CN103954974A - Particulate matter optical thickness remote sensing monitoring method used in urban area - Google Patents

Abstract

The invention discloses a particulate matter optical thickness remote sensing monitoring method used in an urban area. The method includes the following steps that preprocessing, including geometric correction, cloud recognition and atmospheric correction of a clear image, of MODIS data is carried out; surface reflectance of an image to be inversed is obtained according to geometry parameters, obtained after preprocessing, of a clear image and the image to be inversed and according to a BRDF model built in a coupled mode in the urban area; an improved structure function method is put forward according to the special surface structure of the urban area, and a structure function value of the surface reflectance and a structure function vale of apparent reflectance of the image to be inversed are calculated; finally, according to a geometry condition of the image to be inversed, a look-up table of inversion particulate matter optical thicknesses is built, and an aerosol optical thickness is looked up according to the structure function value of the surface reflectance and the structure function value of the apparent reflectance of the image. According to the improved structure function calculation method, influences of matching errors of multiple images on a calculation result can be effectively reduced, higher stability is achieved in the urban area, and inversion accuracy of the particulate matter optical thickness is improved.

Description

A kind of particle optical thickness remote-sensing monitoring method for urban area

Technical field

The present invention relates to a kind of particle optical thickness remote-sensing monitoring method for urban area.

Background technology

City is people's accumulation area, and the particle of urban area has important impact to people's life.Current, what use remote sensing monitoring particle was maximum is the optical thickness of monitoring particle.For the particle optical thickness inverting of land, dense vegetation algorithm is comparative maturity, but can only be applicable to the dense vegetation area that red blue wave band Reflectivity for Growing Season is lower.Urban area, most of earth's surface is all higher at the reflectivity of visible light wave range, for the lower image of spatial resolution, is difficult to find dense vegetation pixel in city, has limited dense vegetation method in the application of urban area.For the inverting of the particle optical thickness in the higher area of Reflectivity for Growing Season, Tanr é etc. has proposed the structure function method (or claiming contrast algorithm) based on image blurring effect.Dark goal method obtains particle information based on path radiation term, and structure function rule is the information of particle of obtaining based on atmospheric transmittance.

Structure function inversion method particle optical thickness is mainly the method based on atmospheric transmittance, and the particle optical thickness obtaining is to take the ratio of transmitance of multiple image to be basis.The variation of transmitance is to have the pixel in a specific range of distance to decide.Because zones of different has its specific space distribution structure therefore structure function adopts the variation of Reflectivity for Growing Season to weigh reflectivity rate of change spatially.During application structure functional based method inverting particle optical thickness, need be with the rate of change of the Reflectivity for Growing Season of reference picture in contrast,, therefore need to calculate to reference picture the structure function value M of the real Reflectivity for Growing Season rate of change after atmospheric correction _p(d, t ₁), and calculate the structure function value of the rate of change of the apparent reflectance treat inversion chart picture and:

$M_{p^{*}}(d,t_2)=M_p(d,t_2)\frac{T({\mathrm{\τ}}_a,{\mathrm{\μ}}_s)F_d({\mathrm{\τ}}_a,{\mathrm{\μ}}_s)}{1-\mathrm{AS}\left({\mathrm{\τ}}_a\right)}---\left(1\right)$

In formula, M _ρ(d, t ₂) for treating the earth's surface real reflectance of inversion chart picture; T is total atmospheric spectral transmittance; τ _afor the optical thickness of atmosphere, comprise gas molecule optical thickness and particle particle optics thickness; μ _εcosine value for solar zenith angle; F _dfor descending built-up radiation; A is the average albedo of object; S is atmosphere hemispherical reflectance.

