance coefficient of each brown carbon source corresponding to each wavelength is determined based on at least one sub-importance coefficient of each brown carbon source corresponding to each wavelength.

[0030] As one possible implementation, determining at least one sub-importance coefficient for each brown carbon source corresponding to each wavelength based on multiple brown carbon source decision models for each wavelength includes:

[0031] The decision models for multiple brown carbon sources corresponding to the second wavelength are iterated. For the current brown carbon source decision model, the aerosol concentration corresponding to each organic aerosol source, the current brown carbon source decision model, and the target brown carbon absorption coefficient corresponding to the second wavelength are input into the pre-trained interpretive machine learning model. The interpretive machine learning model generates the sub-importance coefficients of each organic aerosol source at the second wavelength. The sub-importance coefficients of each organic aerosol source at the second wavelength are used as a sub-importance coefficient of each brown carbon source corresponding to the second wavelength.

[0032] As one possible implementation, determining the absorption spectrum of each brown carbon source based on each wavelength includes:

[0033] Obtain the mass concentration of each brown carbon source corresponding to each wavelength;

[0034] Based on the mass concentration of each brown carbon source corresponding to each wavelength and the brown carbon absorption coefficient corresponding to each wavelength, the mass absorption cross section of each brown carbon source corresponding to each wavelength is calculated.

[0035] The absorption spectrum of each brown carbon source is determined based on the mass absorption cross section of each brown carbon source corresponding to each wavelength.

[0036] Secondly, another embodiment of this application provides a device for determining the light absorption characteristics of atmospheric brown carbon aerosol, the device comprising:

[0037] The acquisition module is used to acquire aerosol observation data and organic aerosol mass spectrometry data. The aerosol observation data includes: filter membrane aerosol absorption coefficients corresponding to multiple wavelengths, correction parameters corresponding to multiple wavelengths, and optical attenuation parameters corresponding to multiple wavelengths. The organic aerosol mass spectrometry data includes: aerosol composition at multiple sampling time points.

[0038] The first determining module is used to determine the brown carbon absorption coefficient corresponding to each wavelength based on the aerosol observation data.

[0039] The second determining module is used to determine multiple organic aerosol sources and the aerosol concentration corresponding to each organic aerosol source based on the organic aerosol mass spectrometry data.

[0040] The training module is used to train multiple brown carbon source decision models based on the brown carbon absorption coefficients corresponding to each wavelength and the aerosol concentrations corresponding to each organic aerosol source, and to determine at least one brown carbon source corresponding to each wavelength based on each brown carbon source decision model.

[0041] The third determining module is used to determine the absorption spectrum of each brown carbon source according to each of the wavelengths corresponding to each of the brown carbon sources. The absorption spectrum of the brown carbon source is used to characterize the change of the light absorption characteristics of the brown carbon source with wavelength.

[0042] As one possible implementation, it also includes: a fourth determining module, used for:

[0043] Based on the absorption spectra of each brown carbon source, the direct radiative forcing corresponding to each brown carbon source is determined. The direct radiative forcing includes: direct radiative forcing of brown carbon aerosols at the top of the atmosphere, direct radiative forcing of brown carbon aerosols at the Earth's surface, and direct radiative forcing of brown carbon aerosols in the atmosphere.

[0044] As one possible implementation, the first determining module is specifically used for:

[0045] Based on the filter membrane aerosol absorption coefficient corresponding to each wavelength, the correction parameters corresponding to each wavelength, the optical attenuation parameters corresponding to each wavelength, and the preset multiple scattering correction parameters, the atmospheric aerosol absorption coefficient corresponding to each wavelength is calculated.

[0046] Based on the atmospheric aerosol absorption coefficient corresponding to the target wavelength and each wavelength, the black carbon absorption coefficient corresponding to each wavelength is calculated.

[0047] The difference between the atmospheric aerosol absorption coefficient and the black carbon absorption coefficient at each wavelength is calculated to obtain the brown carbon absorption coefficient at each wavelength.

[0048] As one possible implementation, the second determining module is specifically used for:

[0049] Based on the organic aerosol mass spectrometry data and the preset positive definite matrix factorization model, the concentration of each organic aerosol component at each time point is obtained, and each organic aerosol component is taken as an organic aerosol source. Based on the concentration of each organic aerosol component at each time point, the aerosol concentration corresponding to each organic aerosol source is obtained.

[0050] As one possible implementation, the training module is specifically used for:

[0051] The brown carbon absorption coefficients corresponding to each wavelength and the aerosol concentrations corresponding to each organic aerosol source are time-aligned to obtain the target brown carbon absorption coefficients corresponding to each wavelength and the target aerosol concentrations corresponding to each organic aerosol source.

[0052] By iterating through each wavelength, for the first wavelength reached, the aerosol concentration corresponding to each organic aerosol source is used as the input variable, and the target brown carbon absorption coefficient corresponding to the first wavelength is used as the prediction variable, thereby training a brown carbon source decision model corresponding to the first wavelength.

[0053] As one possible implementation, the training module is specifically used for:

[0054] Based on the decision models of multiple brown carbon sources corresponding to each wavelength, the importance coefficient of each brown carbon source corresponding to each wavelength is determined. The importance coefficient is used to characterize the degree of influence of each brown carbon source on the brown carbon absorption coefficient at the current wavelength.

[0055] Based on the importance coefficient of each brown carbon source corresponding to each wavelength, at least one brown carbon source corresponding to each wavelength is determined.

[0056] As one possible implementation, the training module is specifically used for:

[0057] Based on the decision models of multiple brown carbon sources corresponding to each wavelength, at least one sub-importance coefficient of each brown carbon source corresponding to each wavelength is determined.

[0058] The importance coefficient of each brown carbon source corresponding to each wavelength is determined based on at least one sub-importance coefficient of each brown carbon source corresponding to each wavelength.

[0059] As one possible implementation, the training module is specifically used for:

[0060] The decision models for multiple brown carbon sources corresponding to the second wavelength are iterated. For the current brown carbon source decision model, the aerosol concentration corresponding to each organic aerosol source, the current brown carbon source decision model, and the target brown carbon absorption coefficient corresponding to the second wavelength are input into the pre-trained interpretive machine learning model. The interpretive machine learning model generates the sub-importance coefficients of each organic aerosol source at the second wavelength. The sub-importance coefficients of each organic aerosol source at the second wavelength are used as a sub-importance coefficient of each brown carbon source corresponding to the second wavelength.

[0061] As one possible implementation, the third determining module is specifically used for:

[0062] Obtain the mass concentration of each brown carbon source corresponding to each wavelength;

[0063] Based on the mass concentration of each brown carbon source corresponding to each wavelength and the brown carbon absorption coefficient corresponding to each wavelength, the mass absorption cross section of each brown carbon source corresponding to each wavelength is calculated.

