CN111737850A - Multisource satellite AOD fusion method based on uncertainty on pixel scale - Google Patents
Multisource satellite AOD fusion method based on uncertainty on pixel scale Download PDFInfo
- Publication number
- CN111737850A CN111737850A CN202010411688.8A CN202010411688A CN111737850A CN 111737850 A CN111737850 A CN 111737850A CN 202010411688 A CN202010411688 A CN 202010411688A CN 111737850 A CN111737850 A CN 111737850A
- Authority
- CN
- China
- Prior art keywords
- aod
- satellite
- fusion
- value
- aatsr
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000007500 overflow downdraw method Methods 0.000 title claims abstract description 49
- 230000004927 fusion Effects 0.000 claims abstract description 99
- 238000000034 method Methods 0.000 claims abstract description 42
- 238000004364 calculation method Methods 0.000 claims abstract description 37
- 238000007781 pre-processing Methods 0.000 claims abstract description 6
- 238000011160 research Methods 0.000 claims description 24
- 238000012795 verification Methods 0.000 claims description 16
- 230000003287 optical effect Effects 0.000 claims description 11
- 238000013441 quality evaluation Methods 0.000 claims description 11
- 238000011156 evaluation Methods 0.000 claims description 10
- 238000012952 Resampling Methods 0.000 claims description 6
- 238000006243 chemical reaction Methods 0.000 claims description 3
- 238000001303 quality assessment method Methods 0.000 claims 1
- 238000012544 monitoring process Methods 0.000 abstract description 5
- 239000000443 aerosol Substances 0.000 description 21
- 210000004027 cell Anatomy 0.000 description 9
- 238000010586 diagram Methods 0.000 description 8
- 241000282414 Homo sapiens Species 0.000 description 5
- 238000007476 Maximum Likelihood Methods 0.000 description 5
- 230000036541 health Effects 0.000 description 5
- 230000006872 improvement Effects 0.000 description 5
- 230000008569 process Effects 0.000 description 3
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 210000003719 b-lymphocyte Anatomy 0.000 description 1
- 230000008033 biological extinction Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 239000003344 environmental pollutant Substances 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 239000002245 particle Substances 0.000 description 1
- 231100000719 pollutant Toxicity 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000005855 radiation Effects 0.000 description 1
- 238000002310 reflectometry Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 230000009897 systematic effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Hardware Design (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a multisource satellite AOD fusion method based on uncertainty on a pixel scale, and belongs to the technical field of satellite remote sensing application. The method comprises the following steps: preprocessing a data set of a multi-source satellite AATSR AOD product; based on the preprocessed uncertainty data set of the multi-source satellite AATSR AOD product, constructing a unit pixel AATSR AOD fusion weight calculation model by using an entropy method, and calculating the fusion weight of the unit pixel AATSR AOD; and performing fusion operation by using a weighted average fusion method to obtain an AOD data set after multi-source satellite fusion. The fusion weight calculation model is constructed by using an entropy method, compared with a lookup table method, the fusion weight calculation model has the advantages of simple operation logic, high operation speed, high accuracy and strong stability, can meet the requirement of high space coverage rate, can ensure the quality of a fusion product, improves the application value of the fusion product, and provides important input data for atmospheric environment monitoring.
Description
Technical Field
The invention relates to the technical field of atmospheric satellite remote sensing application, in particular to a multisource satellite AOD fusion method based on uncertainty on a pixel scale.
Background
Aerosol Optical Depth (AOD) refers to the vertical integral of the Aerosol extinction coefficient from the ground to the top of the atmosphere, describing the attenuation of light by the Aerosol. The method is one of the most important parameters of the aerosol, is also one of the most important parameters of aerosol remote sensing research, and in a plurality of fields related to the aerosol, such as aerosol radiation compelling, atmospheric correction of remote sensing images and the like, the AOD is an important input parameter. The aerosol is a main pollutant in the atmosphere, and the PM2.5 particles can carry and transport harmful substances, thereby seriously threatening the health of human beings.
In order to produce AOD products with high consistency with ground observation data, scholars at home and abroad develop a plurality of targeted fusion algorithms, wherein the most widely applied method is a simple average fusion method and a weighted average fusion method. The simple average fusion method uses a simple mathematical average method, and pixel AOD values of fusion products are represented by values obtained by summing optical thickness values of unit pixel multi-source satellite aerosols participating in fusion and dividing the sum by the number of the multi-source satellite products. The simple weighted average fusion method is characterized in that on the basis of simple weighted average, factors such as earth surface reflectivity or aerosol type and the like which influence the quality of the aerosol inverted by the satellite are quantized into weight factors and then weighted fusion is carried out. However, although the calculation process of these conventional simple average fusion algorithms is simple, the error of the fusion result is large, and the accuracy of the fusion product cannot be guaranteed.
