Satellite-ground cooperative monitoring method for sand and dust weather

Abstract

The invention discloses a sand and dust weather satellite-ground cooperative monitoring method, and belongs to the technical field of atmospheric environment and meteorological disaster monitoring, and the method comprises the steps: calculating a particulate matter concentration change abnormal coefficient of each monitoring station according to a ground environment air quality automatic monitoring network; calculating visible light, intermediate infrared and thermal infrared indexes; extracting a satellite remote sensing monitoring index near each ground monitoring station, and comprehensively comparing ground monitoring to confirm an optimal threshold value of sand and dust discrimination of each index; and extracting a sand and dust weather distribution area, and calculating a sand and dust intensity index. According to the invention, by fusing ground environment air quality monitoring point data and a large-range satellite remote sensing monitoring result, a sand-dust weather integrated monitoring technology method model is comprehensively constructed, dynamic monitoring of regional large-range sand-dust weather distribution is realized, full-coverage sand-dust weather distribution judgment can be carried out on a large-range region, and real-time monitoring of the large-range region is realized. And the sand and dust weather transmission path and the influence thereof can be dynamically reflected.

Description

Technical Field

[0001] The present invention belongs to the technical field of atmospheric environment and meteorological disaster monitoring, and particularly relates to a space-ground collaborative monitoring method for sand-dust weather. Background Art

[0002] Sandstorms pose great hazards to agriculture, transportation, and respiratory health. The northern region of China is a high-incidence area of sand-dust weather. Especially in spring every year, there are frequent long-distance cross-border transmission weather processes from Mongolia to China, which not only cause serious deterioration of urban environmental air quality in a large area, but also affect many social and economic activities such as agricultural production losses, paralysis of transportation infrastructure, and reduction of the stability of the energy industry.

[0003] The sand-dust weather process generally covers a large area and even involves cross-border transmission, with a wide range of influence. The coverage of traditional ground monitoring stations is limited. Remote sensing technology can provide spatially continuous and real-time monitoring data to make up for the deficiencies of ground observations. The remote sensing monitoring technology for sand-dust weather is of great significance in dealing with natural disasters such as sandstorms, environmental protection, and public health management, and provides an important means for monitoring and evaluating the origin, transmission path, and influence range of the sand-dust weather process. It has the characteristics of being macroscopic, dynamic, objective, and accurate. Compared with traditional ground monitoring means in terms of information acquisition, it has unique advantages in terms of spatial range, and can continuously obtain the spatio-temporal changes of the atmospheric particulate matter concentration distribution in a large area, so it can effectively reflect the intensity of the sand-dust weather in the region. At present, the mainstream satellite data for remote sensing monitoring of sand-dust weather at home and abroad can achieve up to once every 10 minutes at the highest, and the spatial resolution can reach up to 1 km at the highest. The atmospheric particulate matter concentration based on ground station monitoring can reach once an hour, which can meet the needs of space-ground collaborative monitoring of sand-dust weather.

[0004] At present, sand-dust monitoring is mainly carried out based on real-time measurement of ground monitoring stations, atmospheric chemical model simulation technology, or satellite remote sensing. However, each single means has different defects. Ground stations are mainly distributed in urban built-up areas, and there is still a certain distance from the sand source, resulting in deficiencies in regional coverage. Especially in recent years, with the frequent cross-border transmission of sand-dust, the defects of ground stations have become increasingly prominent; for atmospheric chemical models, it is difficult to dynamically update the sand flux, and there are also certain uncertainties in the sand generation mechanism; satellites mainly detect sandstorm weather through thermal infrared sensors. Since different spectral characteristics change with different times and spaces, there are still large uncertainties in the remote sensing identification of sand-dust weather, especially the insufficient ability to identify floating dust and blowing dust weather, and it is difficult to guarantee the product accuracy. Moreover, there is still a certain gap between the sand-dust intensity mainly through thermal infrared optical characteristics and the existing sand-dust weather grade classification method. Therefore, the insufficient coverage of traditional ground monitoring and the unstable accuracy of satellite remote sensing monitoring lead to the inability to accurately detect sand-dust weather. Summary of the Invention

