VOCs concentration estimation method, device, equipment and medium

Abstract

The invention relates to a VOCs concentration estimation method, device, equipment and medium, and belongs to the technical field of environmental monitoring, the method comprises the following steps: obtaining column concentration data of pollution gas monitored by a satellite in a target area, profile data of the pollution gas and influence factor data; calculating first resolution near-surface pollution gas concentration data according to the column concentration data and the profile data of the pollution gas, and performing resampling processing on the influence factor data; the first resolution near-surface pollution gas concentration data and the processed influence factor data are input into a downscaling model, second resolution near-surface pollution gas concentration data are output, and the second resolution is larger than the first resolution; and inputting the second resolution near-surface pollution gas concentration data into a VOCs concentration estimation model, and outputting VOCs concentration data of the target area. According to the method, an implementation way is provided for acquiring large-range continuous spatial and temporal distribution information of the VOCs.

Description

Technical Field

[0001] The present invention belongs to the technical field of environmental monitoring, and in particular relates to a VOCs concentration estimation method, device, equipment and medium. Background Art

[0002] As the control of pollutants such as particulate matter, sulfur dioxide, and nitrogen oxides has become increasingly effective, pollution concentrations have dropped significantly, and the ozone problem has become increasingly prominent. The control of ozone and its precursor VOCs has gradually become the focus of current air management. It is very important to timely grasp the distribution and changing trends of VOCs concentrations for air pollution prevention and control.

[0003] In the existing technology, VOCs are mainly monitored by ground equipment, which can monitor the continuous hourly changes of multiple species of VOCs, but it also has certain limitations. On the one hand, the monitoring results only represent the concentration in a small range around the site. On the other hand, the monitoring equipment exists in a small number of ground stations and field experiments, and it is impossible to fully understand the state of air pollution. In comparison, satellite remote sensing can obtain continuous atmospheric distribution information and is an important means of monitoring atmospheric pollutants. However, satellite monitoring also has problems such as coarse temporal and spatial resolution, difficulty in achieving fine-grained supervision of small-scale cities, and inability to directly monitor VOCs. Therefore, there is an urgent need for a VOCs concentration estimation method that can combine multi-source data. Summary of the invention

[0004] In view of the above analysis, the embodiments of the present invention aim to provide a VOCs concentration estimation method, device, equipment and medium to solve the problem that VOCs cannot be continuously monitored over a large range in the prior art.

[0005] The object of the present invention is achieved in that: A first aspect of the present invention provides a VOCs concentration estimation method, comprising: Obtaining column concentration data, profile data, and influencing factor data of pollutant gases monitored by satellite in the target area; Calculate first-resolution near-surface polluted gas concentration data based on the column concentration data and profile data of the polluted gas, and resample the influencing factor data; Inputting the first-resolution near-surface polluted gas concentration data and the processed influencing factor data into a downscaling model, and outputting the second-resolution near-surface polluted gas concentration data, wherein the second resolution is greater than the first resolution; The second-resolution near-ground polluted gas concentration data is input into a VOCs concentration estimation model to output VOCs concentration data of a target area.

[0006] Further, the calculation of first-resolution near-ground polluted gas concentration data based on the column concentration data and profile data of the polluted gas includes: The column concentration data is projected into raster data, and the first-resolution near-surface pollutant gas concentration data is calculated based on the raster data and profile data, which is expressed as follows: in, Represents the grid position as i The profile height is j The concentration of pollutant gases, Represents the grid position as i The pollutant gas column concentration, Indicates the type of pollutant gas, including formaldehyde, glyoxal and nitrogen dioxide, represents the concentration integral of the entire profile, Indicates ground to H Profile concentration integral with height.

[0007] Furthermore, the resampling of the influencing factor data includes: using the Pearson correlation coefficient method to screen the key influencing factor data that affects the concentration of polluted gases in the target area from the influencing factor data, and resampling the key influencing factors using different sampling resolutions to obtain first resolution raster data and second resolution raster data.