Furthermore, Mei et al (2007) completes the fusion of the MODIS and SeaWIFS data for the first time based on the least square method, and verifies the fusion result based on the ground observation data. Leptoukh et al (2007) adopt a maximum likelihood estimation method to perform fusion based on MODIS and MISR aerosol optical thickness data, and find that the coverage of the fused AOD product is obviously improved by comparing the maximum likelihood estimation method with aerosol optical thickness inverted by a single sensor. Chatterjee et al (2010) apply a geostatistical method to the development of a multi-source satellite aerosol fusion algorithm, complete the fusion based on MODIS AOD and MISR AOD, verify the fused AOD product by using AERONET foundation observation aerosol optical thickness data, and find that the fused product is closer to the foundation measurement data. Tang et al (2016) also proposed a Bayesian Maximum Entropy (BME) -based multisource aerosol fusion algorithm, which quantified the uncertainty factors affecting the fusion result by Bayesian maximum entropy method. Xie et al (2018) construct an uncertainty data set as a fusion weight through a lookup table method, and use a maximum likelihood method to fuse the multi-source satellite AOD, and the method has poor stability of the fusion weight calculated through interpolation, and cannot ensure the accuracy of a fusion result.
The least square fusion method and the maximum likelihood fusion method are mainly based on a lookup table method to construct fusion weights, and then the aerosol optical thickness product is fused by using a maximum likelihood estimation method or a least square fusion method. Although the lookup table method can improve the calculation speed to a certain extent, the construction process of the lookup table at the previous stage is time-consuming, and the parameters for establishing the lookup table are not universally applicable and portable. In actual operation calculation, the loading of the lookup table occupies a certain memory space, and the operation processing efficiency is also low. In addition, interpolation calculation in the application process of the lookup table is complex, different calculation results can be obtained by different interpolation methods, the algorithm has no stability, and the obtained fusion weight has high uncertainty, so that the fusion result has no application value.
From the above, the above algorithms do not take into account the uncertainty of the product on the pixel scale. And the advantages of different AOD data sets are fully utilized to produce more complete and accurate aerosol data sets with space-time consistency, and the systematic error and uncertainty of the pixel scale of the original aerosol data set need to be fully considered. Therefore, the multi-source satellite AOD fusion method based on uncertainty in the pixel scale is created, the uncertainty of the product in the pixel scale can be combined, the integrity, the accuracy and the space coverage rate of the multi-source satellite AOD product are improved, and the consistency of the multi-source satellite AOD product and observation data of a foundation station is good.
Disclosure of Invention
The technical problem to be solved by the invention is to provide a multisource satellite AOD fusion method based on uncertainty on a pixel scale, so that the fusion operation speed can be improved, the fusion weight can be ensured to have stable and reliable characteristics, the application value of a fusion product is improved, and higher space coverage rate is met, thereby overcoming the defects of the existing AOD product.
In order to solve the technical problem, the invention provides a multisource satellite AOD fusion method based on uncertainty on a pixel scale, which comprises the following steps:
(1) preprocessing a data set of a multi-source satellite AATSR AOD product;
(2) based on the uncertainty data set of the multi-source satellite AATSR AOD product preprocessed in the step (1), constructing a unit pixel AATSR AOD fusion weight calculation model by using an entropy method, and calculating the fusion weight of the unit pixel AATSR AOD;
(3) and (3) aiming at the fusion weight of the unit pixel AATSR AOD obtained by the calculation in the step (2), performing fusion operation by using a weighted average fusion method to obtain an AOD data set after the multi-source satellite is fused.
In a further improvement, the step (1) of preprocessing the data set of the multi-source satellite AATSR AOD product comprises the following steps:
A. data splicing: extracting longitude and latitude information of a multi-source satellite AATSR AOD product by using an ENVI IDL programming tool, and completing data splicing based on edge longitude and latitude information of each satellite data;
B. data clipping: recording the longitude and latitude of the upper left corner and the lower right corner of the research area, determining the longitude and latitude range of the research area, and cutting the spliced AATSR AOD by taking the longitude and latitude range of the research area as a mask;
C. resampling: uniformly resampling the spatial resolution of the multi-source satellite AATSR AOD product to 1km by 1 km;
D. projection conversion: and uniformly setting the projection mode of the multi-source AATSR AOD product into equal longitude and latitude projection.
Further improving, the fusion weight calculation model of the unit pixel AATSR AOD obtained in the step (2) is as follows:
wherein, wiRepresenting a fusion weight value of the multi-source satellite AOD; q. q.siRepresenting the ratio of any AATSR AOD uncertainty data to its corresponding AOD value, AODiRepresents the value of any AATSR AOD product at the unit pixel, uncartientaryiUncertainty data representing a satellite AOD at a unit pixel; n represents the number of effective AATSR AOD products at a unit pixel, and the value range of n is 0-3: n is 0, which indicates that the fusion AOD value of the pixel is 0; n is 1, which means that only one AATSR AOD product participates in the fusion operation at the pixel, and the fusion result is equal to the value of the satellite AOD; n is 2 or 3, which indicates that two or three AATSR AOD products participate in the fusion operation at the pixel, and the fusion weight values of the single AATSR AOD product at the pixel are respectively calculated at the moment.
In a further improvement, the fusion numerical calculation formula of the weighted average fusion method in the step (3) is as follows:
wherein m represents the number of AOD pixels participating in fusion; tau isiIndicates the unit pixel AOD value, wiDenotes the fusion weight value, τ, of the corresponding AODfusionThe fused AOD values are shown.
And (4) carrying out precision verification and quality evaluation on the AOD data set obtained in the step (3) after the multi-source satellite is fused by using the ground observation data.