[0064] The absorption spectrum of each brown carbon source is determined based on the mass absorption cross section of each brown carbon source corresponding to each wavelength.

[0065] Thirdly, another embodiment of this application provides an electronic device, including: a processor, a storage medium, and a bus, wherein the storage medium stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor communicates with the storage medium via the bus, and the processor executes the machine-readable instructions to perform the steps of any of the methods described in the first aspect above.

[0066] Fourthly, another embodiment of this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, performs the steps of any of the methods described in the first aspect above.

[0067] The beneficial effects of this application are as follows: By acquiring aerosol observation data and organic aerosol mass spectrometry data, and determining the brown carbon absorption coefficient corresponding to each wavelength, multiple organic aerosol sources and their corresponding aerosol concentrations can be identified. Based on the brown carbon absorption coefficients and aerosol concentrations corresponding to each wavelength, multiple brown carbon source decision models can be trained. Furthermore, based on each brown carbon source decision model, at least one brown carbon source corresponding to each wavelength can be identified. Thus, based on each brown carbon source corresponding to each wavelength, the absorption spectrum of each brown carbon source can be determined. Starting from these two types of raw observation data (aerosol observation data and organic aerosol mass spectrometry data), and through multi-stage fusion analysis, accurate inversion of each brown carbon source and its absorption spectrum can be achieved. This realizes cross-scale correlation from chemical composition to climate effects, improving the accuracy and scientific rigor of brown carbon climate impact assessment. Attached Figure Description

[0068] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0069] Figure 1 A flowchart of a method for determining the light absorption characteristics of atmospheric brown carbon aerosol provided in an embodiment of this application;

[0070] Figure 2 A flowchart illustrating the determination of the absorption coefficient of brown carbon at each wavelength in the method for determining the light absorption characteristics of atmospheric brown carbon aerosol provided in this application embodiment;

[0071] Figure 3A flowchart illustrating the process of training multiple brown carbon source decision models in the method for determining the light absorption characteristics of atmospheric brown carbon aerosols provided in this application embodiment;

[0072] Figure 4 A flowchart illustrating the process of determining at least one brown carbon source corresponding to each wavelength in the method for determining the light absorption characteristics of atmospheric brown carbon aerosols provided in this application embodiment;

[0073] Figure 5 A flowchart illustrating the determination of the importance coefficient of each brown carbon source corresponding to each wavelength in the method for determining the light absorption characteristics of atmospheric brown carbon aerosols provided in this application embodiment;

[0074] Figure 6 A flowchart illustrating the determination of the absorption spectra of each brown carbon source in the method for determining the light absorption characteristics of atmospheric brown carbon aerosols provided in this application embodiment;

[0075] Figure 7 A schematic diagram of a device for determining the light absorption characteristics of atmospheric brown carbon aerosol provided in an embodiment of this application;

[0076] Figure 8 This is a schematic diagram of the electronic device structure provided in an embodiment of this application. Detailed Implementation

[0077] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. It should be understood that the accompanying drawings in this application are for illustrative and descriptive purposes only and are not intended to limit the scope of protection of this application. Furthermore, it should be understood that the schematic drawings are not drawn to scale. The flowcharts used in this application illustrate operations implemented according to some embodiments of this application. It should be understood that the operations in the flowcharts may not be implemented in sequence, and steps without logical contextual relationships may be reversed or implemented simultaneously. In addition, those skilled in the art, guided by the content of this application, may add one or more other operations to the flowcharts, or remove one or more operations from the flowcharts.

[0078] Furthermore, the described embodiments are merely some, not all, of the embodiments of this application. The components of the embodiments of this application described and illustrated herein can typically be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of the application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0079] It should be noted that the term "comprising" will be used in the embodiments of this application to indicate the presence of the features declared thereafter, but does not exclude the addition of other features.

[0080] Currently, the observation and estimation of the light absorption characteristics of brown carbon mostly rely on statistical regression methods to establish an empirical relationship between source concentration and brown carbon absorption, thereby enabling the identification and contribution assessment of major brown carbon sources.

[0081] However, traditional technical approaches suffer from drawbacks such as low accuracy and difficulty in capturing the dynamic relationship between various source types and light absorption responses under complex nonlinear environmental processes.

[0082] Based on the above-mentioned problems, this application proposes a method for determining the light absorption characteristics of atmospheric brown carbon aerosols. By establishing a nonlinear response relationship between the sources of each organic aerosol and the absorption coefficients of brown carbon at each wavelength, the brown carbon source that dominates brown carbon light absorption is identified, and its wavelength-dependent mass absorption cross section is derived. This achieves cross-scale correlation from chemical composition to climate effects, improving the accuracy and scientific nature of brown carbon climate impact assessment.

[0083] The following describes in detail the method for determining the light absorption characteristics of atmospheric brown carbon aerosol provided in this application, with reference to several embodiments.

[0084] Figure 1 A flowchart illustrating a method for determining the light absorption characteristics of atmospheric brown carbon aerosols provided in this application embodiment is shown below. Figure 1 As shown, the subject executing this method can be any electronic device with processing capabilities, and the method includes:

[0085] S101. Acquire aerosol observation data and organic aerosol mass spectrometry data.

[0086] Optionally, aerosol observation data and organic aerosol mass spectrometry data can be acquired.

[0087] For example, aerosol observation data from stations in typical climate zones such as humid, semi-humid, and semi-arid regions can be measured using the AE33 black carbon meter.

[0088] For example, organic aerosol mass spectrometry data for a preset duration can be observed using an aerosol mass spectrometer.

[0089] The aerosol observation data includes: filter membrane aerosol absorption coefficients for multiple wavelengths. b ATN (λ), correction parameters corresponding to multiple wavelengths k The organic aerosol mass spectrometry data includes the composition of organic aerosols at multiple sampling time points, as well as the optical attenuation parameter ATN corresponding to multiple wavelengths.

[0090] For example, the multiple wavelengths may include 370nm, 470nm, 520nm, 590nm, 660nm, and 880nm. Optionally, the multiple wavelengths may also include 950nm.

[0091] For example, organic aerosol mass spectrometry data can be characterized as a two-dimensional matrix, where each row represents a sampling time point and each column represents an organic ion fragment. The aerosol mass spectrometry data at each sampling time point contains the signal intensity or quantitative concentration of multiple organic ion fragments.

[0092] S102. Based on aerosol observation data, determine the brown carbon absorption coefficient corresponding to each wavelength.

[0093] Optionally, after obtaining aerosol observation data, the absorption coefficient of brown carbon corresponding to each wavelength can be calculated based on the aerosol observation data.