[0008] Furthermore, the step of inputting the first-resolution near-surface pollutant gas concentration data and the processed influencing factor data into the downscaling model and outputting the second-resolution near-surface pollutant gas concentration data comprises: The trend surface function is used to predict the first-resolution near-surface pollutant gas concentration value based on the first-resolution raster data, and the first-resolution residual between the first-resolution near-surface pollutant gas concentration prediction value and the first-resolution near-surface pollutant gas concentration data is calculated and expressed as: in, represents the predicted value of the first-resolution near-surface pollutant gas concentration, represents the trend surface function, Represents the key influencing factor data in the first resolution raster data, represents the first resolution residual, Indicates the first-resolution near-surface pollutant gas concentration data; The trend surface function is used to predict the second-resolution near-surface pollutant gas concentration value based on the second-resolution raster data, and the first-resolution residual is interpolated to obtain the second-resolution residual. The second-resolution near-surface pollutant gas concentration data is calculated based on the second-resolution near-surface pollutant gas concentration prediction value and the second-resolution residual, which is expressed as: in, represents the predicted value of the second-resolution near-surface pollutant gas concentration, Represents the key influencing factor data in the second resolution raster data, Represents the second-resolution near-surface pollutant gas concentration data, Represents the second resolution residual.

[0009] Furthermore, the trend surface function is obtained by training a random forest model, and the training process includes: collecting near-surface pollutant gas concentration data and influencing factor grid data of the target area as a first training set; training the random forest model with the influencing factor grid data as model input and the near-surface pollutant gas concentration data as model output; The atmospheric diffusion equation is used as the loss function of the random forest model, which is expressed as follows: in, Indicates the concentration data of polluted gases near the ground. u represents the vector wind speed, D represents the diffusion coefficient, Represents the spatial gradient of the near-ground pollutant gas concentration data, Laplacian operator representing the near-surface pollutant gas concentration data.

[0010] Furthermore, it also includes dynamically zoning the target area according to the second resolution near-ground polluted gas concentration data, specifically including: calculating the ratio R of formaldehyde concentration to nitrogen dioxide concentration in the second resolution near-ground polluted gas concentration data, setting the area in the target area where R is less than the first threshold as the artificial source influence area, the area where R is greater than or equal to the first threshold and less than the second threshold as the transition area, and the area where R is greater than or equal to the second threshold as the natural source influence area.

[0011] Furthermore, the training process of the VOCs concentration estimation model includes: collecting historical data of VOCs concentration monitored by ground stations in the target area, and forming a second training set with the historical data of near-ground polluted gas concentration; dividing the second training set into three groups according to dynamic partitioning rules, and each group is trained separately with the historical data of near-ground polluted gas concentration and related influencing factor data as input and the historical data of VOCs concentration as output, to obtain three VOCs concentration estimation models, wherein the near-ground polluted gas includes at least formaldehyde, and the related influencing factors include at least temperature, humidity and normalized vegetation index.

[0012] A second aspect of the present invention provides a VOCs concentration estimation device, comprising: A data acquisition module is used to obtain column concentration data, profile data of polluted gases, and influencing factor data monitored by satellite in the target area; A data processing module, used to calculate first-resolution near-surface polluted gas concentration data according to the column concentration data and profile data of the polluted gas, and to resample the influencing factor data; A downscaling module, used for inputting the first-resolution near-surface polluted gas concentration data and the processed influencing factor data into a downscaling model, and outputting the second-resolution near-surface polluted gas concentration data, wherein the second resolution is greater than the first resolution; The concentration estimation module is used to input the second-resolution near-ground polluted gas concentration data into the VOCs concentration estimation model and output the VOCs concentration data of the target area.

[0013] A third aspect of the present invention provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the VOCs concentration estimation method described in any embodiment is implemented.

[0014] A fourth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the VOCs concentration estimation method described in any one of the embodiments.

[0015] Compared with the prior art, the present invention can achieve at least one of the following beneficial effects: The VOCs concentration estimation method provided by the present invention calculates the near-surface polluted gas concentration data with lower resolution by combining the column concentration data monitored by satellite and the profile data monitored by ground monitoring stations, and then uses a downscaling model to obtain the near-surface polluted gas concentration data with higher resolution based on the near-surface polluted gas concentration data with lower resolution and influencing factor data, and then uses the VOCs concentration estimation model to convert the near-surface polluted gas concentration data with higher resolution into spatially continuous VOCs concentration data, thereby overcoming the shortcomings of relying solely on ground station monitoring or satellite monitoring, realizing relatively comprehensive atmospheric pollution status monitoring, and providing a way to obtain large-scale continuous VOCs spatiotemporal distribution information. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the embodiments of this specification or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the embodiments of this specification. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0017] Figure 1A flow chart of a VOCs concentration estimation method provided in Example 1 of the present invention; Figure 2 A schematic diagram of a VOCs concentration estimation device provided in Example 2 of the present invention; Figure 3 This is a schematic diagram of the electronic device architecture provided in Example 3 of the present invention. DETAILED DESCRIPTION

[0018] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. It should be noted that, in the absence of conflict, the embodiments and features in the embodiments disclosed in this disclosure can be combined, separated, interchanged and / or rearranged with each other. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0019] Example 1 A specific embodiment of the present invention, as Figure 1 As shown, a VOCs concentration estimation method is disclosed, comprising the following steps: S1. Obtaining column concentration data, profile data, and influencing factor data of pollutant gases monitored by satellite in the target area; Specifically, pollutant gases include formaldehyde, glyoxal and nitrogen dioxide, and their hourly column concentration data are obtained through geostationary satellites such as Himawari and GeoKompsat 2B; pollutant gas concentration profile data are obtained through ground monitoring stations or atmospheric chemistry numerical model simulations; influencing factor data are human activities and natural environment influencing factor data, including but not limited to population density, road network density, night light index, normalized vegetation index, surface reflectivity, geographic elevation, surface type data, temperature, relative humidity, air pressure, wind speed, wind direction, and boundary layer height.

[0020] S2. Calculate first-resolution near-surface polluted gas concentration data according to the column concentration data and profile data of the polluted gas, and resample the influencing factor data; In this embodiment, step S2 specifically includes: S201, projecting the column concentration data into grid data, and calculating the first resolution near-surface pollutant gas concentration data according to the grid data and the profile data, as shown below: in, Represents the grid position as i The profile height isj The concentration of pollutant gases, Represents the grid position as i The pollutant gas column concentration, Indicates the type of pollutant gas, including formaldehyde, glyoxal and nitrogen dioxide, represents the concentration integral of the entire profile, Indicates ground to H Profile concentration integral with height.

[0021] S202, using the Pearson correlation coefficient method to screen key influencing factor data that affect the concentration of polluted gases in the target area from the influencing factor data, and resampling the key influencing factors using different sampling resolutions to obtain first resolution raster data and second resolution raster data.

[0022] For example, the key influencing factors selected by the Pearson correlation coefficient method are temperature, pressure, humidity, light index, normalized difference vegetation index, surface type and road network density. Then, the key influencing factors are sampled at different resolutions by least squares interpolation, and the first resolution raster data (such as 7km×8km) with a resolution equal to that of the column concentration data, and the second resolution raster data (such as 1km×1km) with a resolution greater than that of the column concentration data are obtained. Among them, meteorological factor data can be obtained through the China Meteorological Data Network; normalized difference vegetation index data can be obtained by using medium resolution imaging spectrometer data, and the maximum composite method is used to generate monthly products; night light index data can be obtained by using VIIRS / DNB night lighting products, including continuous or low-intensity lights from urban lights, small-scale residential areas and traffic flow; surface type data can use the global geographic information public product GlobalLand30 to set numbers for different surface types; road network density data can use the road network density dataset from the Earth System Science Data Center.

[0023] S3, inputting the first-resolution near-surface polluted gas concentration data and the processed influencing factor data into a downscaling model, and outputting the second-resolution near-surface polluted gas concentration data, wherein the second resolution is greater than the first resolution; In this embodiment, step S3 specifically includes: S301, using a trend surface function to predict the first resolution near-surface pollutant gas concentration value according to the first resolution raster data, and calculating the first resolution residual between the first resolution near-surface pollutant gas concentration prediction value and the first resolution near-surface pollutant gas concentration data, expressed as: in, represents the predicted value of the first-resolution near-surface pollutant gas concentration, represents the trend surface function, Represents the key influencing factor data in the first resolution raster data, represents the first resolution residual, Represents the first-resolution near-ground pollutant gas concentration data.

[0024] S302, using a trend surface function to predict the second resolution near-surface pollutant gas concentration value according to the second resolution raster data, interpolating the first resolution residual to obtain the second resolution residual, and calculating the second resolution near-surface pollutant gas concentration data according to the second resolution near-surface pollutant gas concentration prediction value and the second resolution residual, expressed as: in, represents the predicted value of the second-resolution near-surface pollutant gas concentration, Represents the key influencing factor data in the second resolution raster data, Represents the second-resolution near-surface pollutant gas concentration data, Represents the second resolution residual.

[0025] In this embodiment, the trend surface function is obtained by training a random forest model, and the training process includes: (I) Collecting the ground pollutant gas concentration data and influencing factor grid data of the target area as the first training set; (II) The random forest model is trained using the impact factor grid data as the model input and the near-surface pollutant gas concentration data as the model output; The atmospheric diffusion equation is used as the loss function of the random forest model, which is expressed as follows: in, Indicates the concentration data of polluted gases near the ground. u represents the vector wind speed, D represents the diffusion coefficient, Represents the spatial gradient of the near-ground pollutant gas concentration data, Laplacian operator representing the near-surface pollutant gas concentration data.

[0026] In some embodiments, between steps S3 and S4, the process further includes dynamically partitioning the target area according to the second resolution near-ground polluted gas concentration data, specifically including: The ratio R of formaldehyde concentration to nitrogen dioxide concentration in the second resolution near-ground polluted gas concentration data is calculated, and the area in the target area where R is less than the first threshold is set as the anthropogenic source influence area, the area where R is greater than or equal to the first threshold and less than the second threshold is set as the transition area, and the area where R is greater than or equal to the second threshold is set as the natural source influence area.

[0027] Specifically, the study area was divided into human activity areas and areas with high vegetation coverage such as forests and grasslands, combining the surface type, night light index, and normalized difference vegetation index. The areas with night light index greater than 1 were regarded as human activity areas, and the areas with NDVI greater than 0.4 were regarded as areas with high vegetation coverage. The m% and n% quantiles of the R values ​​of the two types were calculated (Q 1 m,Q 2 n). m and n can be set to 90 and 10 respectively, then the initial threshold is set to: A1=Q 1 90. A2=Q 2 10; A1 is the first threshold and A2 is the second threshold. Use a sliding window (window size 3km) to spatially smooth the thresholds to eliminate local outlier interference. The recommended range for A1 is 1-3.5, and the recommended range for A2 is 3-7.

[0028] If the nitrogen dioxide data is missing, the target area can also be divided according to the ratio R2 of glyoxal concentration to formaldehyde concentration in the second-resolution near-ground pollutant gas concentration data, and the dynamic thresholds B1 and B2 can be set in combination with the surface type data. The area with R2<B1 is the main impact area of ​​anthropogenic sources, the area with R2≥B2 is the main impact area of ​​natural sources, and the area with B1≤R2<B2 is the transition area.

[0029] Compared with the method of zoning based solely on surface type, the ratio method can more comprehensively consider the key variable of atmospheric movement. As an important factor affecting the distribution and transmission of pollutants, the dynamic changes of atmospheric movement are often difficult to accurately capture through static surface type data. By calculating the concentration ratio of specific pollutants, the source and diffusion of pollutants in the atmosphere can be indirectly reflected, providing a better zoning basis for subsequent calculations.

[0030] S4. Input the second-resolution near-ground polluted gas concentration data into a VOCs concentration estimation model, and output VOCs concentration data of the target area.

[0031] Specifically, existing studies have shown that the concentrations of formaldehyde, glyoxal, etc. are linearly related to the concentration of VOCs. Therefore, it is possible to estimate the spatially continuous concentration of VOCs based on the discrete near-ground formaldehyde and glyoxal concentrations. This can be achieved, for example, using a multivariate linear regression equation or a random forest model. This embodiment does not limit the specific structure of the VOCs concentration estimation model and can be flexibly applied according to the data scale.

[0032] Compared with the prior art, the VOCs concentration estimation method provided in this embodiment calculates the near-surface polluted gas concentration data with lower resolution by combining the column concentration data monitored by satellite and the profile data monitored by ground monitoring stations, and then uses the downscaling model to obtain the near-surface polluted gas concentration data with higher resolution based on the near-surface polluted gas concentration data with lower resolution and the influencing factor data, and then uses the VOCs concentration estimation model to convert the near-surface polluted gas concentration data with higher resolution into spatially continuous VOCs concentration data, thereby overcoming the shortcomings of relying solely on ground station monitoring or satellite monitoring, realizing relatively comprehensive atmospheric pollution status monitoring, and providing a way to obtain large-scale continuous VOCs spatiotemporal distribution information.

[0033] In this embodiment, the training process of the VOCs concentration estimation model includes: Collect the measured VOCs concentration data of the target area in the specific historical period, and form a second training set with the second resolution near-ground polluted gas concentration data in the first training set; The second training set is divided into three groups according to the dynamic partitioning rules. Each group is trained separately with the second resolution near-ground polluted gas concentration data and related influencing factors as model input and the measured VOCs concentration data as model output, and three VOCs concentration estimation models are obtained. The near-ground polluted gas includes at least formaldehyde, and the related influencing factors include at least temperature, humidity and normalized vegetation index.

[0034] Specifically, temperature is the most important factor affecting the relationship between VOCs and formaldehyde. During training, the data in each partition is further grouped according to the ambient temperature (such as 7°C as a step). If the number of samples in the group is less than a certain threshold, the multivariate linear regression model is trained as the VOCs concentration estimation model. Otherwise, the random forest regression model is trained as the VOCs concentration estimation model. Considering that there are usually fewer ground stations and the transition zone often lacks ground monitoring data, a weighted method can be used to obtain a concentration estimation model.

[0035] Example 2 This embodiment provides a VOCs concentration estimation device, such as Figure 2 As shown, including: A data acquisition module is used to obtain column concentration data, profile data of polluted gases, and influencing factor data monitored by satellite in the target area; A data processing module, used to calculate first-resolution near-surface polluted gas concentration data according to the column concentration data and profile data of the polluted gas, and to resample the influencing factor data; A downscaling module, used for inputting the first-resolution near-surface polluted gas concentration data and the processed influencing factor data into a downscaling model, and outputting the second-resolution near-surface polluted gas concentration data, wherein the second resolution is greater than the first resolution; The concentration estimation module is used to input the second-resolution near-ground polluted gas concentration data into the VOCs concentration estimation model and output the VOCs concentration data of the target area.

[0036] Example 3 This embodiment provides an electronic device, such as Figure 3 As shown, it includes a memory and a processor, the memory stores a computer program, and when the computer program is executed by the processor, the VOCs concentration estimation method described in any of the above embodiments is implemented.

[0037] Example 4 This embodiment provides a computer-readable storage medium having a computer program stored thereon. When the program is executed by a processor, the VOCs concentration estimation method described in any of the above embodiments is implemented.

[0038] Computer readable storage media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. 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 disk read-only memory (CD-ROM), digital versatile disk (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 temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0039] The professionals should further realize that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to the function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0040] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be implemented using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0041] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of 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 estimating VOCs concentration, characterized in that: include: Obtaining column concentration data, profile data, and influencing factor data of pollutant gases monitored by satellite in the target area; Calculate first-resolution near-surface polluted gas concentration data based on the column concentration data and profile data of the polluted gas, and resample the influencing factor data; Inputting the first-resolution near-surface polluted gas concentration data and the processed influencing factor data into a downscaling model, and outputting the second-resolution near-surface polluted gas concentration data, wherein the second resolution is greater than the first resolution; The second-resolution near-ground polluted gas concentration data is input into a VOCs concentration estimation model to output VOCs concentration data of a target area.

2. The VOCs concentration estimation method according to claim 1, characterized in that: The method of calculating first-resolution near-ground polluted gas concentration data according to the column concentration data and profile data of the polluted gas comprises: The column concentration data is projected into raster data, and the first-resolution near-surface pollutant gas concentration data is calculated based on the raster data and profile data, which is expressed as follows: in, Represents the grid position as i The profile height is j The concentration of pollutant gases, Represents the grid position as i The pollutant gas column concentration, Indicates the type of pollutant gas, including formaldehyde, glyoxal and nitrogen dioxide, represents the concentration integral of the entire profile, Indicates ground to H Profile concentration integral with height.

3. The VOCs concentration estimation method according to claim 1, characterized in that: The resampling process of the impact factor data includes: The Pearson correlation coefficient method is used to screen the key influencing factor data that affects the concentration of polluted gases in the target area from the influencing factor data, and the key influencing factors are resampled using different sampling resolutions to obtain first resolution raster data and second resolution raster data.

4. The VOCs concentration estimation method according to claim 3, characterized in that: The step of inputting the first-resolution near-surface polluted gas concentration data and the processed influencing factor data into the downscaling model and outputting the second-resolution near-surface polluted gas concentration data comprises: The trend surface function is used to predict the first-resolution near-surface pollutant gas concentration value based on the first-resolution raster data, and the first-resolution residual between the first-resolution near-surface pollutant gas concentration prediction value and the first-resolution near-surface pollutant gas concentration data is calculated and expressed as: in, represents the predicted value of the first-resolution near-surface pollutant gas concentration, represents the trend surface function, Represents the key influencing factor data in the first resolution raster data, represents the first resolution residual, Indicates the first-resolution near-surface pollutant gas concentration data; The trend surface function is used to predict the second-resolution near-surface pollutant gas concentration value based on the second-resolution raster data, the first-resolution residual is interpolated to obtain the second-resolution residual, and the second-resolution near-surface pollutant gas concentration data is calculated based on the second-resolution near-surface pollutant gas concentration prediction value and the second-resolution residual, which is expressed as: in, represents the predicted value of the second-resolution near-surface pollutant gas concentration, Represents the key influencing factor data in the second resolution raster data, Represents the second-resolution near-surface pollutant gas concentration data, Represents the second resolution residual.

5. The VOCs concentration estimation method according to claim 4, characterized in that: The trend surface function is obtained by training a random forest model, and the training process includes: Collect the near-ground pollutant gas concentration data and influencing factor grid data of the target area as the first training set; The random forest model is trained using the influencing factor grid data as the model input and the near-surface pollutant gas concentration data as the model output; The atmospheric diffusion equation is used as the loss function of the random forest model, which is expressed as follows: in, Indicates the concentration data of polluted gases near the ground. u represents the vector wind speed, D represents the diffusion coefficient, Represents the spatial gradient of the near-ground pollutant gas concentration data, Laplacian operator representing the near-surface pollutant gas concentration data.

6. The VOCs concentration estimation method according to any one of claims 1 to 5, characterized in that: It also includes dynamic zoning of the target area based on the second-resolution near-surface polluted gas concentration data, including: The ratio R of formaldehyde concentration to nitrogen dioxide concentration in the second resolution near-ground polluted gas concentration data is calculated, and the area in the target area where R is less than the first threshold is set as the anthropogenic source influence area, the area where R is greater than or equal to the first threshold and less than the second threshold is set as the transition area, and the area where R is greater than or equal to the second threshold is set as the natural source influence area.

7. The VOCs concentration estimation method according to claim 6, characterized in that: The training process of the VOCs concentration estimation model includes: Collect historical data of VOCs concentration monitored by ground stations in the target area, and form a second training set with the historical data of near-ground polluted gas concentration; The second training set is divided into three groups according to the dynamic partitioning rules. Each group is trained with the historical data of near-ground polluted gas concentration and related influencing factor data as input and the historical data of VOCs concentration as output, and three VOCs concentration estimation models are obtained. The near-ground pollutant gas includes at least formaldehyde, and the related influencing factors include at least temperature, humidity and normalized vegetation index.

8. A VOCs concentration estimation device, characterized in that: The device comprises: A data acquisition module is used to obtain column concentration data, profile data of polluted gases, and influencing factor data monitored by satellite in the target area; A data processing module, used to calculate first-resolution near-surface polluted gas concentration data according to the column concentration data and profile data of the polluted gas, and to resample the influencing factor data; A downscaling module, used for inputting the first-resolution near-surface polluted gas concentration data and the processed influencing factor data into a downscaling model, and outputting the second-resolution near-surface polluted gas concentration data, wherein the second resolution is greater than the first resolution; The concentration estimation module is used to input the second-resolution near-ground polluted gas concentration data into the VOCs concentration estimation model and output the VOCs concentration data of the target area.

9. An electronic device, characterized in that: It comprises a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the VOCs concentration estimation method according to any one of claims 1 to 7 is implemented.

10. A storage medium, characterized in that: A computer program is stored thereon, and when the program is executed by a processor, the VOCs concentration estimation method according to any one of claims 1 to 7 is implemented.

Images