Further improvement, in the step (4), accuracy verification is performed on the AOD data set after the multi-source satellite is fused by using a time-space matching verification method, and the specific method is as follows: taking longitude and latitude and observation time of the ground observation data as basic input parameters, matching and comparing and verifying a ground observation mean value within +/-30 minutes and a mean value of AOD (active optical device) of a satellite in a 5 x 5 pixel window around an observation station, and judging the usability of a matched result, wherein the requirements are as follows: and ensuring that at least two effective observation values exist in each foundation observation data within +/-30 minutes, and ensuring that no less than 5 effective satellite AOD values exist in each window when the average value of the satellite AOD is calculated in each 5 multiplied by 5 pixel window.
Further improvement, the quality evaluation is carried out on all successfully matched data in the precision verification method, the quality evaluation method comprises five evaluation indexes, and each evaluation index comprises excellent, good and general three quality standards; the five evaluation indexes are respectively:
the satellite AOD mean value MSA represents the mean value of satellite AOD values successfully matched with the ground observation AOD data;
the ground AOD mean value MAA represents the mean value of ground observation AOD values successfully matched with the satellite AOD data;
relative average deviation RMB, which represents the ratio of satellite AOD mean MSA to ground AOD mean MAA;
root mean square error RMSE, which represents the degree to which the satellite AOD mean deviates from the true value;
and the correlation coefficient R represents a statistical index of the correlation between the satellite AOD data and the ground-based AOD data, and the closer R is to 1, the better the correlation between the two variables is.
Further improved, the calculation formula of the satellite AOD mean MSA is as follows:
wherein q represents the effective AOD pixel number in a 5 × 5 pixel window, tausatllite，iRepresenting the effective satellite AOD value in the unit pixel;
the calculation formula of the foundation AOD mean value MAA is as follows:
wherein p represents the number of effective foundation observation AODs within +/-30 minutes; tau isAeronet，iExpressing the effective foundation observation AOD value in the unit pixel;
the calculation formula of the relative average deviation RMB is as follows:
RMB＝MSA/MAA
the RMB is used for representing the phenomena of overestimation and underestimation of the satellite AOD value, the general value range is 0-2, when the RMB is 1-2, the satellite AOD value is indicated to have the overestimation phenomenon, and when the RMB is 0-1, the satellite AOD value is indicated to have the underestimation phenomenon;
the calculation formula of the root mean square error RMSE is as follows:
wherein s represents the number of AODs successfully matched with the AOD data of the ground observation; tau issatllite，iA value representing the satellite AOD; tau isAeronetThe larger the RMSE value, the worse the quality of the satellite AOD value.
And (4) further improving that the index for performing quality evaluation on the AOD data set after the multi-source satellite fusion in the step (4) comprises a space coverage rate.
Further improved, the calculation method of the spatial coverage rate is as follows: using IDL programming to judge whether the AOD value of a unit pixel in a research area is greater than 0, if so, indicating that the pixel is covered by AOD; then using the formula:
and calculating the spatial Coverage ratio Coverage in the research area, wherein x is the number of pixels of which the AOD value of the pixels in the research area is greater than 0, and Y is the sum of the total number of the pixels in the research area.
After adopting such design, the invention has at least the following advantages:
the invention uses entropy method to construct fusion weight calculation model based on AATSR AOD uncertainty data set, compared with the existing method of constructing fusion weight by lookup table method, the invention has simple operation logic; the operation speed is high and is the same as that of the existing simple average fusion algorithm, but the accuracy of the fusion result is much higher than that of the simple weighted average fusion algorithm; and the stability is strong, the fusion weight evaluation accuracy is high, and the transportability is strong.
The multi-source satellite AOD product fusion method provided by the invention combines the spatial distribution characteristics of a plurality of satellite AODs, and provides important input data for monitoring the instantaneous change trend of the atmospheric environment and researching the influence of atmospheric pollution on human health. The multi-source satellite data fusion method not only meets the requirement of higher space coverage rate, but also ensures the quality of the fusion product and improves the application value of the fusion product while improving the space coverage rate of the fusion product.
Drawings
The foregoing is only an overview of the technical solutions of the present invention, and in order to make the technical solutions of the present invention more clearly understood, the present invention is further described in detail below with reference to the accompanying drawings and the detailed description.
FIG. 1 is a flow chart of the multi-source satellite AOD fusion method based on uncertainty on pixel scale of the present invention.
FIG. 2 is a schematic diagram of matching verification of multi-source AATSR AOD values and statistical data of a foundation observation station in the fusion method, wherein a cell A represents a unit pixel with longitude and latitude of the foundation observation station as a center and image resolution as side length, a cell B represents a satellite observation AOD observation result (such as 6) in a matching range, and a white cell represents a pixel without satellite AOD data.
FIG. 3 is a schematic diagram of multi-source AATSR AOD space coverage rate calculation in the fusion method of the present invention, wherein gray cells represent pixels covered by AOD, and white cells represent pixels not covered by AOD.
Fig. 4 is the spatial distribution of AERONET and CARSNET sites in the continental china in the fusion method of the present invention, the light gray spots and the dark gray spots represent AERONET and CARSNET sites, respectively.
FIG. 5 is a scatter plot of three AATSR ADV, ORAC, SU products in the fusion method of the present invention and the fused L2 AOD product of the present invention with ground based observation data. The solid line represents the 1-1 line and the scatter points represent the average of the AERONET and CARSNET AODs within the different AOD value domains.
FIG. 6 is a statistical chart of the average daily spatial coverage of three AATSR AODs in the fusion method of the present invention and fused AOD products.
FIG. 7 is a statistical graph of the monthly mean spatial coverage of three AATSR AODs in 2007 in the fusion method of the present invention and fused AOD products.