[0094] The absorption coefficient of brown carbon at each wavelength refers to the absorption coefficient of brown carbon at each wavelength.

[0095] For example, after obtaining aerosol observation data, the aerosol observation data can be corrected to eliminate interference such as filter membrane effect and multiple scattering. By using the difference in optical properties between black carbon and brown carbon, the absorption coefficient of brown carbon corresponding to each wavelength can be separated from the corrected aerosol observation data.

[0096] S103. Based on the organic aerosol mass spectrometry data, determine the multiple sources of organic aerosols and the corresponding aerosol concentrations for each source.

[0097] Optionally, the organic aerosol mass spectrometry data can be decomposed to obtain multiple organic aerosol sources and the corresponding aerosol concentrations for each organic aerosol source.

[0098] For example, sources of organic aerosols may include traffic emissions, biomass combustion, and secondary organic aerosols. Aerosol concentration refers to mass concentration.

[0099] S104. Based on the brown carbon absorption coefficient corresponding to each wavelength and the aerosol concentration corresponding to each organic aerosol source, train multiple brown carbon source decision models, and determine at least one brown carbon source corresponding to each wavelength based on each brown carbon source decision model.

[0100] Optionally, after obtaining the brown carbon absorption coefficient corresponding to each wavelength and the aerosol concentration corresponding to each organic aerosol source, the brown carbon absorption coefficient corresponding to each wavelength can be used as the dependent variable, and the aerosol concentration corresponding to each organic aerosol source can be used as the independent variable to train multiple brown carbon source decision models. Through multiple brown carbon source decision models, the mapping relationship between aerosol sources and light absorption bands can be learned respectively, and the nonlinear response relationship between each organic aerosol source and the brown carbon absorption coefficient corresponding to each wavelength can be established, thereby more accurately capturing the dynamic response in complex environmental processes.

[0101] Among them, the brown carbon source decision model can be based on a nonlinear machine learning model, which includes: random forest model, XgBoost model, LightGBM model, gradient boosting decision tree regression model, and adaptive boosting model, etc.

[0102] Optionally, after obtaining multiple brown carbon source decision models, the mapping relationship between aerosol sources and light absorption bands learned by each brown carbon source decision model can be analyzed to determine at least one brown carbon source corresponding to each wavelength.

[0103] Brown carbon source refers to the source of brown carbon, that is, the pollution source that causes brown carbon production.

[0104] For example, the decision models for each brown carbon source can be analyzed using the pre-trained interpretation models to obtain at least one brown carbon source corresponding to each wavelength.

[0105] S105. Determine the absorption spectrum of each brown carbon source based on the wavelength corresponding to each brown carbon source.

[0106] Optionally, after obtaining each brown carbon source corresponding to each wavelength, regression modeling can be performed on each brown carbon source corresponding to each wavelength, and absorption contribution can be assigned to each brown carbon source to obtain the absorption spectrum of each brown carbon source.

[0107] Among them, the absorption spectrum of brown carbon source is used to characterize how the light absorption characteristics of brown carbon source change with wavelength.

[0108] In this embodiment, by acquiring aerosol observation data and organic aerosol mass spectrometry data, and determining the brown carbon absorption coefficient corresponding to each wavelength, multiple organic aerosol sources and their corresponding aerosol concentrations are identified. Based on these brown carbon absorption coefficients and concentrations, multiple brown carbon source decision models can be trained. Each brown carbon source decision model then identifies at least one brown carbon source corresponding to each wavelength. Based on these brown carbon sources, their absorption spectra are determined. This approach, starting from both aerosol observation data and organic aerosol mass spectrometry data, and through multi-stage fusion analysis, ultimately achieves accurate inversion of each brown carbon source and its absorption spectra. This realizes a cross-scale correlation from chemical composition to climate effects, improving the accuracy and scientific rigor of brown carbon climate impact assessment.

[0109] In one possible implementation, Figure 2 A flowchart illustrating the determination of the absorption coefficient of brown carbon at each wavelength in the method for determining the light absorption characteristics of atmospheric brown carbon aerosol provided in this application embodiment is shown below. Figure 2 As shown, in S102 above, based on aerosol observation data, the absorption coefficient of brown carbon corresponding to each wavelength is determined, including:

[0110] S201. Based on the aerosol absorption coefficient of the filter membrane corresponding to each wavelength, the correction parameters corresponding to each wavelength, the optical attenuation parameters corresponding to each wavelength, and the preset multiple scattering correction parameters, the atmospheric aerosol absorption coefficient corresponding to each wavelength is calculated.

[0111] Alternatively, the Virkkula algorithm can be used to extract the filter membrane aerosol absorption coefficients corresponding to multiple wavelengths from aerosol observation data. b ATN (λ) is converted into the atmospheric aerosol absorption coefficient, and the atmospheric aerosol absorption coefficient corresponding to each wavelength is obtained.

[0112] Among them, the atmospheric aerosol absorption coefficients corresponding to each wavelength b abs (λ) refers to the aerosol absorption coefficient in the atmosphere at each wavelength.

[0113] For example, taking a wavelength λ as an example, the atmospheric aerosol absorption coefficient corresponding to the wavelength λ can be calculated according to the following formula (1). b abs (λ):

[0114] (1)

[0115] Where λ is the wavelength. b abs(λ) is the atmospheric aerosol absorption coefficient corresponding to the wavelength λ. b ATN (λ) is the aerosol absorption coefficient of the filter membrane corresponding to the wavelength λ. k Here, ATN is the correction parameter corresponding to the wavelength λ, and ATN is the optical attenuation parameter corresponding to the wavelength λ. C ref These are the correction parameters for the multiple scattering effect. C ref The default value for Teflon filter membranes is 1.57.

[0116] S202. Based on the atmospheric aerosol absorption coefficient corresponding to the target wavelength and each wavelength, the black carbon absorption coefficient corresponding to each wavelength is calculated.

[0117] It is understandable that, given that black carbon aerosols exhibit nearly uniform absorption characteristics across all wavelengths, their absorption wavelength index (AAE) is high. BC Theoretically, the value is 1; however, the absorption range of brown carbon aerosols is usually limited to wavelengths less than 700 nm. They can be separated from atmospheric aerosol absorption parameters by the difference in optical properties between black carbon and brown carbon. Specifically, at 880 nm, only black carbon is considered to contribute. From this, the absorption amount of black carbon at other wavelengths can be inferred, which can avoid systematic errors and improve regional applicability.

[0118] Optionally, the absorption wavelength index (AAE) of black carbon is used. BC The absorption coefficient of black carbon at each wavelength is calculated by setting it to 1 and using the following formula (2). b abs,BC (λ):

[0119]

[0120] (2)

[0121] Where λ is the wavelength, and the target wavelength is 880nm. b abs (880) is the absorption coefficient of black carbon corresponding to a wavelength of 880nm.

[0122] S203. Calculate the difference between the atmospheric aerosol absorption coefficient and the black carbon absorption coefficient corresponding to each wavelength to obtain the brown carbon absorption coefficient corresponding to each wavelength.

[0123] Optionally, after obtaining the black carbon absorption coefficients for each wavelength, the atmospheric aerosol absorption coefficients for each wavelength can be calculated separately. b abs (λ) Black carbon absorption coefficient corresponding to each wavelength b abs,BCThe difference (λ) is used to obtain the brown carbon absorption coefficient corresponding to each wavelength. b abs,BrC (λ), specifically, can be referred to the following formula (3):

[0124] (3)

[0125] Where λ is the wavelength. b abs (λ) is the atmospheric aerosol absorption coefficient corresponding to the wavelength λ. b abs,BrC (λ) is the absorption coefficient of brown carbon corresponding to the wavelength λ. b abs,BC (λ) is the black carbon absorption coefficient corresponding to the wavelength λ.

[0126] By using the aerosol absorption coefficient of the filter membrane corresponding to each wavelength, the correction parameters corresponding to each wavelength, the optical attenuation parameters corresponding to each wavelength, and the preset multiple scattering correction parameters, the atmospheric aerosol absorption coefficient corresponding to each wavelength is calculated. Then, using the atmospheric aerosol absorption coefficient corresponding to the target wavelength and each wavelength, the black carbon absorption coefficient corresponding to each wavelength is calculated. The difference between the atmospheric aerosol absorption coefficient corresponding to each wavelength and the black carbon absorption coefficient corresponding to each wavelength is then calculated to obtain the brown carbon absorption coefficient corresponding to each wavelength. This method can decouple black carbon and brown carbon based on prior physical knowledge, accurately separate the true light absorption signal of brown carbon from the aerosol observation data, and achieve optical decoupling of black carbon and brown carbon without chemical labeling or additional tracers. It avoids the systematic errors caused by the empirical ratio method and improves the physical rationality and spatiotemporal applicability of brown carbon absorption identification.

[0127] In one possible implementation, the step S103 above, which determines multiple organic aerosol sources and the corresponding aerosol concentrations based on organic aerosol mass spectrometry data, includes:

[0128] Based on the organic aerosol mass spectrometry data and the preset positive definite matrix factorization model, the concentration of each organic aerosol component at each time point is obtained. Each organic aerosol component is then treated as an organic aerosol source, and the aerosol concentration corresponding to each organic aerosol source is obtained based on the concentration of each organic aerosol component at each time point.

[0129] Optionally, the concentration of each organic ion fragment at each time point in the organic aerosol mass spectrometry data can be iterated. For the concentration of the current organic ion fragment at the current time point, the concentration of the current organic ion fragment at the current time point can be input into a preset positive matrix factorization (PMF) model to decompose and obtain the concentration of each organic aerosol component at each time point.

[0130] For example, it can be achieved by referring to the following formula (4):

[0131] (4)

[0132] in, org i,j Indicates time points in organic aerosol mass spectrometry data i organic ion fragments j concentration, factor i,p Indicates a point in time i Organic aerosol components p concentration, ions p,j Indicates organic aerosol components p organic ion fragments j concentration, e i,j Indicates a point in time i organic ion fragments j The remaining portion that was not fitted by the model.

[0133] Optionally, after obtaining the concentration of each organic aerosol component at each time point, the concentrations of each organic aerosol component at each time point can be combined to obtain the aerosol concentration corresponding to each organic aerosol source.

[0134] Among them, the aerosol concentration corresponding to the organic aerosol source is the aerosol concentration sequence of the organic aerosol source at each time point.

[0135] For example, an organic aerosol component is used as an organic aerosol source, and the concentrations of the organic aerosol component at various time points are combined to obtain a concentration sequence of the organic aerosol source, which is used as the aerosol concentration corresponding to the organic aerosol source.

[0136] By using a positive definite matrix factorization model, several major pollution sources can be identified from complex organic aerosol mass spectrometry observation data, and the mass concentration change sequences of each pollution source in time and space can be obtained, providing key input variables for subsequent operations, thereby helping to achieve accurate source tracing of brown carbon and climate effect assessment.

[0137] In one possible implementation, Figure 3 A flowchart illustrating the process of training multiple brown carbon source decision models in the method for determining the light absorption characteristics of atmospheric brown carbon aerosols provided in this application embodiment, with reference to... Figure 3 As shown, in S104 above, based on the brown carbon absorption coefficient corresponding to each wavelength and the aerosol concentration corresponding to each organic aerosol source, multiple brown carbon source decision models are trained, including:

[0138] S301. Time alignment processing is performed on the brown carbon absorption coefficient corresponding to each wavelength and the aerosol concentration corresponding to each organic aerosol source to obtain the target brown carbon absorption coefficient corresponding to each wavelength and the target aerosol concentration corresponding to each organic aerosol source.

[0139] Optionally, the brown carbon absorption coefficients corresponding to each wavelength and the aerosol concentrations corresponding to each organic aerosol source are time-aligned to match the brown carbon absorption coefficients corresponding to each wavelength and the aerosol concentrations corresponding to each organic aerosol source to the same time series.

[0140] For example, the target brown carbon absorption coefficients corresponding to each wavelength aligned in the same time series and the target aerosol concentrations corresponding to each organic aerosol source can be obtained by averaging or interpolation.

[0141] S302. Traverse each wavelength. For the first wavelength traversed, take the aerosol concentration corresponding to each organic aerosol source as the input variable and the target brown carbon absorption coefficient corresponding to the first wavelength as the prediction variable to train the brown carbon source decision model corresponding to the first wavelength.

[0142] Optionally, it is possible to iterate through each wavelength, and for the first wavelength reached, the aerosol concentration corresponding to each organic aerosol source is used as the input variable, and the target brown carbon absorption coefficient corresponding to the first wavelength is used as the prediction variable. The brown carbon source decision model corresponding to the first wavelength is trained, thereby realizing the decomposition of light absorption source contribution by wavelength band and breaking through the traditional extensive mode of "overall average attribution".

[0143] For example, the target brown carbon absorption coefficient corresponding to each wavelength and the target aerosol concentration corresponding to each organic aerosol source can be divided into a training set and a test set. The aerosol concentration corresponding to each organic aerosol source in the training set is used as the input variable, and the target brown carbon absorption coefficient corresponding to the first wavelength in the training set is used as the prediction variable. Multiple brown carbon source decision models corresponding to the first wavelength are trained respectively.

[0144] Among them, the decision models for multiple brown carbon sources corresponding to the first wavelength can be based on different nonlinear machine learning models.

[0145] For example, multiple brown carbon source decision models corresponding to the first wavelength can be trained based on five different nonlinear machine learning models.

[0146] For example, the trained decision models for each brown carbon source are applied to the test dataset consisting of the remaining 30% of the data points, and the correlation coefficient (R²) between the observed and predicted values ​​is calculated. 2 The robustness and applicability of each brown carbon source decision model are evaluated by using parameters such as slope, root mean square error (RMSE), and mean absolute error (MAE), thereby obtaining multiple brown carbon source decision models corresponding to the first wavelength.

[0147] Optionally, after the traversal is completed, multiple brown carbon source decision models corresponding to each wavelength are obtained.

[0148] In one possible implementation, Figure 4 A flowchart illustrating the process of determining at least one brown carbon source corresponding to each wavelength in the method for determining the light absorption characteristics of atmospheric brown carbon aerosols provided in this application embodiment is shown below. Figure 4 As shown, in S104 above, at least one brown carbon source corresponding to each wavelength is determined according to the brown carbon source decision model, including:

[0149] S401. Based on the decision models of multiple brown carbon sources corresponding to each wavelength, determine the importance coefficient of each brown carbon source corresponding to each wavelength.

[0150] Optionally, the decision models for each brown carbon source can be analyzed using pre-trained explanatory models to obtain the importance coefficients of each brown carbon source corresponding to each wavelength.

[0151] For example, the aerosol concentration corresponding to each organic aerosol source, the target brown carbon absorption coefficient corresponding to each wavelength, and the brown carbon source decision model can be input into the pre-trained interpretation model for analysis to obtain the importance coefficient of each brown carbon source corresponding to each wavelength.

[0152] The explanatory model can be the SHAP ex post-explanatory machine learning model.

[0153] The importance coefficient is used to characterize the degree of influence of each brown carbon source on the brown carbon absorption coefficient at the current wavelength.

[0154] S402. Based on the importance coefficient of each brown carbon source corresponding to each wavelength, determine at least one brown carbon source corresponding to each wavelength.

[0155] Optionally, taking a wavelength as an example, the organic aerosol sources can be ranked according to the importance coefficient of each brown carbon source corresponding to that wavelength, and the ranking result of each organic aerosol source at that wavelength can be obtained.

[0156] Optionally, after obtaining the ranking results of each organic aerosol source at that wavelength, a proportional vote can be conducted using the Delphi expert method to obtain the voting results. The voting results include the ranking results of each organic aerosol source.

[0157] For example, before voting, it is necessary to evaluate the decision models for each brown carbon source at that wavelength. If a certain brown carbon source decision model differs significantly from other models, it is determined that the brown carbon source decision model is overfitting; if the R-value of a certain brown carbon source decision model is significantly lower than that of other models, it is determined that the brown carbon source decision model is overfitting. 2 If the efficiency of a brown carbon source decision model is significantly lower than or significantly higher than that of other brown carbon source decision models, then the model is deemed ineffective and will not participate in the final voting decision.

[0158] For example, after obtaining the voting results, a preset number of aerosol sources are determined from the voting results according to a preset quantity threshold, and used as multiple brown carbon sources.

[0159] By using multiple brown carbon source decision models corresponding to each wavelength, the importance coefficient of each brown carbon source corresponding to each wavelength is determined. Based on the importance coefficient of each brown carbon source corresponding to each wavelength, at least one brown carbon source corresponding to each wavelength is determined. This can objectively quantify the relative influence of each aerosol source on the light absorption of brown carbon at a specific wavelength, avoiding subjective bias caused by artificially setting weights.

[0160] In one possible implementation, Figure 5 A flowchart illustrating the determination of the importance coefficients of each brown carbon source corresponding to each wavelength in the method for determining the light absorption characteristics of atmospheric brown carbon aerosols provided in this application embodiment, with reference to... Figure 5 As shown, in S401 above, the importance coefficients of each brown carbon source corresponding to each wavelength are determined based on multiple brown carbon source decision models corresponding to each wavelength, including:

[0161] S501. Based on the decision models of multiple brown carbon sources corresponding to each wavelength, determine at least one sub-importance coefficient of each brown carbon source corresponding to each wavelength.

[0162] Optionally, multiple brown carbon source decision models corresponding to each wavelength can be grouped, and at least one sub-importance coefficient of each brown carbon source corresponding to each wavelength can be determined according to the different wavelengths.

[0163] For example, the aerosol concentration corresponding to each organic aerosol source, the target brown carbon absorption coefficient corresponding to each wavelength, and the brown carbon source decision model can be input into the pre-trained interpretation model for analysis to obtain at least one sub-importance coefficient of each brown carbon source corresponding to each wavelength.

[0164] Among them, the sub-importance coefficient is used to characterize the degree of influence of each brown carbon source on the brown carbon absorption coefficient at the current wavelength, as indicated by the current brown carbon source decision model.

[0165] S502. Determine the importance coefficient of each brown carbon source corresponding to each wavelength based on at least one sub-importance coefficient of each brown carbon source corresponding to each wavelength.

[0166] Optionally, after obtaining the sub-importance parameters of each brown carbon source corresponding to each wavelength, the sub-importance parameters of each brown carbon source corresponding to each wavelength can be summed to obtain the importance coefficient of each brown carbon source corresponding to each wavelength.

[0167] In one example, the importance parameters of each brown carbon source corresponding to each wavelength can be summed to obtain the importance coefficient of each brown carbon source corresponding to each wavelength.

[0168] Specifically, taking a wavelength as an example, the importance coefficient of each brown carbon source corresponding to that wavelength can be understood as the global importance of the aerosol concentration corresponding to each organic aerosol source to the absorption coefficient of the target brown carbon at that wavelength.

[0169] By employing multiple brown carbon source decision models corresponding to various wavelengths, at least one sub-importance coefficient for each brown carbon source at each wavelength is determined. Based on this sub-importance coefficient, the overall importance coefficient for each brown carbon source at each wavelength is then determined. By integrating the results of multiple independent models, the random errors caused by the structure or training process of individual models are effectively reduced, enhancing the reliability of the obtained importance coefficients for each brown carbon source at each wavelength. Simultaneously, this approach enables a more comprehensive and multi-layered attribution analysis, more closely approximating the essence of complex nonlinear response relationships in the real atmospheric environment, and contributing to revealing the key source classes and their mechanisms of action that dominate brown carbon light absorption.

[0170] In one possible implementation, step S501 above determines at least one sub-importance coefficient for each brown carbon source corresponding to each wavelength based on multiple brown carbon source decision models for each wavelength, including:

[0171] The decision models for multiple brown carbon sources corresponding to the second wavelength are iterated. For the current brown carbon source decision model, the aerosol concentration corresponding to each organic aerosol source, the current brown carbon source decision model, and the target brown carbon absorption coefficient corresponding to the second wavelength are input into the pre-trained interpretive machine learning model. The interpretive machine learning model generates the sub-importance coefficients of each organic aerosol source under the second wavelength. The sub-importance coefficients of each organic aerosol source under the second wavelength are used as a sub-importance coefficient of each brown carbon source corresponding to the second wavelength.

[0172] Optionally, taking any wavelength as the second wavelength as an example, multiple brown carbon source decision models corresponding to the second wavelength can be traversed. For the current brown carbon source decision model encountered, the aerosol concentration corresponding to each organic aerosol source, the current brown carbon source decision model, and the target brown carbon absorption coefficient corresponding to the second wavelength are input into a pre-trained interpretive machine learning model. The interpretive machine learning model generates sub-importance coefficients for each organic aerosol source at the second wavelength. These sub-importance coefficients are then used as sub-importance coefficients for each brown carbon source corresponding to the second wavelength. This explicitly expresses the nonlinear relationships originally hidden in the model parameters as sub-importance coefficients, making the entire analysis chain scientifically verifiable and logically closed-loop. Simultaneously, it avoids biases caused by a single interpretive path and achieves fine-grained attribution analysis from multiple model perspectives.

[0173] In one possible implementation, Figure 6 A flowchart illustrating the determination of the absorption spectra of various brown carbon sources in the method for determining the light absorption characteristics of atmospheric brown carbon aerosols provided in this application embodiment, with reference to... Figure 6 As shown, in S105 above, the absorption spectrum of each brown carbon source is determined according to each wavelength corresponding to each brown carbon source, including:

[0174] S601. Obtain the mass concentration of each brown carbon source corresponding to each wavelength.

[0175] Optionally, the mass concentration of each brown carbon source corresponding to each wavelength can be obtained from the organic aerosol mass spectrometry data. The mass concentration can be the corresponding aerosol concentration.

[0176] S602. Based on the mass concentration of each brown carbon source corresponding to each wavelength and the brown carbon absorption coefficient corresponding to each wavelength, calculate the mass absorption cross section of each brown carbon source corresponding to each wavelength.

[0177] Optionally, taking a wavelength as an example, at that wavelength, the brown carbon absorption coefficient can be assigned to each brown carbon source using a multiple linear regression model to obtain the mass absorption cross section of each brown carbon source at that wavelength.

[0178] For example, this can be achieved by referring to the following formula (5):

[0179] (5)

[0180] Among them, [BrC i [conc.] indicates the first i The mass concentration of the brown carbon source m i It is the regression coefficient, representing the first... i Mass absorption cross section of a brown carbon source babs,BrC (λ) is the absorption coefficient of brown carbon corresponding to the wavelength λ.

[0181] By calculating the mass absorption cross section of each brown carbon source at each wavelength and the absorption coefficient of brown carbon at each wavelength, the light absorption capacity per unit mass of each source can be determined. This enables the wavelength-dependent light absorption characteristics of brown carbon from different sources to be inferred under real-world conditions without offline sampling or chemical labeling, thus overcoming the limitations of traditional laboratory measurements.

[0182] S603. Determine the absorption spectrum of each brown carbon source based on the mass absorption cross section of each brown carbon source corresponding to each wavelength.

[0183] Optionally, taking a wavelength as an example, after obtaining the mass absorption cross section of each brown carbon source corresponding to that wavelength, the absorption spectrum of each brown carbon source can be constructed based on the mass absorption cross section and the brown carbon absorption coefficient.

[0184] In one possible implementation, the method further includes:

[0185] Based on the absorption spectra of each brown carbon source, the corresponding direct radiative forcing is determined.

[0186] Alternatively, the direct radiative forcing corresponding to each brown carbon source can be determined using the Santa Barbara DISORT atmospheric radiative transfer (SBDART) model.

[0187] Direct radiative forcing includes: direct radiative forcing of brown carbon aerosols at the top of the atmosphere, direct radiative forcing of brown carbon aerosols at the Earth's surface, and direct radiative forcing of brown carbon aerosols in the atmosphere.

[0188] For example, the direct radiative forcing of brown carbon aerosols at the top of the atmosphere (TOA) and the surface (BOA) is the difference in net radiative flux at the surface and top of the atmosphere when brown carbon aerosols are present and when they are not. The direct radiative forcing of brown carbon aerosols at the atmosphere (ATM) is the difference in direct radiative forcing of brown carbon aerosols at the top of the atmosphere and the surface. Specifically, refer to the following formulas (6)-(9):

[0189] (6)

[0190] (7)

[0191] (8)

[0192] (9)

[0193] in,F ↓ and F ↑ represent the downward and upward radiative flux, respectively. F ( a )and F △ represents the net radiative flux with and without brown carbon aerosols, respectively. F TOA , △ F BOA and △ F ATM These are direct radiative forcing of brown carbon aerosols at the top of the atmosphere, the Earth's surface, and the atmosphere, respectively.

[0194] For example, the parameter settings in the SBDART radiative transfer model are constrained using aerosol optical thickness, asymmetry coefficient, single scattering albedo, and MODIS surface albedo data observed concurrently. Based on the typical climate region where the station is located, tropical atmospheric profiles or mid-latitude atmospheric profiles are selected, and cloudless conditions are set.

[0195] Based on this, the absorption spectra of each brown carbon source were used as input variables, and the SBDART radiative transfer model was run once each with and without input variables. Using the results of the two model runs, the net radiative flux at the top of the atmosphere and the surface was calculated according to formula (6) with and without brown carbon aerosols. Then, the brown carbon radiative forcing at the top of the atmosphere, the surface, and the atmosphere was calculated according to formulas (7), (8), and (9), respectively.

[0196] Following the above procedure, by sequentially using the absorption spectrum of each brown carbon source as the input variable, the direct radiative forcing caused by the absorption of each brown carbon source can be quantified, and the direct radiative forcing corresponding to each brown carbon source can be obtained.

[0197] By retrieving absorption spectra from real observation data, the direct radiative forcing corresponding to each brown carbon source can be determined, providing first-hand localized optical parameters for radiative forcing calculations, which greatly enhances the scientific rigor and credibility of climate impact assessments.

[0198] Based on the same inventive concept, this application also provides an apparatus for determining the light absorption characteristics of atmospheric brown carbon aerosol, which corresponds to the method for determining the light absorption characteristics of atmospheric brown carbon aerosol. Since the principle of the apparatus in this application is similar to the method for determining the light absorption characteristics of atmospheric brown carbon aerosol described above, the implementation of the apparatus can refer to the implementation of the method, and the repeated parts will not be described again.

[0199] Reference Figure 7 As shown, Figure 7This is a schematic diagram of a device for determining the light absorption characteristics of atmospheric brown carbon aerosol provided in an embodiment of this application. The device includes: an acquisition module 701, a first determination module 702, a second determination module 703, a training module 704, and a third determination module 705.

[0200] The acquisition module 701 is used to acquire aerosol observation data and organic aerosol mass spectrometry data. The aerosol observation data includes: filter membrane aerosol absorption coefficients corresponding to multiple wavelengths, correction parameters corresponding to multiple wavelengths, and optical attenuation parameters corresponding to multiple wavelengths. The organic aerosol mass spectrometry data includes: organic aerosol composition at multiple sampling time points.

[0201] The first determining module 702 is used to determine the brown carbon absorption coefficient corresponding to each wavelength based on aerosol observation data.

[0202] The second determining module 703 is used to determine multiple organic aerosol sources and the aerosol concentration corresponding to each organic aerosol source based on organic aerosol mass spectrometry data.

[0203] Training module 704 is used to train multiple brown carbon source decision models based on the brown carbon absorption coefficient corresponding to each wavelength and the aerosol concentration corresponding to each organic aerosol source, and to determine at least one brown carbon source corresponding to each wavelength based on each brown carbon source decision model.

[0204] The third determining module 705 is used to determine the absorption spectrum of each brown carbon source according to each wavelength. The absorption spectrum of the brown carbon source is used to characterize the change of the light absorption characteristics of the brown carbon source with wavelength.

[0205] As one possible implementation, it also includes: a fourth determining module, used for:

[0206] Based on the absorption spectra of each brown carbon source, the direct radiative forcing corresponding to each brown carbon source is determined. The direct radiative forcing includes: direct radiative forcing of brown carbon aerosols at the top of the atmosphere, direct radiative forcing of brown carbon aerosols at the Earth's surface, and direct radiative forcing of brown carbon aerosols in the atmosphere.

[0207] As one possible implementation, the first determining module 702 is specifically used for:

[0208] Based on the filter membrane aerosol absorption coefficient corresponding to each wavelength, the correction parameters corresponding to each wavelength, the optical attenuation parameters corresponding to each wavelength, and the preset multiple scattering correction parameters, the atmospheric aerosol absorption coefficient corresponding to each wavelength is calculated.

[0209] Based on the atmospheric aerosol absorption coefficient corresponding to the target wavelength and each wavelength, the black carbon absorption coefficient corresponding to each wavelength is calculated.

[0210] The difference between the atmospheric aerosol absorption coefficient and the black carbon absorption coefficient at each wavelength is calculated to obtain the brown carbon absorption coefficient at each wavelength.

[0211] As one possible implementation, the second determining module 703 is specifically used for:

[0212] Based on the organic aerosol mass spectrometry data and the preset positive definite matrix factorization model, the concentration of each organic aerosol component at each time point is obtained. Each organic aerosol component is then treated as an organic aerosol source, and the aerosol concentration corresponding to each organic aerosol source is obtained based on the concentration of each organic aerosol component at each time point.

[0213] As one possible implementation, training module 704 is specifically used for:

[0214] Time alignment processing was performed on the brown carbon absorption coefficients corresponding to each wavelength and the aerosol concentrations corresponding to each organic aerosol source to obtain the target brown carbon absorption coefficients corresponding to each wavelength and the target aerosol concentrations corresponding to each organic aerosol source.

[0215] By iterating through each wavelength, for the first wavelength reached, the aerosol concentration corresponding to each organic aerosol source is used as the input variable, and the target brown carbon absorption coefficient corresponding to the first wavelength is used as the prediction variable, thus training a brown carbon source decision model corresponding to the first wavelength.

[0216] As one possible implementation, training module 704 is specifically used for:

[0217] Based on the decision models of multiple brown carbon sources corresponding to each wavelength, the importance coefficient of each brown carbon source corresponding to each wavelength is determined. The importance coefficient is used to characterize the degree of influence of each brown carbon source on the brown carbon absorption coefficient at the current wavelength.

[0218] Based on the importance coefficient of each brown carbon source corresponding to each wavelength, at least one brown carbon source corresponding to each wavelength is determined.

[0219] As one possible implementation, training module 704 is specifically used for:

[0220] Based on the decision models of multiple brown carbon sources corresponding to each wavelength, at least one sub-importance coefficient of each brown carbon source corresponding to each wavelength is determined.

[0221] The importance coefficient of each brown carbon source corresponding to each wavelength is determined based on at least one sub-importance coefficient of each brown carbon source corresponding to each wavelength.

[0222] As one possible implementation, training module 704 is specifically used for:

[0223] The decision models for multiple brown carbon sources corresponding to the second wavelength are iterated. For the current brown carbon source decision model, the aerosol concentration corresponding to each organic aerosol source, the current brown carbon source decision model, and the target brown carbon absorption coefficient corresponding to the second wavelength are input into the pre-trained interpretive machine learning model. The interpretive machine learning model generates the sub-importance coefficients of each organic aerosol source under the second wavelength. The sub-importance coefficients of each organic aerosol source under the second wavelength are used as a sub-importance coefficient of each brown carbon source corresponding to the second wavelength.

[0224] As one possible implementation, the third determining module 705 is specifically used for:

[0225] Obtain the mass concentration of each brown carbon source corresponding to each wavelength;

[0226] Based on the mass concentration of each brown carbon source corresponding to each wavelength and the brown carbon absorption coefficient corresponding to each wavelength, the mass absorption cross section of each brown carbon source corresponding to each wavelength is calculated.

[0227] The absorption spectrum of each brown carbon source is determined based on the mass absorption cross section of each brown carbon source corresponding to each wavelength.

[0228] The processing flow of each module in the device and the interaction flow between each module can be referred to the relevant descriptions in the above method embodiments, and will not be detailed here.

[0229] This application also provides an electronic device, such as... Figure 8 As shown, Figure 8 The schematic diagram of the electronic device structure provided in this application embodiment includes: a processor 801 and a memory 802, and optionally, a bus 803. The memory 802 stores machine-readable instructions executable by the processor 801. When the electronic device is running, the processor 801 and the memory 802 communicate through the bus 803, and the processor 801 executes the machine-readable instructions to perform the steps of the above-described method for determining the light absorption characteristics of atmospheric brown carbon aerosol.

[0230] This application also provides a computer-readable storage medium storing a computer executing the above-described program, wherein the computer program is run by a processor to determine the light absorption characteristics of atmospheric brown carbon aerosol.

[0231] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems and devices described above can be referred to the corresponding processes in the method embodiments, and will not be repeated here. In the several embodiments provided in this application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. Furthermore, multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed mutual coupling or direct coupling or communication connection can be through some communication interfaces; the indirect coupling or communication connection of devices or modules can be electrical, mechanical, or other forms.

[0232] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. If the functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes: USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, optical disks, and other media capable of storing program code.

[0233] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for determining the light absorption properties of atmospheric brown carbon aerosols, characterized in that, The method comprises the following steps: acquiring aerosol observation data and organic aerosol mass spectrum data, the aerosol observation data comprising filter membrane aerosol absorption coefficients corresponding to a plurality of wavelengths, correction parameters corresponding to a plurality of wavelengths and optical attenuation parameters corresponding to a plurality of wavelengths, and the organic aerosol mass spectrum data comprising organic aerosol compositions at a plurality of sampling time points; determining brown carbon absorption coefficients corresponding to each wavelength according to the aerosol observation data; determining a plurality of organic aerosol sources and aerosol concentrations corresponding to each organic aerosol source according to the organic aerosol mass spectrum data; training a plurality of brown carbon source decision models according to the brown carbon absorption coefficients corresponding to each wavelength and the aerosol concentrations corresponding to each organic aerosol source, and determining at least one brown carbon source corresponding to each wavelength according to each brown carbon source decision model; determining absorption spectra of each brown carbon source according to each brown carbon source corresponding to each wavelength, wherein the absorption spectra of the brown carbon source are used to represent the change of light absorption characteristics of the brown carbon source with wavelength.

2. The method according to claim 1, wherein The method further comprises the following steps: determining direct radiation forcing corresponding to each brown carbon source according to the absorption spectra of each brown carbon source, wherein the direct radiation forcing comprises brown carbon aerosol direct radiation forcing at the top of the atmosphere, brown carbon aerosol direct radiation forcing at the ground and brown carbon aerosol direct radiation forcing in the atmosphere.

3. The method according to claim 1, wherein The step of determining brown carbon absorption coefficients corresponding to each wavelength according to the aerosol observation data comprises the following steps: calculating atmospheric aerosol absorption coefficients corresponding to each wavelength according to the filter membrane aerosol absorption coefficients corresponding to each wavelength, the correction parameters corresponding to each wavelength, the optical attenuation parameters corresponding to each wavelength and preset multiple scattering correction parameters; calculating black carbon absorption coefficients corresponding to each wavelength according to the atmospheric aerosol absorption coefficients corresponding to the target wavelength and each wavelength; calculating the difference between the atmospheric aerosol absorption coefficients corresponding to each wavelength and the black carbon absorption coefficients corresponding to each wavelength to obtain the brown carbon absorption coefficients corresponding to each wavelength.

4. The method according to claim 1, wherein The step of determining a plurality of organic aerosol sources and aerosol concentrations corresponding to each organic aerosol source according to the organic aerosol mass spectrum data comprises the following steps: decomposing the organic aerosol mass spectrum data and a preset positive definite matrix factorization model to obtain the concentrations of each organic aerosol component at each time point, taking each organic aerosol component as an organic aerosol source, and obtaining the aerosol concentrations corresponding to each organic aerosol source according to the concentrations of each organic aerosol component at each time point.

5. The method of claim 1, wherein the method is a method for determining atmospheric brown carbon aerosol absorption optical properties, and The step of training a plurality of brown carbon source decision models according to the brown carbon absorption coefficients corresponding to each wavelength and the aerosol concentrations corresponding to each organic aerosol source comprises the following steps: performing time alignment processing on the brown carbon absorption coefficients corresponding to each wavelength and the aerosol concentrations corresponding to each organic aerosol source to obtain target brown carbon absorption coefficients corresponding to each wavelength and target aerosol concentrations corresponding to each organic aerosol source; By iterating through each wavelength, for the first wavelength reached, the aerosol concentration corresponding to each organic aerosol source is used as the input variable, and the target brown carbon absorption coefficient corresponding to the first wavelength is used as the prediction variable, thereby training a brown carbon source decision model corresponding to the first wavelength.

6. The method according to claim 5, wherein The step of determining at least one brown carbon source corresponding to each wavelength based on each brown carbon source decision model includes: Based on the decision models of multiple brown carbon sources corresponding to each wavelength, the importance coefficient of each brown carbon source corresponding to each wavelength is determined. The importance coefficient is used to characterize the degree of influence of each brown carbon source on the brown carbon absorption coefficient at the current wavelength. Based on the importance coefficient of each brown carbon source corresponding to each wavelength, at least one brown carbon source corresponding to each wavelength is determined.

7. The method according to claim 6, wherein The step of determining the importance coefficient of each brown carbon source corresponding to each wavelength based on the decision model of multiple brown carbon sources corresponding to each wavelength includes: Based on the decision models of multiple brown carbon sources corresponding to each wavelength, at least one sub-importance coefficient of each brown carbon source corresponding to each wavelength is determined. The importance coefficient of each brown carbon source corresponding to each wavelength is determined based on at least one sub-importance coefficient of each brown carbon source corresponding to each wavelength.

8. The method according to claim 7, wherein The step of determining at least one sub-importance coefficient for each brown carbon source corresponding to each wavelength based on multiple brown carbon source decision models for each wavelength includes: The decision models for multiple brown carbon sources corresponding to the second wavelength are iterated. For the current brown carbon source decision model, the aerosol concentration corresponding to each organic aerosol source, the current brown carbon source decision model, and the target brown carbon absorption coefficient corresponding to the second wavelength are input into the pre-trained interpretive machine learning model. The interpretive machine learning model generates the sub-importance coefficients of each organic aerosol source at the second wavelength. The sub-importance coefficients of each organic aerosol source at the second wavelength are used as a sub-importance coefficient of each brown carbon source corresponding to the second wavelength.

9. The method of claim 1, wherein the method is a method for determining atmospheric brown carbon aerosol absorption optical properties, and Determining the absorption spectrum of each brown carbon source based on each wavelength includes: Obtain the mass concentration of each brown carbon source corresponding to each wavelength; Based on the mass concentration of each brown carbon source corresponding to each wavelength and the brown carbon absorption coefficient corresponding to each wavelength, the mass absorption cross section of each brown carbon source corresponding to each wavelength is calculated. The absorption spectrum of each brown carbon source is determined based on the mass absorption cross section of each brown carbon source corresponding to each wavelength.

10. An electronic device, comprising: include: The device includes a processor and a memory, the memory storing machine-readable instructions executable by the processor, which, when the electronic device is in operation, are executed by the processor to perform the steps of the method for determining the light absorption characteristics of atmospheric brown carbon aerosol as described in any one of claims 1 to 9.

Images