Structure function method has been applied to the inverting of the data particle optical thicknesses such as TM, AVHRR, theoretically, by a pixel in computed image and it, closes on the Reflectivity for Growing Season difference △ ρ of certain pixel _i,jwith spoke luminance difference (or apparent reflectance difference ) relation just can obtain the particle optical thickness of this area, but in practical application, due to the impact of the matching precision between different images, △ ρ _i,jwith (or ) locus not quite identical, this brings very large error to inversion result.For reducing the impact on inversion accuracy of the differences in spatial location brought due to images match precision, for different satellite data types and the provincial characteristics in inverting area, researchist has proposed different computing method, and definition structure function is:

$M^2\left(d\right)=\frac{1}{n(m-d)}\underset{i=1}{\overset{n}{\mathrm{\Σ}}}\underset{j=1}{\overset{m-d}{\mathrm{\Σ}}}{({\mathrm{\ρ}}_{i,j}-{\mathrm{\ρ}}_{i,j+d})}^2---\left(2\right)$

The method, for the particle optical thickness in AVHRR inverting arid and semi-arid area, has obtained good efficiency of inverse process.Gin-Rong Liu etc. (2002) are improved to structure function definition:

$M^2\left(d\right)\frac{1}{3(n-d)(m-d)}\underset{i=1}{\overset{n-d}{\mathrm{\Σ}}}\underset{j=1}{\overset{m-d}{\mathrm{\Σ}}}[{({\mathrm{\ρ}}_{i,j}-{\mathrm{\ρ}}_{i,j+d})}^2+{({\mathrm{\ρ}}_{i,j}-{\mathrm{\ρ}}_{i+d,j})}^2+{({\mathrm{\ρ}}_{i,j}-{\mathrm{\ρ}}_{i+d,j+d})}^2]---\left(3\right)$

For these two kinds of methods, first method is certain pixel and closes on pixel contrast with a line fixed range, second method is that the pixel that closes on of certain pixel and its same a line, same row and 45 degree direction constant spacings contrasts, for first method, second method has higher stability, under stronger earth's surface change condition, can keep relative stability.But these two kinds of method d value in all directions is fixed, and for the special area of this class, city, the otherness on earth's surface is very obvious, and the setting of different d values has larger difference to the result of calculation of structure function, and error rate is higher.

Therefore, prior art needs further improvement and develops.

Summary of the invention

In view of above-mentioned the deficiencies in the prior art, the present invention proposes a kind of particle optical thickness remote-sensing monitoring method for urban area, improves the accuracy of particle optical thickness inverting.

For solving the problems of the technologies described above, the present invention program comprises:

For a particle optical thickness remote-sensing monitoring method for urban area, it comprises the following steps:

The pre-service of A, MODIS data, comprises geometric correction, cloud identification, the Atmospheric Correction of picture rich in detail;

B, according to the geometric parameter for the treatment of the picture rich in detail that obtains in inversion chart picture and steps A, the urban area BRDF model of coupling structure, obtains the Reflectivity for Growing Season for the treatment of inversion chart picture;

C, for this special surface infrastructure of urban area, improved structure function method is proposed, calculate and to treat the Reflectivity for Growing Season structure function value of inversion chart picture and the structure function value of apparent reflectance;

D, according to the geometric condition for the treatment of inverting image, build the look-up table of inverting particle optical thickness, by treating that the Reflectivity for Growing Season of inverting image and the structure function value of apparent reflectance search its particle optical thickness.

Described particle optical thickness remote-sensing monitoring method, wherein, also comprises geometric correction step in described steps A: first utilize longitude and latitude data in MODIS data as reference mark, meanwhile, by interpolation algorithm, calculate the actual longitude and latitude data of each pixel; Then for the data of each scanning strip, carry out geometric correction, finally the geometric correction result of all scanning strips has been spliced to geometric correction.

Described particle optical thickness remote-sensing monitoring method, wherein, also comprises picture rich in detail Atmospheric Correction step in described steps A: the radiation value L (μ being obtained by satellite sensor _v) can be expressed from the next:

${\mathrm{\ρ}}_t=\frac{L\left({\mathrm{\μ}}_v\right)-L_0\left({\mathrm{\μ}}_v\right)}{F_{d}T\left({\mathrm{\μ}}_v\right)+S(L\left({\mathrm{\μ}}_v\right)-L_0\left({\mathrm{\μ}}_v\right))}---\left(4\right)$

L (μ in formula (4) _v) be the radiance that sensor receives, ρ _tapparent reflectance, L ₀(μ _v) be path radiation term, F _d=u _sf ₀t(u _s) be the descending built-up radiation of the sun, F0 is the solar irradiance on atmospheric envelope top); the transmitance between sensor and target, direct projection transmitance, t ' _d(θ _v) be scattering transmitance; Under known observation condition, set one group of ρ _tvalue and corresponding sensor, obtain one group of radiance L (μ by MODTRAN4 or the simulation of 6S radiative transfer model _v), and obtain parameter path radiation term, transmitance, atmosphere hemisphere albedo and the descending built-up radiation of the sun of picture rich in detail Atmospheric Correction, will in above-mentioned parameter substitution formula (4), carry out picture rich in detail Atmospheric Correction.

Described particle optical thickness remote-sensing monitoring method, wherein, also comprises cloud identification step in described steps A: once read in the band class information of the MODIS data of image series, then pointwise detects, and finally generates cloud identification document.

Described particle optical thickness remote-sensing monitoring method, wherein, also comprises the structure function computing method that propose for this complicated earth surface of urban area in described step C:

$M^2\left(d\right)=\frac{1}{{(d_{\max }-d_{\min })}^2(n-d_{\min })(m-d_{\min })}\underset{i=1}{\overset{n-d}{\mathrm{\Σ}}}\underset{j=1}{\overset{m-d}{\mathrm{\Σ}}}\underset{d_j=d_{\min }}{\overset{d_j=d_{\max }}{\mathrm{\Σ}}}\underset{d_i=d_{\min }}{\overset{d_i=d_{\max }}{\mathrm{\Σ}}}\left[{({\mathrm{\ρ}}_{i,j}-{\mathrm{\ρ}}_{i+d_i,j+d_j})}^2\right]---\left(5\right)$

Wherein m*n is the size of calculation window, d _max, d _minthe maximum, the minimum pixel distance that refer to the pixel spacing that participates in calculating, ρ is reflectivity.

A kind of particle optical thickness remote-sensing monitoring method for urban area provided by the invention, carries out pre-service to the MODIS data of image series, and then computation structure functional value, according to structure function value, build the look-up table of inverting particle optical thickness, according to the geometric parameter of different images in image series, build the look-up table of particle optical thickness, can effectively reduce the impact on result of calculation of matching error due to multiple image, in urban area, there is higher stability, according to the structure function value of clean image, the earth's surface two that coupling builds is to Reflectivity Model, calculate the structure function value of the Reflectivity for Growing Season for the treatment of inversion chart picture and the structure function value of apparent reflectance (or structure function value of radiation value), according to the equation relation of the two, can obtain the particle optical thickness for the treatment of inversion chart picture, empirical tests, the measurement result of the environmental pollution monitoring result of inversion result and environmental monitoring station and the AERONET of this area survey station has good consistance.

Accompanying drawing explanation

Fig. 1 is the schematic flow sheet of particle optical thickness remote-sensing monitoring method in the present invention;

Fig. 2 is the schematic flow sheet of an embodiment in the present invention;

Fig. 3 is the comparison diagram of traditional detection method and detection method of the present invention;

Fig. 4 is the maximum error figure of the Reflectivity for Growing Season structure function value of distinct methods calculating;

Fig. 5 is the root-mean-square error figure of the Reflectivity for Growing Season structure function value of distinct methods calculating.

Embodiment

The invention provides a kind of particle optical thickness remote-sensing monitoring method for urban area, for making object of the present invention, technical scheme and effect clearer, clear and definite, below the present invention is described in more detail.Should be appreciated that specific embodiment described herein, only in order to explain the present invention, is not intended to limit the present invention.

The invention provides a kind of particle optical thickness remote-sensing monitoring method for urban area, as shown in Figure 1, it comprises the following steps:

The pre-service of step 101:MODIS data, comprises geometric correction, cloud identification, the Atmospheric Correction of picture rich in detail;

Step 102: according to the geometric parameter of the picture rich in detail for the treatment of to obtain in inversion chart picture and step 101, the urban area BRDF model that coupling builds, obtains the Reflectivity for Growing Season for the treatment of inversion chart picture;

Step 103: for this special surface infrastructure of urban area, propose improved structure function method, calculate and treat the Reflectivity for Growing Season structure function value of inversion chart picture and the structure function value of apparent reflectance;

Step 104: build the look-up table of inverting particle optical thickness according to the geometric condition for the treatment of inverting image, by treating that the Reflectivity for Growing Season of inverting image and the structure function value of apparent reflectance search its particle optical thickness.

Further, in described step 101, also comprise geometric correction step: first utilize longitude and latitude data in MODIS data as reference mark, meanwhile, by interpolation algorithm, calculate the actual longitude and latitude data of each pixel; Then for the data of each scanning strip, carry out geometric correction, finally the geometric correction result of all scanning strips has been spliced to geometric correction.

Described particle optical thickness remote-sensing monitoring method, wherein, also comprises picture rich in detail Atmospheric Correction step in described steps A: the radiation value L (μ being obtained by satellite sensor _v) can be expressed from the next:

${\mathrm{\ρ}}_t=\frac{L\left({\mathrm{\μ}}_v\right)-L_0\left({\mathrm{\μ}}_v\right)}{F_{d}T\left({\mathrm{\μ}}_v\right)+S(L\left({\mathrm{\μ}}_v\right)-L_0\left({\mathrm{\μ}}_v\right))}---\left(4\right)$

L (μ in formula (4) _v) be the radiance that sensor receives, ρ _tapparent reflectance, L ₀(μ _v) be path radiation term, F _d=u _sf ₀t(u _s) be the descending built-up radiation of the sun, F0 is the solar irradiance on atmospheric envelope top); the transmitance between sensor and target, direct projection transmitance, t ' _d(θ _v) be scattering transmitance; Under known observation condition, set one group of ρ _tvalue and corresponding sensor, obtain one group of radiance L (μ by MODTRAN4 or the simulation of 6S radiative transfer model _v), and obtain parameter path radiation term, transmitance, atmosphere hemisphere albedo and the descending built-up radiation of the sun of picture rich in detail Atmospheric Correction, will in above-mentioned parameter substitution formula (4), carry out picture rich in detail Atmospheric Correction.

More specifically, in described step 104, also comprise cloud identification step: once read in the band class information of the MODIS data of image series, then pointwise detects, and finally generates cloud identification document.

Especially, in described step 103, also comprise:

$M^2\left(d\right)=\frac{1}{{(d_{\max }-d_{\min })}^2(n-d_{\min })(m-d_{\min })}\underset{i=1}{\overset{n-d}{\mathrm{\Σ}}}\underset{j=1}{\overset{m-d}{\mathrm{\Σ}}}\underset{d_j=d_{\min }}{\overset{d_j=d_{\max }}{\mathrm{\Σ}}}\underset{d_i=d_{\min }}{\overset{d_i=d_{\max }}{\mathrm{\Σ}}}\left[{({\mathrm{\ρ}}_{i,j}-{\mathrm{\ρ}}_{i+d_i,j+d_j})}^2\right]---\left(5\right)$

Wherein m*n is the size of calculation window, d _max, d _minrefer to maximum, the minimum pixel distance of the pixel spacing that participates in calculating

From, ρ is reflectivity, to each pixel in window with and its distance at d _min-d _maxbetween the average of pixel calculated difference value quadratic sum, for the special area of this class, city, the otherness on earth's surface is very obvious, improved structure function computing method have better stability in urban area.As shown in Figure 3, the comparison diagram of traditional detection method and detection method of the present invention, sequence from left to right, the 3rd width figure adopts particle optical thickness remote-sensing monitoring method of the present invention to process to obtain image, and its successful is better than traditional detection method.

In order further to describe the present invention, below enumerate more detailed embodiment and describe, as shown in Figure 2.

1, the pre-service of MODIS data

In particle inverting research, first to carry out pre-service to MODIS data, this step is very necessary.Atmospheric Correction and cloud sign through geometric correction, picture rich in detail, obtain reliable MODIS data.

(1-1) geometric correction of MODIS data has following steps: first utilize longitude and latitude data in MODIS1B data as reference mark, meanwhile, by interpolation algorithm, calculate the actual longitude and latitude data of each pixel.Secondly, for the data of each scanning strip, carry out geometric correction.Finally, the geometric correction result of all scanning strips is just spliced and can be completed geometric correction all processes.

(1-2) Atmospheric Correction of picture rich in detail.Radiation value L (the μ being obtained by satellite sensor _v) can be represented by formula (4) distortion:

$L\left({\mathrm{\μ}}_v\right)=L_0\left({\mathrm{\μ}}_v\right)+\frac{{\mathrm{\ρ}}_t}{1-{\mathrm{\ρ}}_{t}S}{F}_{d}T\left({\mathrm{\μ}}_v\right)---\left(6\right)$

In formula (4-1): L (μ _v) be the radiance that sensor receives, L ₀(μ _v) be path radiation term, F _d=u _sf ₀t(u _s) be the descending built-up radiation (F of the sun ₀the solar irradiance on atmospheric envelope top), be transmitance between sensor and target ( direct projection transmitance, t ' _d(θ _v) be scattering transmitance).Under known observation condition (geometric parameter of the sun and sensor, atmosphere profile, Reflectivity for Growing Season etc.), set one group of ρ _tvalue and corresponding sensor height, obtain one group of radiance L (μ by radiative transfer model simulations such as MODTRAN4,6S _v), substitution formula (4), more just can obtain the required parameter of atmospheric correction (path radiation term, transmitance, atmosphere hemisphere albedo and the descending built-up radiation of the sun) through simple algebraic operation.By formula (4-1), can solve ρ _t:

${\mathrm{\ρ}}_t=\frac{L\left({\mathrm{\μ}}_v\right)-L_0\left({\mathrm{\μ}}_v\right)}{F_{d}T\left({\mathrm{\μ}}_v\right)+S(L\left({\mathrm{\μ}}_v\right)-L_0\left({\mathrm{\μ}}_v\right))}---\left(4\right)$

Reflectivity for Growing Season and the setting of respective sensor height see the following form shown in 1.

Table 1

The atmospheric correction parameter substitution equation (4) of the simulations such as the radiance that sensor is received and MODTRAN4 or 6S just can carry out atmospheric correction.

Reflectivity for Growing Season, the sensor height of corresponding radiance value and setting in table 1 are mapped, and substitution equation (4) simultaneous obtains the system of equations being comprised of five equations:

$\left\{\begin{array}{c}{L}_1=L_0\\ L_2=L_0+\frac{0.1}{1-0.1S}{F}_{d}T\left({\mathrm{\μ}}_v\right)\\ L_3=L_0+\frac{0.2}{1-0.2S}{F}_{d}T\left({\mathrm{\μ}}_v\right)\\ L_4=\frac{0.1}{1-0.1S}{F}_d\\ L_5=\frac{0.2}{1-0.2S}{F}_d\end{array}\right.---\left(7\right)$

By five radiation brightness value L of simulation ₁-L ₅system of equations above substitution, solves atmospheric correction parameter.

$\left\{\begin{array}{c}{L}_0=L_1\\ T\left({\mathrm{\μ}}_v\right)=\frac{L_3-L_2}{L_5-L_4}\\ S=\frac{1-2a}{0.2-0.2a}\\ F_d=\frac{(L_3-L_1)(1-0.2S)}{0.2T\left({\mathrm{\μ}}_v\right)}\end{array}\right.---\left(8\right)$

Each pixel through type (4) in atmospheric correction coefficient and image is calculated and can count the Reflectivity for Growing Season that obtains pixel.

(1-3) cloud of MODIS data sign.The impact of cloud greatly reduces the precision of Atmospheric particulates inverting.In Atmospheric particulates inverting research, judge exactly in remote sensing image and have cloud pixel extremely important.MODIS has a plurality of wave bands can be used for detecting cloud, be mainly to utilize cloud to detect in the difference of reflectivity and radiation brightness value at underlying surfaces such as visible ray and infrared band and vegetation, soil, snow and waters, cloud has higher reflectivity and has low bright temperature value.First require once to read in required band class information, then pointwise detects, and finally generates cloud identification document.

2, treat the estimation of the Reflectivity for Growing Season of inversion chart picture

According to picture rich in detail and the geometric parameter for the treatment of inversion chart picture, the urban area BRDF model that coupling builds, obtains the Reflectivity for Growing Season for the treatment of inversion chart picture.

3, computation structure function

First, use the MODIS first passage data after Atmospheric Correction, the later structure function value calculating method of improvement proposing in formula (5) has been compared in simulation, with the stability of structure function value calculating method in formula (3).The design of analogy method is: from the image of urban area, Beijing MODIS passage one of 250 meters of resolution, shear out two width images of 160 * 160 pixels, between two width images, differ a pixel, suppose that the error of a pixel appears in registration, the d value of two kinds of structure function methods is set, be respectively, formula (3) d is set as 5, 6, 7, 8, 9, (dmin in formula (5), dmax) be set as (3, 6), (3, 8), (4, 8), (4, 9), (5, 8), m * n in formula (3) and (5) is all made as 20 * 20, and with 20 * 20 pixels, being divided into 8 * 8 chunks compares, under different setting parameter conditions by maximum error and the root-mean-square error of the structure function value of this two width image calculation, the result of processing is as Fig. 4, shown in Fig. 5, improved method maximum error and root-mean-square error are all low than traditional method, can find out that improved structure function computing method have better stability in urban area.

Then, computation structure functional value.Suppose in selected image series, clean image and other topographical features for the treatment of inversion chart picture remain unchanged, and do not rain, snow etc. affects the weather characteristics of Reflectivity for Growing Season, there is no the change of the earth's surface types such as destruction of new buildings appearance or inhering architecture.Can be according to the structure function value of clean image, the earth's surface two that coupling builds is to Reflectivity Model:

$M^2\left(d\right)=\frac{1}{{(d_{\max }-d_{\min })}^2(n-d_{\min })(m-d_{\min })}\underset{i=1}{\overset{n-d}{\mathrm{\Σ}}}\underset{j=1}{\overset{m-d}{\mathrm{\Σ}}}\underset{d_j=d_{\min }}{\overset{d_j=d_{\max }}{\mathrm{\Σ}}}\underset{d_i=d_{\min }}{\overset{d_i=d_{\max }}{\mathrm{\Σ}}}\left[{({\mathrm{\ρ}}_{i,j}-{\mathrm{\ρ}}_{i+d_i,j+d_j})}^2\right]---\left(5\right)$

Wherein m*n is the size of calculation window, d _max, d _minthe maximum, the minimum pixel distance that refer to the pixel spacing of participate in calculating, ρ is reflectivity, to each pixel in window with and its distance at d _min-d _maxbetween the average of pixel calculated difference value quadratic sum, by formula (1), can obtain the particle optical thickness for the treatment of inversion chart picture.

4, build the look-up table of particle optical thickness inverting

Structure function method essence is to utilize to treat that the ratio of inversion chart picture and reference picture structure function obtains particle optical thickness to atmospheric transmittance interpolation, therefore, the look-up table that experiment utilizes 6S model generation to contain solar zenith angle, observation zenith angle, atmospheric transmittance and particle optical thickness, the parameter while setting up look-up table arranges as follows: geometric parameter and Reflectivity for Growing Season data consistent; Particle pattern is continental particle; Atmospherical model is middle latitude summer; Band setting is MODIS the first wave band; Adopt the urban area BRDF model realization earth surface reflection property settings building; Particle optical thickness arranges as follows: 0.00001,0.05,0.1,0.2,0.3,0.4,0.5,0.6,0.7,0.8,0.9,1.0,1.2,1.4,1.6,1.8,2.0,2.5.

5, precision test

For understanding the precision of the particle optical thickness of inverting, the present invention contrasts the observation data of the inversion result of Beijing area and AERONET Beijing observation website.During contrast, select optical thickness mean value and the AERONET measured value in 10 * 10 pixel regions centered by AERONET site location in inversion chart picture to contrast, AERONET optical thickness value wherein, what choose is the mean value of the measured value in a hour centered by satellite passes by constantly.Empirical tests known particle optical thickness inversion result and AERONET measurement result have good consistance, and visible improved structure function method is applicable to urban area particle optical thickness inverting, have improved the accuracy that detects particle optical thickness.

Certainly; more than explanation is only preferred embodiment of the present invention; the present invention is not limited to enumerate above-described embodiment; should be noted that; any those of ordinary skill in the art are under the instruction of this instructions; that makes is allly equal to alternative, obvious form of distortion, within all dropping on the essential scope of this instructions, ought to be subject to protection of the present invention.

Claims (5)

1. for a particle optical thickness remote-sensing monitoring method for urban area, it comprises the following steps:

The pre-service of A, MODIS data, comprises geometric correction, cloud identification, the Atmospheric Correction of picture rich in detail;

B, according to the geometric parameter for the treatment of the picture rich in detail that obtains in inversion chart picture and steps A, the urban area BRDF model of coupling structure, obtains the Reflectivity for Growing Season for the treatment of inversion chart picture;

The Reflectivity for Growing Season structure function value of inversion chart picture and the structure function value of apparent reflectance are treated in C, calculating;

D, according to the geometric condition for the treatment of inverting image, build the look-up table of inverting particle optical thickness, by treating that the Reflectivity for Growing Season of inverting image and the structure function value of apparent reflectance search its particle optical thickness.

2. particle optical thickness remote-sensing monitoring method according to claim 1, it is characterized in that, in described steps A, also comprise geometric correction step: first utilize longitude and latitude data in MODIS data as reference mark, meanwhile, by interpolation algorithm, calculate the actual longitude and latitude data of each pixel; Then for the data of each scanning strip, carry out geometric correction, finally the geometric correction result of all scanning strips has been spliced to geometric correction.

3. particle optical thickness remote-sensing monitoring method according to claim 1, is characterized in that, also comprises picture rich in detail Atmospheric Correction step in described steps A: the radiation value L (μ being obtained by satellite sensor _v) can be expressed from the next:

$L\left({\mathrm{\μ}}_v\right)=L_0\left({\mathrm{\μ}}_v\right)+\frac{{\mathrm{\ρ}}_t}{1-{\mathrm{\ρ}}_{t}S}{F}_{d}T\left({\mathrm{\μ}}_v\right)---\left(1\right)$

L (μ in formula (1) _v) be the radiance that sensor receives, ρ _tapparent reflectance, L ₀(μ _v) be path radiation term, F _d=u _sf ₀t(u _s) be the descending built-up radiation of the sun, F ₀it is the solar irradiance on atmospheric envelope top; the transmitance between sensor and target, direct projection transmitance, t ' _d(θ _v) be scattering transmitance; Under known observation condition, set one group of ρ _tvalue and corresponding sensor, obtain one group of radiance L (μ by MODTRAN4 or the simulation of 6S radiative transfer model _v), and obtain parameter path radiation term, transmitance, atmosphere hemisphere albedo and the descending built-up radiation of the sun of picture rich in detail Atmospheric Correction, will in above-mentioned parameter substitution formula (1), carry out picture rich in detail Atmospheric Correction.

4. particle optical thickness remote-sensing monitoring method according to claim 1, is characterized in that, also comprises cloud identification step in described steps A: once read in the band class information of the MODIS data of image series, then pointwise detects, and finally generates cloud identification document.

5. particle optical thickness remote-sensing monitoring method according to claim 1, is characterized in that, also comprises the structure function computing method that propose for this complicated earth surface of urban area in described step C:

$M^2\left(d\right)=\frac{1}{{(d_{\max }-d_{\min })}^2(n-d_{\min })(m-d_{\min })}\underset{i=1}{\overset{n-d}{\mathrm{\Σ}}}\underset{j=1}{\overset{m-d}{\mathrm{\Σ}}}\underset{d_j=d_{\min }}{\overset{d_j=d_{\max }}{\mathrm{\Σ}}}\underset{d_i=d_{\min }}{\overset{d_i=d_{\max }}{\mathrm{\Σ}}}\left[{({\mathrm{\ρ}}_{i,j}-{\mathrm{\ρ}}_{i+d_i,j+d_j})}^2\right]---\left(2\right)$

Wherein m*n is the size of calculation window, d _max, d _minthe maximum, the minimum pixel distance that refer to the pixel spacing that participates in calculating, ρ is reflectivity.