FIG. 8 is a graph of the monthly mean spatial coverage distribution of AATSR AODs and fused AOD products at 2007 in the fusion method of the present invention.
FIG. 9 is a quaternary mean spatial coverage statistical plot of the AATSR AOD and fused AOD products in 2007 in the fusion method of the present invention.
FIG. 10 is a graph of the season-averaged spatial coverage distribution of the AATSR AOD and fused AOD products of 2007 in the fusion method of the present invention.
FIG. 11 is a statistical chart of the annual average spatial coverage of the AATSR AOD and fused AOD products in 2007 in the fusion method of the present invention.
FIG. 12 is a graph of annual average spatial coverage distribution of AATSR AODs and fused AOD products in 2007 in the fusion method of the present invention.
Detailed Description
Referring to fig. 1, the multisource satellite AOD fusion method based on uncertainty on the pixel scale in this embodiment includes the following steps:
(1) and preprocessing the data set of the multi-source satellite AATSR AOD product.
This example is illustrated by three AATSR AOD datasets (ADV, ORAC, SU). Because the spatial resolution and the projection mode of the AOD products from different sources are different, the data set of the multi-source satellite AATSR AOD product needs to be preprocessed before the fusion operation is carried out. The pretreatment step comprises the following steps:
A. data splicing: extracting longitude and latitude information of a multi-source satellite AATSR AOD product by using an ENVI IDL programming tool, and completing data splicing based on edge longitude and latitude information of each satellite data;
B. data clipping: recording the longitude and latitude of the upper left corner and the lower right corner of the research area, determining the longitude and latitude range of the research area, and cutting the spliced AATSR AOD by taking the longitude and latitude range of the research area as a mask;
C. resampling: uniformly resampling the spatial resolution of the multi-source satellite AATSR AOD product to 1km by 1 km;
D. projection conversion: and uniformly setting the projection mode of the multi-source AATSR AOD product into equal longitude and latitude projection.
(2) And (2) constructing a unit pixel AATSR AOD fusion weight calculation model by using an entropy method based on the uncertainty data set of the multi-source satellite AATSR AOD product preprocessed in the step (1), and calculating the fusion weight of the unit pixel AATSR AOD.
The fusion weight calculation model of the unit pixel AATSR AOD is as follows:
wherein, wiRepresenting a fusion weight value of the multi-source satellite AOD; q. q.siRepresenting the ratio of any AATSR AOD uncertainty data to its corresponding AOD value, AODiDenotes the value of any one of AATSR AOD at the unit pixel, noncertinityiUncertainty data representing a satellite AOD at a unit pixel; n represents the number of effective AATSR AOD products at a unit pixel, and the value range of n is 0-3: n is 0, which indicates that the fusion AOD value of the pixel is 0; n is 1, which means that only one AATSR AOD participates in the fusion operation at the pixel, and the fusion result is equal to the value of the satellite AOD; n is 2 or 3, which indicates that two or three AATSR AODs participate in the fusion operation at the pixel, and the fusion weight values of a single AATSR AOD product at the pixel are calculated respectively at the moment.
(3) And (3) aiming at the fusion weight of the unit pixel AATSR AOD obtained by the calculation in the step (2), performing fusion operation by using a weighted average fusion method to obtain an AOD data set after the multi-source satellite is fused.
The fusion numerical calculation formula of the weighted average fusion method is as follows:
wherein m represents the number of AOD pixels participating in fusion; tau isiIndicates the unit pixel AOD value, wiDenotes the fusion weight value, τ, of the corresponding AODfusionThe fused AOD values are shown.
The embodiment also comprises a step of performing precision verification and quality evaluation on the multi-source satellite fused AOD data set obtained in the step (3) by using the ground observation data.
In the step, a space-time matching verification method is adopted for precision verification, and the specific method comprises the following steps: taking longitude and latitude and observation time of the ground observation data as basic input parameters, matching and comparing and verifying a ground observation mean value within +/-30 minutes and a mean value of AOD (active optical device) of a satellite in a 5 x 5 pixel window around an observation station, and judging the usability of a matched result, wherein the requirements are as follows: and ensuring that at least two effective observation values exist in each foundation observation data within +/-30 minutes, and ensuring that no less than 5 effective satellite AOD values exist in each window when the average value of the satellite AOD is calculated in each 5 multiplied by 5 pixel window.
As shown in fig. 2, which is a schematic diagram of a spatiotemporal contrast matching method, in the diagram, a cell a is a unit pixel with longitude and latitude of a ground observation station as a center and image resolution as side length, and an AOD assignment rule of the pixel is as follows: counting all AOD observation records within +/-30 minutes of the site, summing all the observation AOD records, and dividing the sum by the observation times to obtain:
wherein p represents the total number of records of the foundation observation AOD within +/-30 minutes, and in order to increase the verification accuracy, p is required to be more than or equal to 2, namely more than two records of the foundation observation AOD in a unit time window are required.
Fig. 2 further shows that the 5 × 5 cells shown represent a matching range with satellite observation data centered on the ground observation station, and B cells in the figure represent satellite observation AOD observations within the matching range, so that the matching mean of the satellite AODs is:
wherein q represents the number of satellite observation AODs in the matching range, and in order to increase the verification accuracy, q is required to be more than or equal to 5, namely the number of the satellite AOD inversion results in a 5 x 5 space window with the longitude and latitude of a foundation station as the center is more than 5.