[0005] The object of the present invention is to propose a satellite-ground collaborative monitoring method for sand and dust weather. By integrating the spectral characteristics of ground stations and satellite remote sensing, a satellite-ground collaborative monitoring model for sand and dust weather is constructed to accurately and comprehensively identify the dynamic changes in the spatio-temporal distribution of sand and dust. It can not only comprehensively reflect the dynamic changes in the regional sand and dust distribution, but also accurately reflect the impact of sand and dust weather on the changes in the urban atmospheric environment, thus making up for the limitations of traditional monitoring in terms of insufficient coverage and unstable monitoring result accuracy, which lead to the inability to accurately detect sand and dust weather.

[0006] To achieve the above object, the present invention is implemented by the following technical solutions: The method includes:

[0007] According to the automatic monitoring network of ground environmental air quality, calculate the abnormal coefficient of the change in particulate matter concentration at each monitoring station, and judge the impact of sand and dust weather on each station one by one;

[0008] According to the monitoring results of wide-coverage multi-spectral geostationary satellite remote sensing, calculate the visible light, mid-infrared and thermal infrared indices;

[0009] Extract the satellite remote sensing monitoring indices near each ground monitoring station, and comprehensively compare the ground monitoring to confirm the optimal threshold for each index to be used for sand and dust discrimination;

[0010] Extract the distribution area of sand and dust weather, calculate the sand and dust intensity index, and divide the impact level of sand and dust weather.

[0011] In one solution, the calculation of the abnormal coefficient of the change in particulate matter concentration at each monitoring station includes:

[0012] According to the ground environmental air quality monitoring network, obtain the hourly concentration monitoring data of PM 2.5 , PM 10 for two elements at each station in the recent 3 days;

[0013] Calculate the statistical characteristics of the 3-day coarse particulate matter at each station, including the average value and standard deviation of PM 10 and the ratio of coarse particulate matter;

[0014] Compare the statistical characteristics of the real-time monitoring and historical monitoring results at each station, and calculate the coefficient of variation of the real-time monitoring PM 10 concentration and the ratio of coarse particulate matter at the station;

[0015] Comprehensively judge the impact of sand and dust weather on the station, and divide all stations into stations affected by sand and dust and stations not affected by sand and dust; When the PM 10 mass concentration is greater than 150 μg / m³, CVCPM is greater than 2, CVCPMR is greater than 2, and RPM 10 is greater than 0.6, it is judged that the station is affected by sand and dust weather, otherwise it is judged not to be affected by sand and dust weather.

[0016] In one solution, calculating the visible light, mid-infrared, and thermal infrared indices includes:

[0017] Satellite remote sensing monitoring includes multispectral satellite remote sensing monitoring. Based on the multispectral satellite remote sensing monitoring data, DN values near five bands at 0.47μm, 2.1μm, 3.8μm, 11μm, and 12μm are extracted, and the radiance is calculated according to the calibration coefficient;

[0018] Calculate the apparent reflectance near the two bands at 0.47μm and 2.1μm;

[0019] Calculate the brightness temperature near the three bands at 3.8μm, 11μm, and 12μm according to Planck's formula;

[0020] Calculate the normalized dust index, mid-infrared band brightness temperature difference value, and thermal infrared band brightness temperature difference value index respectively.

[0021] In one solution, the optimal thresholds of each index for dust discrimination include:

[0022] According to the geographical coordinates of the ground stations, match the satellite remote sensing monitoring data at the same moment. Extract the average values of three parameters, NDDI, BTDMI, and BTDTI, in the vicinity by extracting a 3*3 window centered on each station coordinate to form a comprehensive dust observation data set;

[0023] According to the station classification, divide the comprehensive dust observation data set into two categories: the subset affected by dust and the subset not affected by dust, and calculate the average value and standard deviation of the three indices of the two subsets respectively. The average values of the three indices of the subset affected by dust are respectively denoted as The standard deviation is denoted as The three indices of the subset not affected by dust The standard deviation is denoted as

[0024] Calculate the optimal segmentation thresholds T NDDI , T BTDMI , T BTDDI

[0025] According to the three multispectral satellite remote sensing indices, combined with the thresholds T NDDI , T BTDMI , T BTDDI Perform pixel-by-pixel classification to identify dust weather. If a pixel meets the three conditions of NDDI>T NDDI , BTDMI>T BTDMI and BTDDI>T BTDDI , then the pixel is judged as a dust pixel, otherwise it is a non-dust pixel.

[0026] In one solution, calculating the dust intensity index and dividing the dust weather impact levels includes:

[0027] Matching the ground site monitoring PM of dust pixels 10 concentration and three satellite remote sensing monitoring indices, constructing a dust monitoring dataset X and a PM 10 concentration matrix Y;

[0028] Using the geographically weighted regression method to construct a satellite remote sensing PM 10 concentration remote sensing inversion method model;

[0029] Evaluating the regional PM 10 concentration remote sensing inversion: Adopting the weighted least squares method, obtaining a weight function matrix according to the geographically weighted regression model;

[0030] Obtaining the optimal bandwidth of each ground site according to the weight function matrix, and obtaining a weight function matrix of each input parameter according to the optimal bandwidth;

[0031] Performing spatial matching on the weight function matrix of each input parameter and the geographical location of the ground site corresponding to the input parameter to obtain a spatial function weight matrix;

[0032] Obtaining the regression coefficients of the constant term, NDDI, BTDMI, and BTDDI respectively according to the spatial function weight matrix;

[0033] According to the NDDI, BTDMI, BTDDI corresponding to each pixel in the satellite remote sensing data and the regression coefficients, combining with the geographically weighted regression model to obtain the PM of each pixel in the evaluation area 10 concentration;

[0034] If the PM 10 concentration of the dust pixel is less than 300 μg / m 3 then it is classified as floating dust. If the PM 10 concentration of the dust pixel is greater than or equal to 300 μg / m 3 and less than 600 μg / m 3 then it is classified as blowing dust. If the PM 10 concentration of the dust pixel is greater than or equal to 600 μg / m 3 then it is classified as sandstorm.

[0035] Advantages of the present invention:

[0036] The present invention constructs a satellite-ground collaborative monitoring method for sand-dust weather. By integrating the spectral characteristics of ground stations and satellite remote sensing, a satellite-ground collaborative monitoring model for sand-dust weather is constructed to accurately and comprehensively identify the dynamic changes in the spatio-temporal distribution of sand-dust. It can not only comprehensively reflect the dynamic changes in the regional sand-dust distribution, but also accurately reflect the impact of sand-dust weather on the changes in the urban atmospheric environment, thus making up for the limitations of traditional monitoring, such as insufficient monitoring coverage and unstable monitoring result accuracy, which lead to the inability to accurately detect sand-dust weather. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is a flowchart of a satellite-ground collaborative monitoring method for sand-dust weather according to the present invention;

[0038] Figure 2 It is a flowchart for calculating the anomaly coefficient of the change in particulate matter concentration at each monitoring station;

[0039] Figure 3 It is a flowchart for calculating visible light, mid-infrared and thermal infrared indices;

[0040] Figure 4 It is a flowchart for confirming the optimal threshold for each index used in sand-dust discrimination;

[0041] Figure 5 It is a flowchart for calculating the sand-dust intensity index and dividing the impact level of sand-dust weather. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0042] To facilitate the understanding of the present invention, the present invention will be described more comprehensively below with reference to the relevant drawings. The typical embodiments of the present invention are shown in the drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. On the contrary, the purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive.

[0043] Unless otherwise defined, all technical and scientific terms used in the present invention have the same meaning as understood by those skilled in the technical field to which the present invention belongs. The terms used in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. To facilitate the understanding of the present invention, the present invention will be described more comprehensively below with reference to the relevant drawings. The typical embodiments of the present invention are shown in the drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. On the contrary, the purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive.

[0044] As Figure 1 shown, a satellite-ground collaborative monitoring method for sand-dust weather specifically includes:

[0045] S1. Calculate the anomaly coefficient of the particulate matter concentration change at each monitoring station based on the ground ambient air quality automatic monitoring network, and judge the impact of sand and dust weather on each station one by one.

[0046] As Figure 2 shown, the step S1 further includes:

[0047] S1.1 Obtain the hourly concentration monitoring data of the two elements of PM 2.5 , PM 10 at each station in the most recent 3 days according to the ground ambient air quality monitoring network;

[0048] S1.2 Calculate the 3-day statistical characteristics of coarse particulate matter at each station, including PM 10 and the average value and standard deviation of the coarse particulate matter ratio ((PM 10 -PM 2.5 ) / PM 10 );

[0049] S1.3 Compare the statistical characteristics of the real-time monitoring and historical monitoring results at each station, calculate the coefficient of variation of the real-time monitoring PM 10 concentration and the coarse particulate matter ratio at the station, and the calculation formula is as follows:

[0050]

[0051] where PM 10 and PM 2.5 are the hourly concentrations of PM 10 and PM 2.5 at the station, is the historical average concentration of the station, σ PM10 is the historical standard deviation of the station, CVCPM is the coefficient of variation of the coarse particulate matter concentration, CVCPMR is the coefficient of variation of the coarse particulate matter concentration change rate, and RPM 10 is the coarse particulate matter concentration ratio.

[0052] S1.4 Comprehensively judge the impact of sand and dust weather on the station, and divide all stations into stations affected by sand and dust and stations not affected by sand and dust. If the four conditions that the PM 10 mass concentration is greater than 150 μg / m³, CVCPM is greater than 2, CVCPMR is greater than 2, and RPM 10 is greater than 0.6 are all met, it is judged that the station is affected by sand and dust weather, otherwise it is judged not to be affected by sand and dust weather.

[0053] S2. Calculate the visible light, mid-infrared and thermal infrared indices according to the monitoring results of wide-coverage multi-spectral geostationary satellite remote sensing.

[0054] As Figure 3 shown, the step S2 further includes:

[0055] The satellite remote sensing monitoring described in S2.1 includes multispectral satellite remote sensing monitoring. Based on the multispectral satellite remote sensing monitoring data, the DN values near the five bands of 0.47μm, 2.1μm, 3.8μm, 11μm, and 12μm are extracted, and the radiance is calculated according to the calibration coefficient.

[0056] S2.2 Calculate the apparent reflectance near the two bands of 0.47μm and 2.1μm. The calculation formula is as follows:

[0057]

[0058] Among them, ρ λ is the apparent reflectance at band λ, L λ is the solar spectral radiance value received by the satellite sensor at band λ, D is the Earth-Sun distance, ESUN is the average solar spectral irradiance at the top of the atmosphere, and cos(θ) is the cosine of the solar zenith angle θ.

[0059] S2.3 Calculate the brightness temperature near the three bands of 3.8μm, 11μm, and 12μm according to Planck's formula. The calculation formula is as follows:

[0060]

[0061] Among them, c1 = 14387.7K, c2 = 1.19014×10 8 Wμm 4 m -2 sr -1 , L is the radiance value, which can be directly read from the satellite detection data.

[0062] S2.4 Calculate the Normalized Difference Dust Index (NDDI), the Bright Temperature Difference Index in the Mid-Infrared band (BTDMI), and the Bright Temperature Difference Index in the Thermal Infrared band (BTDTI) respectively. The calculation formulas are as follows:

[0063]

[0064] BTDMI = BT 3.8 -BT 11

[0065] BTDTI = BT 12 -BT11

[0066] Among them, ρ 0.47 and ρ 2.1 are the apparent reflectances of the satellite near the two bands of 0.47 μm and 2.1 μm respectively, and BT 3.8 、BT 11 and BT 12 are the brightness temperatures near the three bands of 3.8 μm, 11 μm and 12 μm respectively.

[0067] S3. Extract the satellite remote sensing monitoring indexes near each ground monitoring site, and comprehensively compare the ground monitoring to confirm the optimal threshold of each index for dust discrimination.

[0068] As Figure 4 shown, the step S3 further includes:

[0069] S3.1 According to the geographical coordinates of the ground site, match the satellite remote sensing monitoring data at the same moment, and extract the average values of the three parameters of NDDI, BTDMI and BTDTI in the vicinity by extracting a 3*3 window centered on each site coordinate to form a comprehensive dust observation data set;

[0070] S3.2 According to the site classification in the step S1.4, divide the comprehensive dust observation data set into two categories: the subset affected by dust and the subset not affected by dust, and calculate the average value and standard deviation of the three indexes of the two subsets respectively. Among them, the average values of the three indexes of the subset affected by dust are respectively denoted as The standard deviation is denoted as The three indexes of the subset not affected by dust The standard deviation is denoted as

[0071] S3.3 Calculate the optimal segmentation thresholds T NDDI 、T BTDMI 、T BTDDI , and the calculation method is as follows:

[0072]

[0073]

[0074] S3.4 According to the three multispectral satellite remote sensing indexes, combined with the thresholds T NDDI 、T BTDMI 、T BTDDI

[0075] Perform pixel-by-pixel classification to identify dust weather. If the pixel satisfies NDDI>T NDDI 、BTDMI>T BTDMI

[0076] and BTDDI>T BTDDI If the three conditions are met, the pixel is determined as a dust pixel; otherwise, it is a non-dust pixel.

[0077] S4. Extract the distribution area of sand and dust weather, calculate the sand and dust intensity index, and divide the influence level of sand and dust weather.

[0078] Such as Figure 5 shown, the step S4 further includes:

[0079] S4.1 Match the ground station monitoring PM of dust pixels 10 concentration and three satellite remote sensing monitoring indices, and construct the dust monitoring dataset X and PM 10 concentration matrix Y;

[0080] S4.2 Use the geographically weighted regression method to construct the following satellite remote sensing PM 10 concentration remote

[0081] inversion method model:

[0082] PM 10 (u i ,v i ) =

[0083] β0(u i ,v i ) + β1(u i ,v i ) × NDDI(u i ,v i ) + β2(u i ,v i ) × BTMDI(u i ,v i ) + β3(u i ,v i ) × BTDDI(u i ,v i )

[0084] Among them, β0(u i ,v i ) is the regression coefficient of the constant term at the observation point (u i ,v i ), β1(u i ,v i ) is

[0085] the regression coefficient of the NDDI at the observation point (u i ,v i ), β2(u i ,v i ) is the BTDMI at the observation

[0086] Regression coefficient at the measurement point (u i , v i ), β3(u i , v i ) is the regression coefficient of the BTDDI at the observation point (u i , v i ), and PM 10 (u i , v i ) is the PM i , v i ) concentration at the observation point (u 10 concentration.

[0087] S4.3 Remote sensing inversion of PM 10 concentration in the assessment area:

[0088] Using the weighted least squares method, obtain the weight function matrix according to the geographically weighted regression model. The weight function matrix is as follows:

[0089] β(u i , v i ) = [X T W(u i , v i )X] -1 X T W(u i , v i )Y

[0090] where β is the regression coefficient, W is the weight function matrix, Y is the PM 10 concentration matrix, and X is the input parameter matrix. The input parameters include: constant term, NDDI, BTDMI, and BTDDI;

[0091] Using the cross-validation method, obtain the optimal bandwidth for each ground station according to the weight function matrix, and obtain the weight function matrix of each input parameter according to the optimal bandwidth;

[0092] Perform spatial matching on the weight function matrix of each input parameter and the geographical location of the ground station corresponding to the input parameter to obtain the spatial function weight matrix;

[0093] Through the Kriging spatial interpolation method, obtain the regression coefficients of the constant term, NDDI, BTDMI, and BTDDI respectively according to the spatial function weight matrix;

[0094] According to the NDDI, BTDMI, BTDDI, and the regression coefficients corresponding to each pixel in the satellite remote sensing data, combine the geographically weighted regression model to obtain the PM 10 concentration of each pixel in the assessment area.

[0095] S4.4 Classify the dust intensity according to the PM10 concentration of each dust pixel. The classification method is as follows: If the PM 10 concentration of the dust pixel is less than 300 μg / m 3 then it is classified as floating dust. If the PM 10 concentration of the dust pixel is greater than or equal to 300 μg / m 3 and less than 600 μg / m 3 then it is classified as wind-blown dust. If the PM 10 concentration of the dust pixel is greater than or equal to 600 μg / m 3 then it is classified as sandstorm.

[0096] It should be noted that a space-ground collaborative monitoring method for dust weather is a comprehensive monitoring method. It combines the advantages of accurate ground monitoring stations and wide satellite remote sensing coverage, and realizes accurate detection of dust weather processes through collaborative monitoring. Therefore, this method overcomes the disadvantages of insufficient coverage of traditional monitoring stations or low accuracy of satellite remote sensing for dust monitoring, making the space-ground collaborative dust weather monitoring more comprehensive and accurate.

[0097] In summary, the present invention is based on the deep integration of satellite remote sensing and ground monitoring networks. By fusing the spectral characteristics of ground stations and satellite remote sensing, a space-ground collaborative monitoring model for dust weather is constructed to accurately and comprehensively identify the dynamic changes in the spatio-temporal distribution of dust. It can not only comprehensively reflect the dynamic changes in regional dust distribution, but also accurately reflect the impact of dust weather on the changes in the urban atmospheric environment, thus making up for the limitations of traditional monitoring in insufficient coverage and unstable monitoring results accuracy, which lead to the inability to accurately detect dust weather, and objectively quantitatively characterizing the regional dust weather distribution and intensity level. Therefore, the space-ground collaborative monitoring method model for dust weather proposed by this invention provides a new and effective technical means for accurately detecting the origin, influence range, etc. of dust weather.

[0098] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the above method embodiments. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM), etc.

[0099] It should be understood that the detailed description of the technical solutions of the present invention by means of the preferred embodiments above is illustrative rather than restrictive. Those of ordinary skill in the art can modify the technical solutions described in each embodiment on the basis of reading the specification of the present invention, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of each embodiment of the present invention.

Claims

1. A space-ground collaborative monitoring method for sand and dust weather, characterized in that: The method described includes: Calculating the anomaly coefficient of the particulate matter concentration change at each monitoring site according to the ground environmental air quality automatic monitoring network, and judging the impact of sand and dust weather site by site; Calculating visible light, mid-infrared and thermal infrared indices according to the monitoring results of wide-coverage multi-spectral geostationary satellite remote sensing; Extracting the satellite remote sensing monitoring indices near each ground monitoring site, and comprehensively comparing with ground monitoring to confirm the optimal threshold for each index to be used for sand and dust discrimination; Extracting the distribution area of sand and dust weather, calculating the sand and dust intensity index, and dividing the impact level of sand and dust weather.

2. The satellite-ground collaborative monitoring method for sandstorm weather according to claim 1, wherein The calculation of the anomaly coefficient of the particulate matter concentration change at each monitoring site includes: According to the ground ambient air quality monitoring network, obtain the hourly concentration monitoring data of PM 2.5 and PM 10 for two elements at each station in the most recent 3 days; Calculate the 3-day statistical characteristics of coarse particulate matter for each site, including PM 10 as well as the mean and standard deviation of the coarse particulate matter ratio; Compare the statistical characteristics of real-time monitoring and historical monitoring results at each site, and calculate the coefficient of variation of the PM 10 concentration and the ratio of coarse particulate matter at the site in real-time monitoring; Comprehensively judge the impact of dust weather on stations, and divide all stations into stations affected by dust and stations not affected by dust; PM 10 When the mass concentration is greater than 150 μg / m³, CVCPM is greater than 2, CVCPMR is greater than 2, and RPM 10 is greater than 0.6 and all four conditions are met simultaneously, it is judged that the station is affected by dust weather; otherwise, it is judged not to be affected by dust weather.

3. A satellite-ground collaborative monitoring method for sand and dust weather according to claim 1, characterized in that, The calculation of visible light, mid-infrared and thermal infrared indices includes: Satellite remote sensing monitoring includes multi-spectral satellite remote sensing monitoring. According to the multi-spectral satellite remote sensing monitoring data, DN values near 5 bands of 0.47μm, 2.1μm, 3.8μm, 11μm and 12μm are extracted, and the radiance is calculated according to the calibration coefficient; Calculating the apparent reflectance near the two bands of 0.47μm and 2.1μm; Calculating the brightness temperature near the three bands of 3.8μm, 11μm and 12μm according to Planck's formula; Calculating the normalized sand and dust index, the mid-infrared band brightness temperature difference value and the thermal infrared band brightness temperature difference value index respectively.

4. The sandstorm weather satellite-ground collaborative monitoring method according to claim 1, characterized in that: The optimal threshold for each index to be used for sand and dust discrimination includes: According to the geographical coordinates of the ground site, matching the satellite remote sensing monitoring data at the same moment, and extracting the average values of three parameters of the normalized sand and dust index NDDI and the mid-infrared band brightness temperature difference value BTDMI near the 3*3 window centered on each site coordinate to form a sand and dust comprehensive observation data set; According to the site classification, the comprehensive dust observation dataset is divided into two categories: the subset affected by dust and the subset not affected by dust. The mean and standard deviation of the three indices of the two subsets are calculated respectively. The mean values of the three indices of the subset affected by dust are denoted as The standard deviation is denoted as The three indices of the subset not affected by dust The standard deviation is denoted as Calculate the optimal segmentation thresholds T of the three indices according to the mean value and the standard deviation NDDI , T BTDMI , T BTDDI Based on three multispectral satellite remote sensing indices and combined with thresholds T NDDI , T BTDMI , T BTDDI perform pixel-by-pixel classification to identify sand and dust weather. If a pixel satisfies NDDI > T NDDI , BTDMI > T BTDMI and BTDDI > T BTDDI and meets all three conditions, then the pixel is judged as a sand and dust pixel; otherwise, it is a non-sand and dust pixel.

5. A satellite-ground collaborative monitoring method for sand and dust weather according to claim 1, characterized in that: The calculation of the sand and dust intensity index and the division of the impact level of sand and dust weather include: Match the ground station monitoring PM of dust pixels 10 concentration and three satellite remote sensing monitoring indices to construct the dust monitoring dataset X and the PM 10 concentration matrix Y; Construct a remote sensing inversion method model for satellite remote sensing PM 10 concentration using the geographically weighted regression method; PM in the evaluation area 10 Remote sensing inversion of concentration: Using the weighted least squares method, obtain the weight function matrix according to the geographically weighted regression model; The weight function matrix obtains the optimal bandwidth of each ground site, and obtains the weight function matrix of each input parameter according to the optimal bandwidth; Spatially matching the weight function matrix of each input parameter and the geographical location of the ground site corresponding to the input parameter to obtain the spatial function weight matrix; The spatial function weight matrix respectively obtains the regression coefficients of the constant term, the normalized sand and dust index NDDI, the mid-infrared band brightness temperature difference value BTDMI and the thermal infrared band brightness temperature difference value index BTDDI; Combining the Normalized Dust Index NDDI, the Brightness Temperature Difference Value BTDMI in the mid-infrared band, and the regression coefficient corresponding to each pixel in the satellite remote sensing data, and using the geographically weighted regression model to obtain the PM of each pixel in the evaluation area 10 concentration; Dust pixel PM 10 Concentration less than 300 μg / m 3 Then it is classified as floating dust, dust pixel PM 10 Concentration greater than or equal to 300 μg / m 3 And less than 600 μg / m 3 Then it is classified as blowing dust, dust pixel PM 10 Concentration greater than or equal to 600 μg / m 3 Then it is classified as sandstorm.

Images