All-day full-coverage near-surface NO2 concentration inversion method and system

Abstract

The invention discloses an all-day full-coverage near-surface NO2 concentration inversion method and system, and belongs to the technical field of atmospheric environment information. The method comprises the steps that satellite observation data and foundation observation data are acquired and preprocessed, and a matched data set is obtained; according to the matched data set, constructing a first NO2 concentration inversion model and a second NO2 concentration inversion model, and respectively obtaining a first NO2 concentration inversion result and a second NO2 concentration inversion result by using the two models; the first NO2 concentration inversion result and the second NO2 concentration inversion result are fused based on the Poisson image adjustment theory, and a comprehensive domain NO2 concentration inversion result is obtained; and based on discrete cosine transform, correcting an inversion error of a full domain NO2 concentration inversion result to obtain the final near-surface NO2 concentration. Poisson adjustment and discrete cosine transform are applied, and satellite-earth observation data are combined, so that all-day full-coverage near-surface NO2 concentration data are obtained.

Description

Technical Field

[0001] The present invention belongs to the field of atmospheric environment information technology, and in particular relates to a near-surface NO2 concentration inversion method and system with full coverage throughout the day. Background Art

[0002] NO 2 The sources of monitoring data can be divided into ground-based site observations, airborne observations, and satellite remote sensing observations. Although ground-based site observations can provide us with relatively accurate real-time data of the near-ground or troposphere around the site, their observation range is limited, and the sites are sparse and unevenly distributed. Airborne observations are highly maneuverable and can obtain vertical column concentration profiles within the flight area. The results are often used to verify satellite remote sensing products, but due to the high cost, large-area observations cannot be achieved and cannot be widely used. In contrast, satellite remote sensing observations can achieve long-term continuous large-area observations. Therefore, high-resolution satellite data and other auxiliary variables are used to invert and obtain near-ground NO 2 Concentration has become the current development trend.

[0003] However, the current NO 2 Monitoring is still mainly done by polar-orbiting satellites, which usually observe the same area once every 1-3 days, making it difficult to capture short-term pollution events and the complete process of daily changes. Although the development of geostationary satellites has made continuous time monitoring possible, the current monitoring of NO 2 The geostationary satellite sensors monitoring the NO 2 The inversion is mainly based on ultraviolet spectral differential absorption, and the spectral window used is usually 400~500nm. This determines that the current technical solution cannot obtain NO at night. 2 Observation; but NO at night 2 The distribution of NO cannot be ignored. In addition, due to the influence of cloud cover, etc., the current NO 2 The instantaneous inversion results have large-scale data missing, which limits the commercial application of the inversion results. Summary of the invention

[0004] The purpose of the present invention is to overcome the deficiencies in the prior art and provide a near-surface NO2 concentration inversion method and system covering all day and night. 2 Concentration inversion to achieve full coverage of NO in the region 2 Concentration acquisition.

[0005] The present invention provides the following technical solutions:

[0006] In a first aspect, a near-surface NO2 concentration inversion method with full coverage throughout the day is provided, comprising: obtaining satellite observation data and ground-based observation data, and preprocessing the satellite observation data and the ground-based observation data to obtain a matching data set; constructing a first NO2 concentration inversion method based on the matching data set. 2 Concentration inversion model and the second NO 2 The concentration inversion model uses the first NO 2 Concentration inversion model and the second NO 2 The concentration inversion model was used to obtain the first NO 2 The concentration inversion results and the second NO 2 Concentration inversion results; Based on Poisson image adjustment theory, the first NO 2 The concentration inversion results and the second NO 2 The concentration inversion results are combined to obtain the comprehensive NO 2 Concentration inversion results; The full domain NO is corrected based on discrete cosine transform 2 The inversion error of the concentration inversion result is used to obtain the final near-surface NO 2 concentration.

[0007] As an optional technical solution of the present invention, the satellite observation data includes satellite brightness temperature data, meteorological parameters, normalized difference vegetation index and surface elevation data.

[0008] As an optional technical solution of the present invention, the satellite observation data and the ground-based observation data are preprocessed, including: resampling the meteorological parameters, normalized vegetation index and surface elevation data; matching the ground-based observation data with the satellite brightness temperature data and the resampled meteorological parameters, normalized vegetation index and surface elevation data.

[0009] As an optional technical solution of the present invention, the first NO 2 The concentration inversion model is expressed as:

[0010] ;

[0011] The second NO 2 The concentration inversion model is expressed as:

[0012] ;

[0013] in, Indicates the first NO 2 Concentration inversion model, Indicates the second NO 2 Concentration inversion model, represents a function, represents the brightness temperature measured by the first channel of the satellite sensor, represents the brightness temperature measured by the second channel of the satellite sensor, represents the brightness temperature measured by the third channel of the satellite sensor, represents the brightness temperature measured by the fourth channel of the satellite sensor, represents the solar altitude angle, is the relative humidity of the surface, represents the air temperature at 2 meters above the ground. represents the surface pressure, represents the meridional component of wind speed at 10 meters above the ground surface, represents the latitudinal component of wind speed at 10 meters above the ground, represents the boundary layer height, represents the atmospheric precipitable water content, represents the near-surface ozone mass mixing ratio, represents the total amount of atmospheric column ozone, represents the normalized difference vegetation index, Represents surface elevation data.

[0014] As an optional technical solution of the present invention, the first NO 2 The concentration inversion results and the second NO 2 The concentration inversion results are combined to obtain the comprehensive NO 2 Concentration inversion results, including:

[0015] In the case where the first NO2 concentration inversion result is not included and the second NO 2 Area of ​​concentration inversion results The objective function f to be solved satisfies:

[0016] ;

[0017] in, , represents the gradient operator, z represents the area The vector field on Indicates area The boundary of , S represents the full domain, Indicates area No.1 in China 2 The concentration inversion result is a known function. Indicates that it is located f on Indicates that it is located on ;

[0018] For a grid p in the full domain S, the four grids above, below, left, and right adjacent to grid p are defined as the neighborhood space N of grid p.p , expressed as:

[0019] ;

[0020] Among them, q represents the adjacent grid of grid p, ;

[0021] area The border It is expressed as:

[0022] ;

[0023] The discretized form of the objective function f to be solved is expressed as:

[0024] ;

[0025] in, , represents the value of the objective function f to be solved at grid p, represents the value of the objective function f to be solved at grid q, represents the gradient vector between grids p and q, , Represents the second NO at grid p before Poisson image adjustment 2 Concentration inversion results, Indicates the second NO at grid q before Poisson image adjustment 2 Concentration inversion results;

[0026] for , satisfy:

[0027] hour,

[0028] get ;

[0029] in, Indicates the first NO 2 The value of the concentration inversion result at grid q;

[0030] At this time, solve the objective function f, in the region The fused NO 2 The concentration inversion results show that the first NO 2 Concentration inversion results and fused NO 2 The concentration inversion results are combined to form a comprehensive NO 2 Concentration inversion results.

[0031] As an optional technical solution of the present invention, the full range NO correction based on discrete cosine transform 2The inversion error of the concentration inversion result is used to obtain the final near-surface NO 2 Concentrations include:

[0032] The ground-based observation data XG station With the full domain NO 2 Concentration inversion results XG domain The ratio is expressed as:

[0033] ;

[0034] in, The closer the absolute value of is to 1, the smaller the inversion error is;

[0035] right Perform spatial interpolation to obtain the initial distribution of full domain error , expressed as:

[0036] ;

[0037] Where i represents the i-th ground-based observation site, n represents the total number of ground-based observation sites, , represents the ground-based observation data a of the i-th ground-based observation station i The same time and location as the full domain NO 2 Concentration inversion results b i The ratio of B represents the overall domain NO at the grid (u, v). 2 Inversion concentration, u represents the longitude coordinate, v represents the latitude coordinate, , ,u i represents the longitude coordinate of the i-th ground-based observation station, v i represents the latitude coordinate of the i-th ground-based observation station;

[0038] The discrete cosine transform includes a forward transform and an inverse transform, and the initial distribution of the full domain error Perform iterative correction, expressed as:

[0039] ;

[0040] Where k represents the kth iteration, represents the relaxation factor, Indicates a 0-1 mask, , A constant representing the matrix type, STDCT represents forward transformation, and ISTDCT represents inverse transformation;

[0041] When the infinite norm of the difference between two consecutive iteration results is less than 10 -3 Or if the number of iterations is greater than 200, the iteration stops and the final iteration result is used as the inversion error distribution. , then the final near-surface NO 2 The concentration is expressed as:

[0042] .

[0043] As an optional technical solution of the present invention, the transformation formula of the forward transformation is expressed as:

[0044] ;

[0045] ;

[0046] The transformation formula of the inverse transformation is expressed as:

[0047] ;

[0048] ;

[0049] in, Represents the error distribution data before iteration, in the form of a three-dimensional matrix; It represents the error distribution data after a complete iteration, in the form of a three-dimensional matrix; M, N and P represent the total number of rows, the total number of columns and the length of the time series respectively; m, j and t represent the row position, column position and observation time of the target grid respectively. Indicated by After a positive transformation, the three-dimensional matrix, h, l and w represent the frequency domain coordinates after the transformation, and their range of variation is the same as that of m, j and t;

[0050] Matrix type constants It is expressed as:

[0051] ;

[0052] in, Represents parameters, h∈[0, M-1 ], l∈[0, N-1 ], w∈[0 ,P-1 ].

[0053] Secondly, a near-surface NO2 concentration inversion system with full coverage throughout the day is provided, including:

[0054] A data acquisition module is used to acquire satellite observation data and ground-based observation data, and pre-process the satellite observation data and the ground-based observation data to obtain a matching data set;

[0055] A model building prediction module is used to build a first NO according to the matching data set. 2 Concentration inversion model and the second NO 2 The concentration inversion model uses the first NO 2 Concentration inversion model and the second NO2 The concentration inversion model was used to obtain the first NO 2 The concentration inversion results and the second NO 2 Concentration inversion results;

[0056] A data fusion module is used to transform the first NO 2 The concentration inversion results and the second NO 2 The concentration inversion results are combined to obtain the comprehensive NO 2 Concentration inversion results;

[0057] An error correction module is used to correct the full range NO based on discrete cosine transform 2 The inversion error of the concentration inversion result is used to obtain the final near-surface NO 2 concentration.

[0058] Compared with the prior art, the present invention has the following beneficial effects:

[0059] The present invention provides a near-surface NO 2 The concentration inversion method matches satellite observation data with ground-based observation data and applies Poisson adjustment theory and discrete cosine transform to obtain full coverage of NO2 all day long. 2 Concentration data, improve NO 2 At the same time, this method uses a geostationary satellite sensor that is not specifically used for trace gas detection, giving full play to its high time resolution advantage in infrared detection, and achieves hourly NO concentration inversion accuracy near the surface. 2 Inversion of . BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 It is a flow chart of a near-surface NO2 concentration inversion method with full coverage throughout the day in an embodiment of the present invention;

[0061] Figure 2 Schematic diagram of Poisson image adjustment in an embodiment of the present invention;

[0062] Figure 3 It is the near-surface NO at all times during the whole day from January to February 2020 in the embodiment of the present invention. 2 Concentration and ground-based observation 10-fold cross-validation comparison chart;

[0063] Figure 4 It is the first NO in the embodiment of the present invention. 2 The concentration inversion results and the second NO 2 Concentration inversion results superposition and full-scale NO 2 Effect comparison chart of concentration inversion results;

[0064] Figure 5In the embodiment of the present invention, on March 2, 2020, based on the first NO 2 The near-surface NO concentration inversion model at all times of the day 2 Concentration distribution diagram;

[0065] Figure 6 It is the embodiment of the present invention based on the second NO on March 2, 2020 2 The near-surface NO concentration inversion model at all times of the day 2 Concentration distribution;

[0066] Figure 7 It is the near-surface NO at all times of the day based on discrete cosine transform on March 2, 2020 in the embodiment of the present invention. 2 Concentration distribution. DETAILED DESCRIPTION

[0067] The present invention will be further described below in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and cannot be used to limit the protection scope of the present invention.

[0068] Example 1

[0069] This embodiment provides a near-surface NO2 concentration inversion method with full coverage throughout the day, such as Figure 1 As shown, the specific steps include:

[0070] Step 1: Obtain satellite observation data and ground-based observation data, and preprocess the satellite observation data and ground-based observation data to obtain a matching data set.

[0071] 1.1. The satellite observation data include satellite brightness temperature data, meteorological parameters, normalized difference vegetation index and surface elevation data. Since satellite brightness temperature data will be interfered by cloud cover, all satellite observation data use clear sky samples.

[0072] In this embodiment, the satellite is the Japanese Himawari-8 / 9 satellite. The satellite brightness temperature data includes the brightness temperatures tbb7, tbb8, tbb9 and tbb10 measured by the four channels of the Japanese Himawari-8 / 9 satellite AHI sensor Band7 (λ=3.9μm), Band8 (λ=6.2μm), Band9 (λ=6.9μm) and Band10 (λ=7.3μm). The temporal sampling resolution of the satellite brightness temperature data is 1 hour.

[0073] In this embodiment, the meteorological parameters adopt the 1-hour resolution ERA-5 meteorological parameters released by the European Center for Mesoscale Weather Forecasts; the normalized vegetation index adopts the MODIS normalized vegetation index released by NASA; the surface elevation data adopts the SRTM surface elevation data released by NASA. Among them, the meteorological parameters include surface relative humidity RH, air temperature TEMP, surface pressure SP, east-west meridional component of wind at 10 meters above the surface , the north-south latitudinal component of the wind at 10 meters above the ground , boundary layer height BLH, atmospheric precipitable water PW, near-surface ozone mass mixing ratio O3mr and total atmospheric column ozone O3all.

[0074] In this embodiment, the ground-based observation data uses the national ambient air quality monitoring network data operated by the China National Environmental Monitoring Center.

[0075] 1.2. Since the acquired data has obvious structural heterogeneity and differences in spatiotemporal resolution, in order to obtain a spatiotemporal matching data set as a learning sample, the acquired data needs to be preprocessed. The details are as follows:

[0076] 1.2.1. The meteorological parameters, normalized difference vegetation index and surface elevation data are resampled with a sampling resolution of 0.05°.

[0077] 1.2.2. Matching the ground-based observation data with the satellite brightness temperature data and the resampled meteorological parameters, normalized difference vegetation index and surface elevation data.

[0078] Step 2: Construct the first NO according to the matching data set 2 Concentration inversion model and the second NO 2 The concentration inversion model uses the first NO 2 Concentration inversion model and the second NO 2 The concentration inversion model was used to obtain the first NO 2 The concentration inversion results and the second NO 2 Concentration inversion results.

[0079] 2.1. In this embodiment, based on the LightGBM machine learning algorithm, the first NO 2 Concentration inversion model and the second NO 2 Concentration inversion model.

[0080] For cloud-free areas and time periods, all matching data are input into the initialized LightGBM model as explanatory variables to obtain the first NO 2 The concentration inversion model is expressed as:

[0081] .

[0082] The data except the satellite brightness temperature data in the matching data are input into the initialized LightGBM model as explanatory variables to obtain the second NO 2 The concentration inversion model is expressed as:

[0083] .

[0084] in, Indicates the first NO 2 Concentration inversion model, Indicates the second NO 2 Concentration inversion model, represents a function, represents the brightness temperature measured by the first channel (λ=3.9μm) of the satellite sensor, represents the brightness temperature measured by the second channel (λ=6.2μm) of the satellite sensor, represents the brightness temperature measured by the third channel (λ=6.9μm) of the satellite sensor, represents the brightness temperature measured by the fourth channel (λ=7.3μm) of the satellite sensor, represents the solar altitude angle, is the relative humidity of the surface, represents the air temperature at 2 meters above the ground. represents the surface pressure, represents the meridional component of wind speed at 10 meters above the ground surface, represents the latitudinal component of wind speed at 10 meters above the ground, represents the boundary layer height, represents the atmospheric precipitable water content, represents the near-surface ozone mass mixing ratio, represents the total amount of atmospheric column ozone, represents the normalized difference vegetation index, Represents surface elevation data

[0085] 2.2. Use the first NO 2 Concentration inversion model and the second NO 2 The concentration inversion model obtained the first NO 2 The concentration inversion results and the second NO 2 Concentration inversion results.

[0086] In this embodiment, the first NO 2 The concentration inversion results are highly accurate, but due to cloud cover and other reasons, they are not comprehensive. 2 The concentration inversion results do not rely on satellite observations, so the NO concentration is seamless in both day and night time and space. 2However, due to the lack of necessary satellite observation data, its inversion accuracy is lower than that of the first NO 2 Concentration inversion results.

[0087] Step 3: Based on Poisson image adjustment theory, the first NO 2 The concentration inversion results and the second NO 2 The concentration inversion results are combined to obtain the comprehensive NO 2 Concentration inversion results. Specifically include:

[0088] like Figure 2 As shown, S represents the overall domain NO 2 Concentration area, Indicates the second NO 2 The area of ​​concentration inversion results, Indicates the first NO 2 The area where the concentration inversion results are obtained. There are known functions , it is necessary to calculate the definition that meets the fixed boundary conditions The objective function f on is as follows:

[0089] In the case where the first NO2 concentration inversion result is not included and the second NO 2 Area of ​​concentration inversion results The objective function f to be solved satisfies:

[0090] ;

[0091] in, , represents the gradient operator, z represents the area The vector field on Indicates area The boundary of , S represents the full domain, Indicates area No.1 in China 2 The concentration inversion result is a known function. Indicates that it is located f on Indicates that it is located on .

[0092] For the grid p in the full domain S, the four grids above, below, left and right adjacent to the grid p are defined as the neighborhood space Np of the grid p, expressed as:

[0093] ;

[0094] Among them, q represents the adjacent grid of grid p, , when the target grid is at the boundary, the number of its adjacent grids is less than 4.

[0095] area The border It is expressed as:

[0096] .

[0097] The discretized form of the objective function f to be solved is expressed as:

[0098] ;

[0099] in, , represents the value of the objective function f to be solved at grid p, represents the value of the objective function f to be solved at grid q, represents the gradient vector between grids p and q, , Represents the second NO at grid p before Poisson image adjustment 2 Concentration inversion results, Indicates the second NO at grid q before Poisson image adjustment 2 Concentration inversion results.

[0100] From the form of the solved function and the knowledge related to multivariate quadratic functions, we can know that for , satisfy:

[0101] hour,

[0102] get ;

[0103] in, Indicates the first NO 2 The value of the concentration inversion result at grid q.

[0104] At this time, solve the objective function f, in the region The fused NO 2 The concentration inversion results show that the first NO 2 Concentration inversion results and fused NO 2 The concentration inversion results are combined to form a comprehensive NO 2 Concentration inversion results.

[0105] like Figure 2 As shown, the second NO 2 Area of ​​concentration inversion results and the first NO 2 Area of ​​concentration inversion results Composed of comprehensive domain S.

[0106] Step 4: Correction of full-domain NO based on discrete cosine transform 2 The inversion error of the concentration inversion result is used to obtain the final near-surface NO 2 Concentration. Details are as follows:

[0107] 4.1. The ground-based observation data XG station With the full domain NO 2 Concentration inversion results XG domain The ratio is expressed as:

[0108] ;

[0109] Among them, the closer the absolute value of is to 1, the smaller the inversion error is.

[0110] Due to XG station Distribution is limited, so It is not comprehensive data. Perform spatial interpolation to obtain the initial distribution of full domain error , expressed as:

[0111] ;

[0112] Where i represents the i-th ground-based observation site, n represents the total number of ground-based observation sites, , represents the ground-based observation data a of the i-th ground-based observation station i The same time and location as the full domain NO 2 Concentration inversion results b i The ratio of B represents the overall domain NO at the grid (u, v). 2 Inversion concentration, u represents the longitude coordinate, v represents the latitude coordinate, , ,u i represents the longitude coordinate of the i-th ground-based observation station, v i Represents the latitude coordinate of the i-th ground-based observation station.

[0113] 4.2. The discrete cosine transform includes a forward transform and an inverse transform, and the initial distribution of the full domain error Perform iterative correction, expressed as:

[0114] ;

[0115] Where k represents the kth iteration, represents the relaxation factor, Indicates a 0-1 mask, , A constant indicating the matrix type. STDCT indicates forward transform and ISTDCT indicates inverse transform.

[0116] Furthermore, the transformation formula of the forward transformation is expressed as:

[0117] .

[0118] ;

[0119] The transformation formula of the inverse transformation is expressed as:

[0120] ;

[0121] ;

[0122] in, Represents the error distribution data before iteration, in the form of a three-dimensional matrix; It represents the error distribution data after a complete iteration, in the form of a three-dimensional matrix; M, N and P represent the total number of rows, the total number of columns and the length of the time series respectively; m, j and t represent the row position, column position and observation time of the target grid respectively. Indicated by After a positive transformation of the three-dimensional matrix, h, l and w represent the transformed frequency domain coordinates, and their variation range is the same as that of m, j and t.

[0123] Furthermore, the discrete cosine transform can convert the signal from the spatial domain to the frequency domain using the cosine function, thereby representing the data as the weighting of different frequency components, effectively achieving the smoothing of observation errors within the target space-time range.

[0124] Furthermore, constants of matrix type It is expressed as:

[0125] ;

[0126] in, Represents parameters, h∈[0, M-1 ], l∈[0, N-1 ], w∈[0 ,P-1 ].

[0127] 4.3. When the infinite norm of the difference between two consecutive iteration results is less than 10 -3 Or if the number of iterations is greater than 200, the iteration stops and the final iteration result is used as the inversion error distribution. , then the final near-surface NO 2 The concentration is expressed as:

[0128] .

[0129] Example 2

[0130] Based on Example 1, this example provides a specific application of this method, taking the central and eastern parts of China (95°E ~ 130°E, 20°N ~45°N) from January to February 2020 as a demonstration study area.

[0131] like Figure 3 As shown, based on the ten-fold cross-validation method, the first NO at each moment in January-February 2020 is shown. 2 Concentration inversion model Ⅰ )、Second NO 2 Concentration inversion model Ⅱ ), Poisson adjustment and DCT (discrete cosine transform) after satellite-ground fusion, four types of NO 2 The so-called 10-fold cross-validation method is to randomly divide the data set into 10 subsets of equal or approximately equal size, and use 9 of them as training data and 1 as test data in turn, and use the average of the 10 iterations of evaluation indicators as a comprehensive evaluation of the model performance. Figure 3 As shown, MODEL Ⅰ The accuracy is higher than MODEL Ⅱ , its R 2 0.72 and 0.69 respectively. Although the overall accuracy did not change significantly after Poisson adjustment and the RMSE and MAE decreased slightly, the Poisson adjustment ensured that NO 2 The continuity of the distribution in space. Apply DCT satellite-ground fusion to the image after Poisson adjustment to obtain high-precision NO 2 The inversion result is R 2 It is 0.86.

[0132] like Figure 4 As shown, (a) means putting MODEL Ⅰ and MODEL Ⅱ The results of the simple layer overlay, (b) shows the results after Poisson adjustment. At 02:00 local time on January 5, 2020, when MODEL Ⅰ and MODEL Ⅱ When the results of simple layer overlay are used, MODEL Ⅰ Effective inversion replacement MODEL Ⅱ From the results in Figure 2, we can see that there is a jump phenomenon in the concentration distribution.

[0133] like Figure 5-7 As shown, the MODEL time-by-time data for the entire day on March 2, 2020 is shown. Ⅰ 、MODEL Ⅱ NO integrated with satellite and earth observation 2 The results show that due to the influence of cloud cover, MODEL Ⅰ Inversion of NO2 There are large areas of missing data. However, the results of satellite-ground fusion have achieved full coverage of the domain and are highly consistent with ground-based observations. 2 High concentrations are mainly concentrated in Beijing-Tianjin-Hebei and the six surrounding provinces and cities, which is related to traffic emissions, industrial emissions, thermal power generation emissions, etc. In addition, northern Gansu, Inner Mongolia and some parts of Northeast China have NO 2 The concentration value is also high, which is consistent with the conclusions of other studies. In addition to emissions from transportation, industry and thermal power generation, farmland, grassland, etc. also release NO 2 From the perspective of time, NO 2 High concentrations are mainly distributed at night, and during the day, especially between 12 and 18 o'clock, NO 2 The overall concentration is low. This is because the strong sunlight during the day can promote the concentration of NO in the atmosphere. 2 At night, the photolysis reaction is significantly weakened, and the atmospheric boundary layer is significantly reduced, which promotes the formation of NO 2 Increased concentration.

[0134] Example 3

[0135] This embodiment provides a near-surface NO2 concentration inversion system with full coverage throughout the day, including:

[0136] The data acquisition module is used to acquire satellite observation data and ground-based observation data, and pre-process the satellite observation data and the ground-based observation data to obtain a matching data set.

[0137] A model building prediction module is used to build a first NO according to the matching data set. 2 Concentration inversion model and the second NO 2 The concentration inversion model uses the first NO 2 Concentration inversion model and the second NO 2 The concentration inversion model was used to obtain the first NO 2 The concentration inversion results and the second NO 2 Concentration inversion results.

[0138] A data fusion module is used to transform the first NO 2 The concentration inversion results and the second NO 2 The concentration inversion results are combined to obtain the comprehensive NO 2 Concentration inversion results.

[0139] An error correction module is used to correct the full range NO based on discrete cosine transform 2 The inversion error of the concentration inversion result is used to obtain the final near-surface NO 2 concentration.

[0140] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0141] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0142] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0143] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0144] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for inverting near-surface NO2 concentration with full coverage throughout the day, characterized in that: include: Acquire satellite observation data and ground-based observation data, and pre-process the satellite observation data and ground-based observation data to obtain a matching data set; According to the matching data set, a first NO2 concentration inversion model and a second NO2 concentration inversion model are constructed, and the first NO2 concentration inversion model and the second NO2 concentration inversion model are used to obtain a first NO2 concentration inversion result and a second NO2 concentration inversion result respectively; Based on Poisson image adjustment theory, the first NO2 concentration inversion result and the second NO2 concentration inversion result are integrated to obtain a full-domain NO2 concentration inversion result; The inversion error of the full-domain NO2 concentration inversion result is corrected based on discrete cosine transform to obtain the final near-surface NO2 concentration.

2. The all-day and full-coverage near-surface NO2 concentration inversion method according to claim 1 is characterized in that: The satellite observation data include satellite brightness temperature data, meteorological parameters, normalized difference vegetation index and surface elevation data.

3. The all-day and full-coverage near-surface NO2 concentration inversion method according to claim 2 is characterized in that: Preprocessing the satellite observation data and the ground-based observation data includes: Resampling the meteorological parameters, normalized difference vegetation index and surface elevation data; The ground-based observation data are matched with the satellite brightness temperature data and the resampled meteorological parameters, normalized difference vegetation index and surface elevation data.

4. The all-day and full-coverage near-surface NO2 concentration inversion method according to claim 1 is characterized in that: The first NO2 concentration inversion model is expressed as: ； The second NO2 concentration inversion model is expressed as: ； in, represents the first NO2 concentration inversion model, represents the second NO2 concentration inversion model, represents a function, represents the brightness temperature measured by the first channel of the satellite sensor, represents the brightness temperature measured by the second channel of the satellite sensor, represents the brightness temperature measured by the third channel of the satellite sensor, represents the brightness temperature measured by the fourth channel of the satellite sensor, represents the solar altitude angle, is the relative humidity of the surface, represents the air temperature at 2 meters above the ground. represents the surface pressure, represents the meridional component of wind speed at 10 meters above the ground surface, represents the latitudinal component of wind speed at 10 meters above the ground, represents the boundary layer height, represents the atmospheric precipitable water content, represents the near-surface ozone mass mixing ratio, represents the total amount of atmospheric column ozone, represents the normalized difference vegetation index, Represents surface elevation data.

5. The all-day and full-coverage near-surface NO2 concentration inversion method according to claim 1 is characterized in that: The first NO2 concentration inversion result and the second NO2 concentration inversion result are merged based on the Poisson image adjustment theory to obtain a full-domain NO2 concentration inversion result, including: In the area that does not include the first NO2 concentration inversion result but includes the second NO2 concentration inversion result The objective function f to be solved satisfies: ； in, , represents the gradient operator, z represents the area The vector field on Indicates area The boundary of , S represents the full domain, Indicates area The first NO2 concentration inversion result is a known function. Indicates that it is located f on Indicates that it is located on ; For a grid p in the full domain S, the four grids above, below, left, and right adjacent to grid p are defined as the neighborhood space N of grid p. p , expressed as: ； Among them, q represents the adjacent grid of grid p, ; area The border It is expressed as: ； The discretized form of the objective function f to be solved is expressed as: ； in, , represents the value of the objective function f to be solved at grid p, represents the value of the objective function f to be solved at grid q, represents the gradient vector between grids p and q, , represents the second NO2 concentration inversion result at grid p before Poisson image adjustment, It represents the second NO2 concentration inversion result at grid q before Poisson image adjustment; for , satisfy: hour, get ; in, represents the value of the first NO2 concentration inversion result at grid q; At this time, solve the objective function f, in the region The fused NO2 concentration inversion result is obtained above, and the first NO2 concentration inversion result and the fused NO2 concentration inversion result are combined to form a comprehensive NO2 concentration inversion result.

6. The all-day and full-coverage near-surface NO2 concentration inversion method according to claim 1 is characterized in that: The method of correcting the inversion error of the full-domain NO2 concentration inversion result based on discrete cosine transform to obtain the final near-surface NO2 concentration includes: The ground-based observation data XG station Compared with the inversion results of NO2 concentration in the whole area XG domain The ratio is expressed as: ； in, The closer the absolute value of is to 1, the smaller the inversion error is; right Perform spatial interpolation to obtain the initial distribution of full domain error , expressed as: ； Where i represents the i-th ground-based observation site, n represents the total number of ground-based observation sites, , represents the ground-based observation data a of the i-th ground-based observation station i The inversion results of the global NO2 concentration at the same time and location b i The ratio of B represents the global NO2 inversion concentration at the grid (u, v), u represents the longitude coordinate, and v represents the latitude coordinate. , ,u i represents the longitude coordinate of the i-th ground-based observation station, v i represents the latitude coordinate of the i-th ground-based observation station; The discrete cosine transform includes a forward transform and an inverse transform, and the initial distribution of the full domain error Perform iterative correction, expressed as: ； Where k represents the kth iteration, represents the relaxation factor, Indicates a 0-1 mask, , A constant representing the matrix type, STDCT represents forward transformation, and ISTDCT represents inverse transformation; When the infinite norm of the difference between two consecutive iteration results is less than 10 -3 Or if the number of iterations is greater than 200, the iteration stops and the final iteration result is used as the inversion error distribution. , then the final near-surface NO2 concentration is expressed as: 。 7. The all-day and full-coverage near-surface NO2 concentration inversion method according to claim 6 is characterized by: The transformation formula of the forward transformation is expressed as: ； ； The transformation formula of the inverse transformation is expressed as: ； ； in, Represents the error distribution data before iteration, in the form of a three-dimensional matrix; It represents the error distribution data after a complete iteration, in the form of a three-dimensional matrix; M, N and P represent the total number of rows, the total number of columns and the length of the time series respectively; m, j and t represent the row position, column position and observation time of the target grid respectively. Indicated by After a positive transformation, the three-dimensional matrix, h, l and w represent the frequency domain coordinates after the transformation, and their range of variation is the same as that of m, j and t; Matrix type constants It is expressed as: ； in, Represents parameters, h∈[0, M-1 ], l∈[0, N-1 ], w∈[0 ,P-1 ].

8. A near-surface NO2 concentration inversion system with full coverage throughout the day, characterized in that: include: A data acquisition module is used to acquire satellite observation data and ground-based observation data, and pre-process the satellite observation data and the ground-based observation data to obtain a matching data set; A model building prediction module, used to build a first NO2 concentration inversion model and a second NO2 concentration inversion model according to the matching data set, and use the first NO2 concentration inversion model and the second NO2 concentration inversion model to obtain a first NO2 concentration inversion result and a second NO2 concentration inversion result respectively; A data fusion module, for fusing the first NO2 concentration inversion result and the second NO2 concentration inversion result based on Poisson image adjustment theory to obtain a full-domain NO2 concentration inversion result; The error correction module is used to correct the inversion error of the full-domain NO2 concentration inversion result based on discrete cosine transform to obtain the final near-surface NO2 concentration.

Images