The present embodiment also performs quality evaluation on all the successfully matched data in the accuracy verification method, and the quality evaluation method includes five evaluation indexes, each evaluation index includes three quality standards Q1-Q3, Q1 is excellent, Q2 is good, and Q3 is general. The five evaluation indexes are respectively:
the method comprises the following steps that firstly, the Mean Satellite AOD (Mean Satellite-retrieved AOD) MSA represents the Mean value of Satellite AOD values successfully matched with ground observation AOD data; the calculation formula is as follows:
wherein q represents the effective AOD pixel number in a 5 × 5 pixel window, tausatllite，iRepresenting the effective satellite AOD value in a unit pixel element. The larger the MSA value, the higher the mean value of satellite AOD, and vice versa.
Secondly, a ground AOD Mean value (Mean AERONET or CARSNET AOD) MAA represents the Mean value of ground observation AOD values successfully matched with satellite AOD data; the calculation formula is as follows:
wherein p represents the number of effective AERONET or CARSNETAOD within + -30 minutes; tau isAeronet，iRepresenting the effective AERONE or CARSNET AOD value within a unit pixel. The larger the value of MAA is, the higher the mean value of the AOD observed on the foundation is, and vice versa.
Thirdly, Relative average deviation RMB (Relative Mean Bias) represents the ratio of the satellite AOD Mean value MSA to the ground AOD Mean value MAA; the calculation formula is as follows:
RMB＝MSA/MAA (8)
the RMB is used for representing the phenomena of overestimation and underestimation of the satellite AOD value, the general value range is 0-2, when the RMB is 1-2, the satellite AOD value is overestimated, and the phenomenon of overestimation of the satellite AOD is more serious when the RMB is larger. When the RMB is between 0 and 1, the satellite AOD value is underestimated. The smaller the RMB, the more severe the underestimation of the satellite AOD.
Root-Mean-Square Error (RMSE) representing the degree of deviation of the satellite AOD Mean value from a true value; the calculation formula is as follows:
wherein s represents the number of AODs successfully matched with the AOD data of the ground observation; tau issatllite，iA value representing the satellite AOD; tau isAeronetThe larger the RMSE value, the worse the quality of the satellite AOD value.
And a correlation coefficient R represents a statistical index of a correlation relationship between the satellite AOD data and the ground AOD data, wherein the closer R is to 1, the better the correlation between the two variables is.
Also, spatial coverage of the AOD product is one of the reference criteria for evaluating the quality of the AOD product. The AOD product with high coverage rate has important application significance in the fields of atmospheric monitoring and human health. The fusion method further comprises the step of evaluating the quality of the AOD data set after the multi-source satellite is fused by adopting the space coverage rate index. Within the same study area, the larger the spatial coverage of the AOD product, the better the quality of the product.
The calculation method of the space coverage rate comprises the following steps: using IDL programming to judge whether the AOD value of a unit pixel in a research area is greater than 0, if so, indicating that the pixel is covered by AOD; then using the formula:
and calculating the spatial Coverage ratio Coverage in the research area, wherein x is the number of pixels of which the AOD value of the pixels in the research area is greater than 0, and Y is the sum of the total number of the pixels in the research area.
As shown in fig. 3, which shows the principle diagram of AATSR AOD spatial Coverage calculation, a gray cell represents a pixel covered by AOD, for example, 15, and a white cell represents a pixel not covered by AOD, for example, 25, then the spatial Coverage in the research area is equal to 15/(15+25) ═ 37.5%.
Referring to fig. 4-12, the results of performing precision verification and spatial coverage quality evaluation on the fusion AOD product and the single remote sensing satellite AOD product of the present invention by using ground-based observation AOD data are as follows.
As shown in the attached FIG. 5, the root mean square error RMSE of ADV, ORAC and SU of the three AATSR AOD products is 0.18, 0.14 and 0.14 respectively, while the root mean square error RMSE of the fusion AOD product (Fus ion AOD) of the invention is 0.17. The root mean square error of the fusion AOD product of the invention is not different from the root mean square error of the single remote sensing satellite AOD product, the high consistency is kept, and the precision requirement of the fusion product can be met.
As can be seen from the statistical chart of the daily average spatial coverage of the three AATSR AODs and the fused AOD product shown in the attached figure 6, the daily average spatial coverage of the fused AOD product (Fusion AOD) is respectively 43.2%, 39.1% and 39.2% higher than that of ADV, ORAC and SU products, and the improvement range is very large.
From the statistical graph of the monthly mean spatial coverage of the three AATSR AODs and the fused AOD product in 2007 shown in the attached FIG. 7, it can be seen that the monthly mean spatial coverage of the fused AOD product (Fusion AOD) is respectively improved by 20%, 18% and 18% compared with that of ADV, ORAC and SU products. And as can be seen from the distribution diagram of the monthly mean spatial coverage of the three AATSR AOD products of 2007, 1 month, 4 months, 7 months and 10 months shown in fig. 8 and the fused AOD product of the present invention, the monthly mean spatial coverage of the fused AOD product (Fusion AOD) of the present invention is greatly improved compared with the monthly mean spatial coverage of ADV, ORAC and SU products.
From the statistical chart of the quaternary mean spatial coverage of the three AATSR AODs and the fused AOD product in 2007 shown in the attached FIG. 9, it can be seen that the quaternary mean spatial coverage of the fused AOD product (Fusion AOD) is respectively 15.2%, 12.5% and 7.5% higher than that of ADV, ORAC and SU products. And as can be seen from the distribution diagram of the quaternary mean spatial coverage of the three AATSR AOD products in 4 seasons and the fused AOD product of the invention shown in the attached FIG. 10, the quaternary mean spatial coverage of the fused AOD product (fusion AOD) of the invention is improved compared with that of ADV, ORAC and SU products.
From the statistical chart of the annual average spatial coverage of the three AATSR AODs in 2007 and the fused AOD product shown in the attached figure 11, the annual average spatial coverage of the fused AOD product (Fusion AOD) is improved by 6.9%, 5.7% and 2.1% respectively compared with the annual average spatial coverage of ADV, ORAC and SU products. And as can be seen from the distribution diagram of the annual average spatial coverage of the three AATSR AOD products in 2007 shown in the attached figure 12 and the fused AOD product of the invention, the annual average spatial coverage of the fused AOD product (Fusion AOD) of the invention is improved compared with that of ADV, ORAC and SU products.
As can be seen from the attached figures 5-12, compared with three single remote sensing satellites AATSR AOD, the spatial coverage rate of the fused AOD product is improved. Especially the daily average and monthly average time resolution. Therefore, compared with a satellite aerosol optical thickness product inverted by a single sensor, the fused product obtained by the multisource satellite AOD fusion method based on uncertainty on the pixel scale not only improves the spatial coverage, but also ensures the reliability of the quality of the fused product, and can play an important role in the fields of atmospheric monitoring and human health.
The multisource satellite AOD fusion method based on uncertainty on the pixel dimension makes up the limitation of AOD data provided by a single remote sensing sensor in the space coverage range and cannot provide a product with high space coverage, and limits the fusion error to the pixel dimension by constructing a pixel dimension uncertainty expression model, so that the precision and reliability of the fusion product are improved, a brand-new aerosol optical thickness product with high consistency, high precision and wider space coverage range is created, and support data are provided for atmosphere monitoring and human health research.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the present invention in any way, and it will be apparent to those skilled in the art that the above description of the present invention can be applied to various modifications, equivalent variations or modifications without departing from the spirit and scope of the present invention.
Claims (10)
1. A multisource satellite AOD fusion method based on uncertainty on a pixel scale is characterized by comprising the following steps:
(1) preprocessing a data set of a multi-source satellite AATSR AOD product;
(2) based on the uncertainty data set of the multi-source satellite AATSR AOD product preprocessed in the step (1), constructing a unit pixel AATSR AOD fusion weight calculation model by using an entropy method, and calculating the fusion weight of the unit pixel AATSR AOD;
(3) and (3) aiming at the fusion weight of the unit pixel AATSR AOD obtained by the calculation in the step (2), performing fusion operation by using a weighted average fusion method to obtain an AOD data set after the multi-source satellite is fused.
2. The multi-source satellite AOD fusion method based on uncertainty on pixel scale according to claim 1, wherein the step of preprocessing the dataset of multi-source satellite AATSR AOD products in step (1) comprises:
A. data splicing: extracting longitude and latitude information of a multi-source satellite AATSR AOD product by using an ENVI IDL programming tool, and completing data splicing based on edge longitude and latitude information of each satellite data;
B. data clipping: recording the longitude and latitude of the upper left corner and the lower right corner of the research area, determining the longitude and latitude range of the research area, and cutting the spliced AATSR AOD by taking the longitude and latitude range of the research area as a mask;
C. resampling: uniformly resampling the spatial resolution of the multi-source satellite AATSR AOD product to 1km by 1 km;
D. projection conversion: and uniformly setting the projection mode of the multi-source AATSR AOD product into equal longitude and latitude projection.
3. The multi-source satellite AOD fusion method based on uncertainty on pixel dimension according to claim 2, wherein the fusion weight calculation model of unit pixel AATSR AOD obtained in step (2) is:
wherein, wiRepresenting a fusion weight value of the multi-source satellite AOD; q. q.siRepresenting the ratio of any AATSR AOD uncertainty data to its corresponding AOD value, AODiRepresents the value of any AATSR AOD product at the unit pixel, uncartientaryiUncertainty data representing a satellite AOD at a unit pixel; n represents the number of effective AATSR AOD products at a unit pixel, and the value range of n is 0-3: n is 0, which indicates that the fusion AOD value of the pixel is 0; n is 1, which means that only one AATSR AOD product participates in the fusion operation at the pixel, and the fusion result is equal to the value of the satellite AOD; n is 2 or 3, which indicates that two or three AATSR AOD products participate in the fusion operation at the pixel, and the fusion weight values of the single AATSR AOD product at the pixel are respectively calculated at the moment.
4. The multi-source satellite AOD fusion method based on uncertainty on pixel scale according to claim 3, wherein the fusion numerical calculation formula of the weighted average fusion method in the step (3) is:
wherein m represents the number of AOD pixels participating in fusion; tau isiIndicates the unit pixel AOD value, wiDenotes the fusion weight value, τ, of the corresponding AODfusionThe fused AOD values are shown.
5. The multisource satellite AOD fusion method based on uncertainty on pixel scale according to any one of claims 1 to 4, characterized by further comprising the step (4) of performing precision verification and quality evaluation on the multisource satellite fused AOD dataset obtained in the step (3) by using ground-based observation data.
6. The multisource satellite AOD fusion method based on uncertainty on pixel dimension of claim 5, wherein in the step (4), the AOD data set after multisource satellite fusion is subjected to precision verification by adopting a time-space matching verification method, and the specific method is as follows: taking longitude and latitude and observation time of the ground observation data as basic input parameters, matching and comparing and verifying a ground observation mean value within +/-30 minutes and a mean value of AOD (active optical device) of a satellite in a 5 x 5 pixel window around an observation station, and judging the usability of a matched result, wherein the requirements are as follows: and ensuring that at least two effective observation values exist in each foundation observation data within +/-30 minutes, and ensuring that no less than 5 effective satellite AOD values exist in each window when the average value of the satellite AOD is calculated in each 5 multiplied by 5 pixel window.
7. The multi-source satellite AOD fusion method based on uncertainty on pixel scale according to claim 6, wherein the quality evaluation is performed on all successfully matched data in the precision verification method, the quality evaluation method comprises five evaluation indexes, and each evaluation index comprises excellent, good and general three quality standards; the five evaluation indexes are respectively:
the satellite AOD mean value MSA represents the mean value of satellite AOD values successfully matched with the ground observation AOD data;
the ground AOD mean value MAA represents the mean value of ground observation AOD values successfully matched with the satellite AOD data;
relative average deviation RMB, which represents the ratio of satellite AOD mean MSA to ground AOD mean MAA;
root mean square error RMSE, which represents the degree to which the satellite AOD mean deviates from the true value;
and the correlation coefficient R represents a statistical index of the correlation between the satellite AOD data and the ground-based AOD data, and the closer R is to 1, the better the correlation between the two variables is.
8. The multi-source satellite AOD fusion method based on uncertainty on a pixel scale according to claim 7, wherein the calculation formula of the satellite AOD mean MSA is as follows:
wherein q represents the effective AOD pixel number in a 5 × 5 pixel window, tausatllite，iRepresenting the effective satellite AOD value in the unit pixel;
the calculation formula of the foundation AOD mean value MAA is as follows:
wherein p represents the number of effective foundation observation AODs within +/-30 minutes; tau isAeronet，iExpressing the effective foundation observation AOD value in the unit pixel;
the calculation formula of the relative average deviation RMB is as follows:
RMB＝MSA/MAA
the RMB is used for representing the phenomena of overestimation and underestimation of the satellite AOD value, the general value range is 0-2, when the RMB is 1-2, the satellite AOD value is indicated to have the overestimation phenomenon, and when the RMB is 0-1, the satellite AOD value is indicated to have the underestimation phenomenon;
the calculation formula of the root mean square error RMSE is as follows:
wherein s represents the number of AODs successfully matched with the AOD data of the ground observation; tau issatllite，iA value representing the satellite AOD; tau isAeronetThe larger the RMSE value is, the larger the value isThe worse the quality of the star AOD values.
9. The multi-source satellite AOD fusion method based on uncertainty on a pixel scale according to claim 5, wherein the index for quality assessment of the multi-source satellite fused AOD dataset in step (4) comprises spatial coverage.
10. The multi-source satellite AOD fusion method based on uncertainty on pixel scale according to claim 9, wherein the spatial coverage is calculated by: using IDL programming to judge whether the AOD value of a unit pixel in a research area is greater than 0, if so, indicating that the pixel is covered by AOD; then using the formula:
and calculating the spatial Coverage ratio Coverage in the research area, wherein x is the number of pixels of which the AOD value of the pixels in the research area is greater than 0, and Y is the sum of the total number of the pixels in the research area.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010411688.8A CN111737850B (en) | 2020-05-15 | 2020-05-15 | Multi-source satellite AOD fusion method based on uncertainty on pixel scale |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010411688.8A CN111737850B (en) | 2020-05-15 | 2020-05-15 | Multi-source satellite AOD fusion method based on uncertainty on pixel scale |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111737850A true CN111737850A (en) | 2020-10-02 |
| CN111737850B CN111737850B (en) | 2024-04-19 |
Family
ID=72647276
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010411688.8A Active CN111737850B (en) | 2020-05-15 | 2020-05-15 | Multi-source satellite AOD fusion method based on uncertainty on pixel scale |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111737850B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112884342A (en) * | 2021-03-10 | 2021-06-01 | 陕西九州遥感信息技术有限公司 | Water color satellite atmospheric layer top radiation product quality evaluation and cross calibration method |
| CN113435512A (en) * | 2021-06-28 | 2021-09-24 | 内蒙古师范大学 | Multi-satellite multi-source XCO2Data fusion method and system |
| CN114943303A (en) * | 2022-06-16 | 2022-08-26 | 福州大学 | Time sequence AOD reconstruction method based on multi-sensor remote sensing |
| CN115753687A (en) * | 2022-12-23 | 2023-03-07 | 西安工程大学 | Atmospheric pollution remote sensing monitoring method, system and computer readable medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5736959A (en) * | 1991-10-28 | 1998-04-07 | Teledesic Corporation | Earth-fixed cell beam management for satellite communication system using dielectic lens-focused scanning beam antennas |
| CN109213964A (en) * | 2018-07-13 | 2019-01-15 | 中南大学 | A kind of satellite AOD product bearing calibration for merging multi-source feature geographic factor |
| US20200082151A1 (en) * | 2018-09-06 | 2020-03-12 | National Central University | Method of Top-of-Atmosphere Reflectance-Based Spatiotemporal Image Fusion Using Aerosol Optical Depth |
| CN110909309A (en) * | 2019-11-21 | 2020-03-24 | 中国科学院遥感与数字地球研究所 | Method for acquiring hourly high-resolution PM2.5 data |
-
2020
- 2020-05-15 CN CN202010411688.8A patent/CN111737850B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5736959A (en) * | 1991-10-28 | 1998-04-07 | Teledesic Corporation | Earth-fixed cell beam management for satellite communication system using dielectic lens-focused scanning beam antennas |
| CN109213964A (en) * | 2018-07-13 | 2019-01-15 | 中南大学 | A kind of satellite AOD product bearing calibration for merging multi-source feature geographic factor |
| US20200082151A1 (en) * | 2018-09-06 | 2020-03-12 | National Central University | Method of Top-of-Atmosphere Reflectance-Based Spatiotemporal Image Fusion Using Aerosol Optical Depth |
| CN110909309A (en) * | 2019-11-21 | 2020-03-24 | 中国科学院遥感与数字地球研究所 | Method for acquiring hourly high-resolution PM2.5 data |
Non-Patent Citations (3)
| Title |
|---|
| XU, H.等: "A consistent aerosol optical depth (AOD) dataset over mainland China by integration of several AOD products", ATMOSPHERIC ENVIRONMENT, vol. 114, pages 48 - 56, XP029238387, DOI: 10.1016/j.atmosenv.2015.05.023 * |
| 方新;邹滨;刘宁;: "不确定性约束下的AOD空间覆盖STRE建模优化", 武汉大学学报(信息科学版), no. 04, pages 63 - 67 * |
| 李龙;施润和;张璐;张颉;周聪;徐彦平;高炜;: "华东地区MODIS与OMI气溶胶光学厚度数据融合", 地球信息科学学报, no. 10, pages 94 - 103 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112884342A (en) * | 2021-03-10 | 2021-06-01 | 陕西九州遥感信息技术有限公司 | Water color satellite atmospheric layer top radiation product quality evaluation and cross calibration method |
| CN112884342B (en) * | 2021-03-10 | 2024-03-12 | 陕西九州遥感信息技术有限公司 | Quality evaluation and cross calibration method for water-color satellite atmospheric roof radiation product |
| CN113435512A (en) * | 2021-06-28 | 2021-09-24 | 内蒙古师范大学 | Multi-satellite multi-source XCO2Data fusion method and system |
| CN114943303A (en) * | 2022-06-16 | 2022-08-26 | 福州大学 | Time sequence AOD reconstruction method based on multi-sensor remote sensing |
| CN115753687A (en) * | 2022-12-23 | 2023-03-07 | 西安工程大学 | Atmospheric pollution remote sensing monitoring method, system and computer readable medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111737850B (en) | 2024-04-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111737850A (en) | Multisource satellite AOD fusion method based on uncertainty on pixel scale | |
| US20220043182A1 (en) | Spatial autocorrelation machine learning-based downscaling method and system of satellite precipitation data | |
| CN112905560B (en) | Air pollution prediction method based on multi-source time-space big data deep fusion | |
| Birrer et al. | Line-of-sight effects in strong lensing: putting theory into practice | |
| CN116297068B (en) | Aerosol optical thickness inversion method and system based on earth surface reflectivity optimization | |
| CN114091756B (en) | Township tsunami risk assessment method based on Thiessen polygon | |
| Stovall et al. | TLSLeAF: automatic leaf angle estimates from single‐scan terrestrial laser scanning | |
| CN114693281B (en) | Engineering investigation information management system based on cloud platform | |
| Griffiths et al. | Demonstrating 24-hour continuous vertical monitoring of atmospheric optical turbulence | |
| WO2024036739A1 (en) | Reservoir water reserve inversion method and apparatus | |
| CN110503137A (en) | Based on the determination method of the remote sensing image temporal-spatial fusion base image pair of mixing together | |
| CN111079835B (en) | Himapari-8 atmospheric aerosol inversion method based on deep full-connection network | |
| Ma et al. | Characterizing the three-dimensional spatiotemporal variation of forest photosynthetically active radiation using terrestrial laser scanning data | |
| CN111650128B (en) | High-resolution atmospheric aerosol inversion method based on surface reflectivity library | |
| CN112949989B (en) | InSAR micro-deformation cultural heritage influence quantitative characterization method | |
| CN113962426A (en) | Offshore water navigation safety intelligent forecasting method and device | |
| Molenar et al. | Visual air quality simulation techniques | |
| CN115392128A (en) | Method for simulating basin runoff by utilizing space-time convolution LSTM network | |
| Zhang et al. | Incorporating dynamic factors for improving a GIS‐based solar radiation model | |
| CN114020725B (en) | Window sliding GPM data correction method considering spatial distribution | |
| CN117805826B (en) | Minute precipitation estimation method and system based on MIM network and radar jigsaw | |
| Ma et al. | Modelling the ZR Relationship of Precipitation Nowcasting Based on Deep Learning | |
| CN117347282B (en) | Star-based aerosol optical thickness inversion method, device and system and storage medium | |
| Weenink | Local spatio-temporal regression kriging for property price predictions | |
| CN118036238A (en) | Aerosol optical thickness inversion method and electronic equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |