CN103605123B

CN103605123B - Parametrization remote sensing technique based on oxygen A channel aerosol scattering effect

Info

Publication number: CN103605123B
Application number: CN201310646718.3A
Authority: CN
Inventors: 邹铭敏; 陈良富; 陶金花; 张莹; 范萌; 苏林
Original assignee: Institute of Remote Sensing and Digital Earth of CAS
Current assignee: Institute of Remote Sensing and Digital Earth of CAS
Priority date: 2013-12-04
Filing date: 2013-12-04
Publication date: 2016-08-31
Anticipated expiration: 2033-12-04
Also published as: CN103605123A

Abstract

The invention discloses a parameterized remote sensing method based on the scattering effect of the oxygen A channel aerosol, comprising: parameterizing the scattering effect of the oxygen A channel aerosol, and using four parameter factors to equivalently count the scattering effect; selecting four parameters The initial prior value of the factor, and then use 4 parameter factors to modify the calculation equation of the whole atmosphere transmittance to obtain the whole atmosphere effective transmittance model; The observation data of channel A adopts the optimal estimation method to invert the values of the four parameter factors of the aerosol scattering effect under the actual aerosol observation conditions; the values of the four parameter factors obtained by the inversion are used to obtain the final corrected whole layer Atmospheric transmittance model to accurately estimate aerosol scattering effects. The invention provides an accurate and effective technical solution to the aerosol scattering correction problem faced in the process of carrying out visible light-short-wave near-infrared satellite remote sensing in the aerosol high-value area.

Description

Parametric Remote Sensing Method Based on Aerosol Scattering Effect of Oxygen A Channel

技术领域technical field

本发明涉及卫星大气遥感技术领域，尤其涉及基于氧A通道气溶胶散射效应的参数化遥感方法。The invention relates to the technical field of satellite atmospheric remote sensing, in particular to a parametric remote sensing method based on the aerosol scattering effect of an oxygen A channel.

背景技术Background technique

美国于20世纪七十年代最先设计了用于大气遥感探测的卫星传感器HIRS，早期主要用于探测大气中二氧化碳和水汽，反演大气的温度廓线。早期的传感器主要利用大气的热辐射信息，由于热红外波段辐射波长较长，在大气传输过程中不受颗粒物的散射作用影响。In the 1970s, the United States first designed the satellite sensor HIRS for remote sensing of the atmosphere. In the early days, it was mainly used to detect carbon dioxide and water vapor in the atmosphere and retrieve the temperature profile of the atmosphere. Early sensors mainly used the thermal radiation information of the atmosphere. Due to the long wavelength of radiation in the thermal infrared band, it is not affected by the scattering effect of particles in the process of atmospheric transmission.

受大气热辐射传输自身特点的限制，卫星传感器的热辐射观测数据对近地层大气状态参数不敏感，包含的近地层大气状态信息量很少。紫外-可见光-短波近红外遥感方式则可以弥补热红外遥感的这一缺陷，该波段范围内卫星传感器接收的是经过地表反射的太阳辐射，包含了近地层大气状态参数信息。Due to the limitation of the characteristics of atmospheric thermal radiation transmission, the thermal radiation observation data of satellite sensors are not sensitive to the parameters of the near-surface atmospheric state, and contain very little information about the near-surface atmospheric state. The ultraviolet-visible-short-wave near-infrared remote sensing method can make up for this shortcoming of thermal infrared remote sensing. In this band, the satellite sensor receives the solar radiation reflected by the surface, which contains the information of the near-surface atmospheric state parameters.

对于痕量气体遥感来说，由于该波段范围内波长短，易受到底层大气中悬浮的气溶胶颗粒影响，散射效应能明显改变太阳辐射的传输路径，从而给这一段范围内痕量气体的卫星遥感精度带来很大的误差，例如开展温室气体卫星遥感中，ENVISAT上搭载的SCIAMACHY传感器，日本的GOSAT搭载的TANSO-FTS传感器等，其观测数据非常容易受到气溶胶散射的影响。For trace gas remote sensing, due to the short wavelength in this band, it is easily affected by aerosol particles suspended in the bottom atmosphere. The accuracy of remote sensing brings large errors. For example, in the development of satellite remote sensing of greenhouse gases, the SCIAMACHY sensor on ENVISAT and the TANSO-FTS sensor on Japan's GOSAT are very susceptible to the influence of aerosol scattering.

针对气溶胶的散射效应如何校正的问题，产生了相应的一些解决方案。基于多项式拟合方式消除气溶胶散射效应的方法，其主要依据在特定的观测通道内，气体的吸收属于高频信号，而散射是一种慢变的低频信号，因此通过多项式拟合可以去除散射的慢变信号，但此种方法不能有效校正散射对光子路径的改变，因此散射校正的精度有限。另外，基于物理散射模型(DISORT)的校正方法，可以很好地估算气溶胶散射效应，但需要掌握观测视场内气溶胶分布及相应的物理光学参数，如单次散射反照率、相函数、不对称因子等，这些参数在实际观测中难以实时获取，这就使得难以基于物理模型的散射校正的应用。Aiming at the problem of how to correct the scattering effect of aerosol, some corresponding solutions have been produced. The method of eliminating aerosol scattering effect based on polynomial fitting is mainly based on the fact that in a specific observation channel, the absorption of gas is a high-frequency signal, while scattering is a slowly changing low-frequency signal, so polynomial fitting can be used to remove scattering Slowly changing signal, but this method cannot effectively correct the change of photon path caused by scattering, so the accuracy of scattering correction is limited. In addition, the correction method based on the physical scattering model (DISORT) can well estimate the aerosol scattering effect, but it is necessary to know the aerosol distribution in the observation field and the corresponding physical optics parameters, such as single scattering albedo, phase function, Asymmetry factors, etc. These parameters are difficult to obtain in real time in actual observations, which makes it difficult to apply scattering correction based on physical models.

发明内容Contents of the invention

针对现有技术存在的问题，本发明的目的在于提供一种基于氧A通道气溶胶散射效应的参数化遥感方法，该方法为在气溶胶高值区开展可见光-短波近红外卫星遥感过程中面临的气溶胶散射校正问题，提供了精确有效的解决技术手段。Aiming at the problems existing in the prior art, the object of the present invention is to provide a parametric remote sensing method based on the oxygen A channel aerosol scattering effect, which is to solve the problems faced in the visible light-short-wave near-infrared satellite remote sensing process in the aerosol high-value area. The aerosol scattering correction problem provides an accurate and effective technical solution.

为实现上述目的，本发明基于氧A通道气溶胶散射效应的参数化遥感方法，具体为：In order to achieve the above object, the present invention is based on the parameterized remote sensing method of oxygen A channel aerosol scattering effect, specifically:

1)利用氧A通道卫星观测大气，得到可见光-短波红外波段卫星痕量气体遥感数据；1) Use the oxygen A channel satellite to observe the atmosphere, and obtain satellite trace gas remote sensing data in the visible light-short-wave infrared band;

2)根据气溶胶粒子对光子路径物理散射过程，将气太阳光子路径因散射作用而产生的改变量，用4个参数因子来等效表示，该4个参数因子为：含有气溶胶粒子的低层大气厚度ha，反射率因子α，光子路径修正因子ρ，光子散射后路径长度分布形状的调整因子Υ；2) According to the physical scattering process of the photon path by the aerosol particles, the change of the air-solar photon path due to the scattering effect is equivalently represented by 4 parameter factors. The 4 parameter factors are: the lower layer containing aerosol particles Atmospheric thickness ha, reflectivity factor α, photon path correction factor ρ, adjustment factor Υ of path length distribution shape after photon scattering;

3)依据气溶胶粒子垂直分布特征，选定4个参数因子的初始值，并将此初始值加入到大气透过率模型中，得到初始的整层大气有效透过率模型；3) According to the vertical distribution characteristics of aerosol particles, the initial values of the four parameter factors are selected, and these initial values are added to the atmospheric transmittance model to obtain the initial effective transmittance model of the whole layer of the atmosphere;

4)利用含气溶胶散射效应的氧A通道卫星的观测数据，采用最优化估计方法，反演对应的气溶胶观测条件下4个参数化因子的准确值；4) Using the observation data of the Oxygen A channel satellite with aerosol scattering effect, adopt the optimal estimation method to invert the accurate values of the four parameterization factors under the corresponding aerosol observation conditions;

5)利用步骤4)中反演得到的4个参数因子值，构建最终的整层大气有效透过率模型，此最终模型计入了气溶胶的散射过程，利用此模型估算校正气溶胶散射效应。5) Use the 4 parameter factors obtained from the inversion in step 4) to construct the final effective transmittance model of the entire atmosphere. This final model takes into account the scattering process of aerosols, and uses this model to estimate and correct the aerosol scattering effect .

进一步，所述步骤2)中利用以下定义的4个参数因子来等效表征气溶胶的散射效应，4个参数因子具体为：Further, in the step 2), the four parameter factors defined below are used to equivalently characterize the scattering effect of the aerosol, and the four parameter factors are specifically:

A)大气中气溶胶通常分布于边界层中，定义含有气溶胶粒子的低层大气厚度ha，在ha高度以上，太阳光子传输过程不受散射影响；进入如ha高度下的光子，将受气溶胶散射影响而导致传输路径发生变化；A) Aerosols in the atmosphere are usually distributed in the boundary layer. Define the thickness ha of the lower atmosphere containing aerosol particles. Above the ha height, the solar photon transmission process is not affected by scattering; photons entering such as ha height will be scattered by the aerosol The transmission path changes due to the impact;

B)太阳光子在到达含有气溶胶的低层大气顶时，部分光子将被反射向上传输，据此定义反射率因子α，表征被反射的光子比例；B) When solar photons reach the lower atmosphere top containing aerosol, part of the photons will be reflected and transmitted upwards, and the reflectance factor α is defined accordingly, which represents the proportion of reflected photons;

C)未被反射的光子进入低层大气，受气溶胶散射影响其传输路径发生改变，据此定义光子路径修正因子ρ，表征散射对光子传输路径的改变量；C) The unreflected photon enters the lower atmosphere, and its transmission path is changed due to aerosol scattering. Based on this, the photon path correction factor ρ is defined to represent the amount of change in the photon transmission path caused by scattering;

D)考虑到不同气溶胶类型对光子散射过程的差异，在增加路径修正因子ρ的基础上，再定义光子散射后路径长度分布形状的调整因子Υ。D) Considering the difference in the photon scattering process of different aerosol types, on the basis of increasing the path correction factor ρ, define the adjustment factor Υ of the shape of the path length distribution after photon scattering.

进一步，所述步骤3)具体为：Further, the step 3) is specifically:

A)利用S1中定义的4个参数因子，修正整层大气透过率方程，给出新的有效透过率模型：A) Using the four parameter factors defined in S1, the equation of the entire atmospheric transmittance is corrected, and a new effective transmittance model is given:

$\overset{~ ~}{T T} = = α α \cdot &Center Dot; {T T}_{22} + + ((11 - - α α)) \cdot \cdot {T T}_{11} \cdot &Center Dot; {T T}_{22},,$

式中，In the formula,

${T T}_{11} = = exp exp [[- - ((\frac{11}{μ μ} + + \frac{11}{{μ μ}_{00}})) \cdot \cdot ((11 + + δ δ)) \cdot \cdot {τ τ}_{11}]],,$

${T T}_{22} = = exp exp [[- - ((\frac{11}{μ μ} + + \frac{11}{{μ μ}_{00}})) \cdot &Center Dot; {τ τ}_{22}]],,$

其中，in,

$τ_{1} = {&Integral;}_{0}^{h_{a}} k (h) d h, τ_{2} = {&Integral;}_{h_{a}}^{h_{t o p}} k (h) d h,$ δ＝ρ·exp[-γ·(τ₁+τ₂)]， $τ_{1} = {&Integral;}_{0}^{h_{a}} k (h) d h, τ_{2} = {&Integral;}_{h_{a}}^{h_{t o p}} k (h) d h,$ δ=ρ·exp[-γ·(τ ₁ +τ ₂ )],

底层含有气溶胶的大气看作一层L1,上部不含气溶胶的大气作为另外一层L2,T1、T2分别表示云层L1层、L2层的大气透过率；μ、μ₀分别表示太阳天顶角、观测天顶角的余弦；h_a表示L1层顶的高度，h_top表示大气顶的高度；The bottom layer of the atmosphere containing aerosols is regarded as one layer L1, and the upper part of the atmosphere without aerosols is regarded as another layer L2. T1 and T2 represent the atmospheric transmittance of the cloud layer L1 and L2 respectively _; apex angle, the cosine of the observed zenith angle; h _a represents the height of the top of the L1 layer, and h _top represents the height of the top of the atmosphere;

B)得到新的整层大气有效透过率模型后，根据已有的关于观测场景中气溶胶的先验知识，选定对应的4个参数因子的初始值，进而得到初始的整层大气有效透过率模型。B) After obtaining the new effective transmittance model of the whole layer of the atmosphere, according to the existing prior knowledge about the aerosol in the observation scene, select the initial values of the corresponding four parameter factors, and then obtain the initial effective transmittance of the whole layer of the atmosphere Transmittance model.

进一步，所述步骤4)具体为：Further, the step 4) is specifically:

A)基于最优化估计方法反演原理，反演4个参数因子，需要计算4个参数因子的权重函数Jacobian，依据修正的整层大气有效透过率模型，各参数因子的权重计算形式如下：A) Based on the inversion principle of the optimal estimation method, the inversion of 4 parameter factors requires the calculation of the weight function Jacobian of the 4 parameter factors. According to the modified effective transmittance model of the whole layer of the atmosphere, the weight calculation form of each parameter factor is as follows:

$\frac{\partial \partial \overset{~ ~}{T T} [[X x,, {C C}_{{O o}_{22}} ((h h))]]}{\partial \partial {h h}_{c c}} = = {T T}_{22} ((\frac{11}{μ μ} + + \frac{11}{{μ μ}_{00}})) k k (({h h}_{c c})) \cdot \cdot [[α α - - ((11 - - α α)) {T T}_{11} \cdot &Center Dot; exp exp ((- - γ γ \cdot \cdot (({τ τ}_{11} + + {τ τ}_{22})))) \cdot &Center Dot; ρ ρ]]$

$\frac{\partial \partial \overset{~ ~}{T T} [[X x,, {C C}_{{O o}_{22}} ((h h))]]}{\partial \partial α α} = = {T T}_{22} - - {T T}_{11} {T T}_{22}$

$\frac{\partial \partial \overset{~ ~}{T T} [[X x,, {C C}_{{O o}_{22}} ((h h))]]}{\partial \partial ρ ρ} = = ((11 - - α α)) {T T}_{11} {T T}_{22} [[- - ((\frac{11}{μ μ} + + \frac{11}{{μ μ}_{00}})) {τ τ}_{11} exp exp ((- - γ γ (({τ τ}_{11} + + {τ τ}_{22}))))]]$

$\frac{\partial \partial \overset{~ ~}{T T} [[X x,, {C C}_{{O o}_{22}} ((h h))]]}{\partial \partial γ γ} = = ((11 - - α α)) {T T}_{11} {T T}_{22} [[((\frac{11}{μ μ} + + \frac{11}{{μ μ}_{00}})) {τ τ}_{11} (({τ τ}_{11} + + {τ τ}_{22})) \cdot &Center Dot; δ δ]]$

式中，表示氧气垂直浓度廓线；In the formula, Indicates the vertical oxygen concentration profile;

B)依据大气状态参数，参照大气中氧气的浓度及垂直分布特点确定分布，然后按照S3.1中给出的公式计算4个因子初始值状态下的权重函数；B) According to the atmospheric state parameters, refer to the concentration and vertical distribution characteristics of oxygen in the atmosphere to determine distribution, and then calculate the weight function under the initial value state of the four factors according to the formula given in S3.1;

C)最优化估计方法反演过程中，需要4个参数因子的先验协方差信息Sa，考虑到4因子间相互独立性，定义Sa为一对角矩阵，对角元素的值分别设定为100、0.5、1、100，分别对应路径修正因子ρ、反射率因子α、底层大气厚度ha以及分布形状校正因子Υ；并设定同为对角矩阵的误差协方差矩阵Se；C) In the inversion process of the optimal estimation method, the prior covariance information Sa of the four parameter factors is required. Considering the mutual independence among the four factors, Sa is defined as a pair of diagonal matrices, and the values of the diagonal elements are respectively set as 100, 0.5, 1, 100, respectively corresponding to path correction factor ρ, reflectivity factor α, bottom atmosphere thickness ha and distribution shape correction factor Υ; and set the error covariance matrix Se which is also a diagonal matrix;

D)依据最优化估计方法原理，进行一次迭代计算，迭代形式如下：D) According to the principle of the optimal estimation method, an iterative calculation is performed, and the iterative form is as follows:

${X x}_{i i + + 11} = = {X x}_{i i} + + {[[{S S}_{a a}^{- - 11} + + {K K}_{i i}^{T T} {S S}_{ϵ ϵ}^{- - 11} {K K}_{i i}]]}^{- - 11} {{{K K}_{i i}^{T T} {S S}_{ϵ ϵ}^{- - 11} [[Y Y - - F f ((X x))]] - - {S S}_{a a}^{- - 11} [[{X x}_{i i} - - {X x}_{a a}]]}},,$

X_i表示第i次迭代计算的待反演参数向量，此处它包含了4个参数因子；X _i represents the parameter vector to be inverted calculated in the i-th iteration, here it contains 4 parameter factors;

E)重复步骤A)至步骤D)，直到最优化估计方法定义的代价函数达到阈值内，停止迭代，得到4个参数因子的反演结果。E) Repeat step A) to step D) until the cost function defined by the optimal estimation method reaches the threshold, stop the iteration, and obtain the inversion results of the four parameter factors.

本发明提供的基于氧A通道气溶胶散射效应的参数化算法，依据用大气中氧气浓度的恒稳特点，利用典型的氧A特征吸收通道，有效估算气溶胶的散射效应，为在气溶胶高值区开展可见光-短波近红外卫星遥感过程中面临的气溶胶散射校正问题，提供了精确有效的解决技术手段。The parametric algorithm based on the aerosol scattering effect of the oxygen A channel provided by the present invention uses the typical oxygen A characteristic absorption channel to effectively estimate the scattering effect of the aerosol according to the constant and stable characteristics of the oxygen concentration in the atmosphere. The aerosol scattering correction problem faced in the process of carrying out visible light-short-wave near-infrared satellite remote sensing in the value area provides an accurate and effective technical solution.

附图说明Description of drawings

图1为本发明的流程图。Fig. 1 is a flowchart of the present invention.

图2为实施例中一定气溶胶类型条件下散射效应参数化估算校正有效性验证的实例图；Fig. 2 is an example diagram of valid verification of parameterized estimation and correction of scattering effect under certain aerosol type conditions in the embodiment;

图3为实施例中一定气溶胶类型条件下散射效应参数化估算偏差统计图。Fig. 3 is a statistical diagram of parameterized estimation deviation of scattering effect under certain aerosol type conditions in the embodiment.

具体实施方式detailed description

下面，参考附图1、图2、图3，对本发明进行更全面的说明，附图中示出了本发明的示例性实施例。然而，本发明可以体现为多种不同形式，并不应理解为局限于这里叙述的示例性实施例。而是，提供这些实施例，从而使本发明全面和完整，并将本发明的范围完全地传达给本领域的普通技术人员。In the following, the present invention will be described more fully with reference to accompanying drawings 1 , 2 and 3 , in which exemplary embodiments of the present invention are shown. This invention may, however, be embodied in many different forms and should not be construed as limited to the exemplary embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art.

本发明基于氧A通道气溶胶散射效应的参数化遥感方法，首先需要利用氧A通道卫星观测大气，得到可见光-短波红外波段卫星痕量气体遥感数据，基于所得到的数据，处理过程为：The parametric remote sensing method based on the aerosol scattering effect of the oxygen A channel of the present invention first needs to use the oxygen A channel satellite to observe the atmosphere to obtain satellite trace gas remote sensing data in the visible light-short-wave infrared band. Based on the obtained data, the processing process is as follows:

S1.根据气溶胶粒子对光子路径物理散射过程，将气太阳光子路径因散射作用而产生的改变量，用4个参数因子来等效表示；S1. According to the physical scattering process of the photon path by the aerosol particles, the change amount of the air-solar photon path due to scattering is equivalently represented by four parameter factors;

S2.依据气溶胶粒子垂直分布特征等先验知识，选定4个参数因子的初始值，并将此初始值加入到大气透过率模型中，得到初始的整层大气有效透过率模型；S2. Based on the prior knowledge of the vertical distribution characteristics of aerosol particles, select the initial values of the four parameter factors, and add these initial values to the atmospheric transmittance model to obtain the initial effective transmittance model of the entire atmosphere;

S3.利用含气溶胶散射效应的氧A通道的观测数据，采用最优化估计方法，反演对应的气溶胶观测条件下4个散射效应参数化因子的准确值；S3. Using the observation data of the oxygen A channel containing the aerosol scattering effect, the optimal estimation method is used to invert the accurate values of the four scattering effect parameterization factors under the corresponding aerosol observation conditions;

S4.利用S3中反演得到的4个参数因子值，构建最终的整层大气有效透过率模型，此最终模型计入了气溶胶的散射过程，利用此模型估算校正气溶胶散射效应。S4. Using the 4 parameter factor values obtained from the inversion in S3, construct the final effective transmittance model of the whole atmosphere. This final model takes into account the scattering process of aerosols, and uses this model to estimate and correct the aerosol scattering effect.

其中，步骤S1进一步包括：Wherein, step S1 further includes:

S1.1大气中气溶胶通常分布于边界层中，定义含有气溶胶粒子的低层大气厚度ha，在ha高度以上，太阳光子传输过程不受散射影响；进入如ha高度下的光子，将受气溶胶散射影响而导致传输路径发生变化；S1.1 Aerosols in the atmosphere are usually distributed in the boundary layer. Define the thickness ha of the lower atmosphere containing aerosol particles. Above the height of ha, the solar photon transmission process is not affected by scattering; photons entering such a height will be affected by the aerosol The transmission path changes due to the influence of scattering;

S1.2太阳光子在到达含有气溶胶的低层大气顶时，部分光子将被反射向上传输，据此定义反射率因子α，表征被反射的光子比例；S1.2 When solar photons reach the top of the lower atmosphere containing aerosols, part of the photons will be reflected and transmitted upwards. Based on this, the reflectance factor α is defined to represent the proportion of reflected photons;

S1.3未被反射的光子进入低层大气，受气溶胶散射影响其传输路径发生改变，据此定义光子路径修正因子ρ，表征散射对光子传输路径的改变量；S1.3 Unreflected photons enter the lower atmosphere, and their transmission paths are changed due to aerosol scattering. Based on this, the photon path correction factor ρ is defined to represent the amount of change in the photon transmission path caused by scattering;

S1.4考虑到不同气溶胶类型对光子散射过程的差异，在增加路径修正因子ρ的基础上，再定义光子散射后路径长度分布形状的调整因子Υ；S1.4 Considering the difference of different aerosol types in the photon scattering process, on the basis of increasing the path correction factor ρ, define the adjustment factor Υ of the shape of the path length distribution after photon scattering;

S1.5利用以上定义的4个因子来等效表征气溶胶的散射效应。S1.5 Use the four factors defined above to equivalently characterize the scattering effect of aerosols.

其中，步骤S2进一步包括：Wherein, step S2 further includes:

S2.1利用S1中定义的4个参数因子，修正整层大气透过率方程，给出新的有效透过率模型：S2.1 Using the four parameter factors defined in S1, the equation of the entire atmospheric transmittance is corrected, and a new effective transmittance model is given:

$\overset{~ ~}{T T} = = α α \cdot &Center Dot; {T T}_{22} + + ((11 - - α α)) \cdot &Center Dot; {T T}_{11} \cdot &Center Dot; {T T}_{22},,$

式中，In the formula,

${T T}_{11} = = exp exp [[- - ((\frac{11}{μ μ} + + \frac{11}{{μ μ}_{00}})) \cdot \cdot ((11 + + δ δ)) \cdot &Center Dot; {τ τ}_{11}]],,$

其中，in,

S2.2得到新的整层大气有效透过率模型后，根据已有的关于观测场景中气溶胶的先验知识，选定对应的4个参数因子的初始值，进而得到初始的整层大气有效透过率模型；S2.2 After obtaining the new effective transmittance model of the whole layer of the atmosphere, according to the existing prior knowledge about the aerosol in the observation scene, select the initial values of the corresponding four parameter factors, and then obtain the initial whole layer of the atmosphere Effective transmittance model;

其中，步骤S3进一步包括：Wherein, step S3 further includes:

S3.1基于最优化估计方法反演原理，反演4个参数因子，需要计算个因子的权重函数(Jacobian)，依据修正的整层大气有效透过率模型，各因子的权重计算形式如下：S3.1 Based on the inversion principle of the optimal estimation method, the inversion of four parameter factors requires the calculation of the weight function (Jacobian) of each factor. According to the modified effective transmittance model of the whole layer of the atmosphere, the weight calculation form of each factor is as follows:

$\frac{\partial \partial \overset{~ ~}{T T} [[X x,, {C C}_{{O o}_{22}} ((h h))]]}{\partial \partial {h h}_{c c}} = = {T T}_{22} ((\frac{11}{μ μ} + + \frac{11}{{μ μ}_{00}})) k k (({h h}_{c c})) \cdot &Center Dot; [[α α - - ((11 - - α α)) {T T}_{11} \cdot &Center Dot; exp exp ((- - γ γ \cdot \cdot (({τ τ}_{11} + + {τ τ}_{22})))) \cdot &Center Dot; ρ ρ]],,$

$\frac{\partial \partial \overset{~ ~}{T T} [[X x,, {C C}_{{O o}_{22}} ((h h))]]}{\partial \partial α α} = = {T T}_{22} - - {T T}_{11} {T T}_{22},,$

$\frac{\partial \partial \overset{~ ~}{T T} [[X x,, {C C}_{{O o}_{22}} ((h h))]]}{\partial \partial ρ ρ} = = ((11 - - α α)) {T T}_{11} {T T}_{22} [[- - ((\frac{11}{μ μ} + + \frac{11}{{μ μ}_{00}})) {τ τ}_{11} exp exp ((- - γ γ (({τ τ}_{11} + + {τ τ}_{22}))))]],,$

$\frac{\partial \partial \overset{~ ~}{T T} [[X x,, {C C}_{{O o}_{22}} ((h h))]]}{\partial \partial γ γ} = = ((11 - - α α)) {T T}_{11} {T T}_{22} [[((\frac{11}{μ μ} + + \frac{11}{{μ μ}_{00}})) {τ τ}_{11} (({τ τ}_{11} + + {τ τ}_{22})) \cdot \cdot δ δ]],,$

式中，表示氧气垂直浓度廓线；T1、T2分别表示不同云层的大气透过率；μ、μ₀分别表示太阳天顶角、观测天顶角的余弦；τ₁表示含有气溶胶的底层大气的光学厚度；τ₂表示不含气溶胶的上层大气光学厚度；k(hc)表示hc高度处大气的容积吸收系数；δ是利用路径修正因子ρ与分布形状校正因子Υ定义的一个变量。In the formula, Represents the oxygen vertical concentration profile; T1 and T2 represent the atmospheric transmittance of different cloud layers respectively; μ and μ ₀ represent the solar zenith angle and the cosine of the observation zenith angle respectively; τ ₁ represents the optical thickness of the bottom atmosphere containing aerosols ; τ ₂ represents the optical thickness of the upper atmosphere without aerosol; k(hc) represents the volumetric absorption coefficient of the atmosphere at the height of hc; δ is a variable defined by the path correction factor ρ and the distribution shape correction factor Υ.

S3.2依据大气状态参数，参照大气中氧气的浓度及垂直分布特点确定分布，然后按照S3.1中给出的公式计算4个因子初始值状态下的权重函数；S3.2 is determined according to the atmospheric state parameters, referring to the concentration and vertical distribution characteristics of oxygen in the atmosphere distribution, and then calculate the weight function under the initial value state of the four factors according to the formula given in S3.1;

S3.3最优化估计方法反演过程中，需要4个参数因子的先验协方差信息Sa，考虑到4因子间相互独立性，定义Sa为一对角矩阵，对角元素的值分别设定为100、0.5、1、100，分别对应路径修正因子ρ、反射率因子α、底层大气厚度ha以及分布形状校正因子Υ；并设定同为对角矩阵的误差协方差矩阵Se；S3.3 In the inversion process of the optimal estimation method, the prior covariance information Sa of the 4 parameter factors is required. Considering the mutual independence of the 4 factors, Sa is defined as a pair of diagonal matrix, and the values of the diagonal elements are set separately are 100, 0.5, 1, 100, respectively corresponding to the path correction factor ρ, the reflectivity factor α, the bottom atmosphere thickness ha and the distribution shape correction factor Υ; and set the error covariance matrix Se which is also a diagonal matrix;

S3.4利用SCIATRAN模拟的卫星观测值，依据最优化估计方法原理，反演中采用逐线积分正向模型LBLRTM，进行一次迭代计算，迭代形式如下：S3.4 Using the satellite observations simulated by SCIATRAN, and based on the principle of the optimal estimation method, the line-by-line integral forward model LBLRTM is used in the inversion to perform an iterative calculation. The iterative form is as follows:

X_i表示第i次迭代计算的待反演参数向量，此处它包含了4个参数因子；Xi+1表示第i+1次迭代计算的待反演参数向量；S_ε表示误差协方差矩阵；Ki表示第i次计算得到的权重函数；Y表示观测；F(X)表示正向模型计算值；Xa表示待反演参量的先验值。X _i represents the parameter vector to be inverted calculated in the i-th iteration, here it contains 4 parameter factors; Xi+1 represents the parameter vector to be inverted calculated in the i+1 iteration; S _ε represents the error covariance matrix ; Ki represents the weight function obtained from the ith calculation; Y represents the observation; F(X) represents the calculated value of the forward model; Xa represents the prior value of the parameter to be inverted.

S3.5重复S3.1-S3.4步骤，直到最优化估计方法定义的代价函数达到阈值内，停止迭代，得到4个参数因子的反演结果。S3.5 Repeat steps S3.1-S3.4 until the cost function defined by the optimal estimation method reaches the threshold, stop the iteration, and obtain the inversion results of the four parameter factors.

本发明利用氧A通道卫星观测数据、基于等效理论的气溶胶散射效应参数化算法，依据用大气中氧气浓度的恒稳特点，利用典型的氧A特征吸收通道，避开了复杂的物理散射过程，从等效的角度切入，将散射效应参数化，是一种有效的估算散射效应的技术手段。通过该技术手段，本发明可有效估算气溶胶的散射效应，为在气溶胶高值区开展可见光-短波近红外卫星遥感过程中面临的气溶胶散射校正问题，提供了精确有效的解决技术手段。The present invention utilizes the satellite observation data of the oxygen A channel, the parametric algorithm of the aerosol scattering effect based on the equivalent theory, and uses the typical oxygen A characteristic absorption channel based on the constant and stable characteristics of the oxygen concentration in the atmosphere to avoid complicated physical scattering The process, starting from an equivalent point of view and parameterizing the scattering effect, is an effective technical means to estimate the scattering effect. Through this technical means, the present invention can effectively estimate the scattering effect of the aerosol, and provide an accurate and effective solution to the problem of aerosol scattering correction faced in the process of carrying out visible light-short-wave near-infrared satellite remote sensing in the high-value area of aerosol.

Claims

1. The parametric remote sensing method based on the oxygen A channel aerosol scattering effect, is characterized in that, the method is specifically:

1) Use the oxygen A channel satellite to observe the atmosphere, and obtain satellite trace gas remote sensing data in the visible light-short-wave infrared band;

2) According to the physical scattering process of the photon path by the aerosol particles, the change of the air-solar photon path due to the scattering effect is equivalently represented by 4 parameter factors. The 4 parameter factors are: the lower layer containing aerosol particles Atmospheric thickness ha, reflectivity factor α, photon path correction factor ρ, distribution shape correction factor γ;

3) According to the vertical distribution characteristics of aerosol particles, the initial values of the four parameter factors are selected, and these initial values are added to the atmospheric transmittance model to obtain the initial effective transmittance model of the whole layer of the atmosphere;

4) Using the observation data of the Oxygen A channel satellite with aerosol scattering effect, adopt the optimal estimation method to invert the accurate values of the four parameterization factors under the corresponding aerosol observation conditions;

5) Use the 4 parameter factors obtained from the inversion in step 4) to construct the final effective transmittance model of the entire atmosphere. This final model takes into account the scattering process of aerosols, and uses this model to estimate and correct the aerosol scattering effect ;

Described step 4) specifically is:

A) Based on the inversion principle of the optimal estimation method, the inversion of 4 parameter factors requires the calculation of the weight function Jacobian of the 4 parameter factors. According to the modified effective transmittance model of the whole layer of the atmosphere, the weight calculation form of each parameter factor is as follows:

\frac{\partial \partial \overset{~ ~}{T T} [[X x,, {C C}_{{O o}_{22}} ((h h))]]}{\partial \partial {h h}_{c c}} = = {T T}_{22} ((\frac{11}{μ μ} + + \frac{11}{{μ μ}_{00}})) k k (({h h}_{c c})) \cdot \cdot [[α α - - ((11 - - α α)) {T T}_{11} \cdot \cdot exp exp ((- - γ γ \cdot \cdot (({τ τ}_{11} + + {τ τ}_{22})))) \cdot \cdot ρ ρ]]

\frac{\partial \partial \overset{~ ~}{T T} [[X x,, {C C}_{{O o}_{22}} ((h h))]]}{\partial \partial α α} = = {T T}_{22} - - {T T}_{11} {T T}_{22}

\frac{\partial \partial \overset{~ ~}{T T} [[X x,, {C C}_{{O o}_{22}} ((h h))]]}{\partial \partial ρ ρ} = = ((11 - - α α)) {T T}_{11} {T T}_{22} [[- - ((\frac{11}{μ μ} + + \frac{11}{{μ μ}_{00}})) {τ τ}_{11} exp exp ((- - γ γ (({τ τ}_{11} + + {τ τ}_{22}))))]]

\frac{\partial \partial \overset{~ ~}{T T} [[X x,, {C C}_{{O o}_{22}} ((h h))]]}{\partial \partial γ γ} = = ((11 - - α α)) {T T}_{11} {T T}_{22} [[((\frac{11}{μ μ} + + \frac{11}{{μ μ}_{00}})) {τ τ}_{11} (({τ τ}_{11} + + {τ τ}_{22})) \cdot &Center Dot; δ δ]]

In the formula, Represents the oxygen vertical concentration profile; T1 and T2 represent the atmospheric transmittance of different cloud layers respectively; μ and μ ₀ represent the solar zenith angle and the cosine of the observation zenith angle respectively; τ ₁ represents the optical thickness of the bottom atmosphere containing aerosols ; _τ2 represents the optical thickness of the upper atmosphere without aerosol; k(hc) represents the volumetric absorption coefficient of the atmosphere at the height of hc; δ is a variable defined by the path correction factor ρ and the distribution shape correction factor γ;

B) According to the atmospheric state parameters, refer to the concentration and vertical distribution characteristics of oxygen in the atmosphere to determine distribution, and then calculate the weight function under the initial value state of the 4 factors according to the formula given in A);

C) In the inversion process of the optimal estimation method, the prior covariance information Sa of the 4 parameter factors is required. Considering the mutual independence among the 4 factors, Sa is defined as a pair of diagonal matrix, and the values of the diagonal elements are respectively set as 100, 0.5, 1, 100, respectively corresponding to path correction factor ρ, reflectivity factor α, bottom atmosphere thickness ha and distribution shape correction factor γ; and set the error covariance matrix Se which is also a diagonal matrix;

D) According to the principle of the optimal estimation method, an iterative calculation is performed, and the iterative form is as follows:

{X x}_{i i + + 11} = = {X x}_{i i} + + {[[{S S}_{a a}^{- - 11} + + {K K}_{i i}^{T T} {S S}_{ϵ ϵ}^{- - 11} {K K}_{i i}]]}^{- - 11} {{{K K}_{i i}^{T T} {S S}_{ϵ ϵ}^{- - 11} [[Y Y - - F f ((X x))]] - - {S S}_{a a}^{- - 11} [[{X x}_{i i} - - {X x}_{a a}]]}},,

X _i represents the parameter vector to be inverted calculated in the i-th iteration, here it contains 4 parameter factors; Xi+1 represents the parameter vector to be inverted calculated in the i+1 iteration; S _ε represents the error covariance matrix ; Ki represents the weight function obtained from the ith calculation; Y represents the observation; F(X) represents the calculated value of the forward model; Xa represents the prior value of the parameter to be inverted;

E) Repeat step A) to step D) until the cost function defined by the optimal estimation method reaches the threshold, stop the iteration, and obtain the inversion results of the four parameter factors.

2. the parametric remote sensing method based on oxygen A channel aerosol scattering effect as claimed in claim 1, is characterized in that, utilizes 4 parameter factors defined below to represent the scattering effect of aerosol equivalently in described step 2) , the four parameter factors are specifically:

A) Aerosols in the atmosphere are usually distributed in the boundary layer. Define the thickness ha of the lower atmosphere containing aerosol particles. Above the height of ha, the solar photon transmission process is not affected by scattering; photons entering the height of ha will be affected by aerosol scattering As a result, the transmission path changes;

B) When solar photons reach the lower atmosphere top containing aerosol, part of the photons will be reflected and transmitted upwards, and the reflectance factor α is defined accordingly, which represents the proportion of reflected photons;

C) The unreflected photon enters the lower atmosphere, and its transmission path is changed due to aerosol scattering. Based on this, the photon path correction factor ρ is defined to represent the amount of change in the photon transmission path caused by scattering;

D) Considering the difference in the photon scattering process of different aerosol types, on the basis of increasing the path correction factor ρ, define the distribution shape correction factor γ.

3. the parametric remote sensing method based on the oxygen A channel aerosol scattering effect as claimed in claim 1, is characterized in that, described step 3) be specially:

a) Utilize the 4 parameter factors defined in the step A) to correct the atmospheric transmittance equation of the whole layer, and give a new effective transmittance model:

\overset{~ ~}{T T} = = α α \cdot \cdot {T T}_{22} + + ((11 - - α α)) \cdot \cdot T T \cdot \cdot {T T}_{22},,

In the formula,

{T T}_{11} = = exp exp [[- - ((\frac{11}{μ μ} + + \frac{11}{{μ μ}_{00}})) \cdot \cdot ((11 + + δ δ)) \cdot &Center Dot; {τ τ}_{11}]],,

{T T}_{22} = = exp exp [[- - ((\frac{11}{μ μ} + + \frac{11}{{μ μ}_{00}})) \cdot \cdot {τ τ}_{22}]],,

in,

δ=ρ·exp[-γ·(τ ₁ +τ ₂ )],

The bottom layer of the atmosphere containing aerosols is regarded as one layer L1, and the upper part of the atmosphere without aerosols is regarded as another layer L2. T1 and T2 represent the atmospheric transmittance of the cloud layer L1 layer and L2 layer respectively; h _a represents the height of the top of the L1 layer , h _top represents the height of the top of the atmosphere; k(h) represents the volumetric absorption coefficient of the atmosphere at height h;

b) After obtaining the new effective transmittance model of the entire atmosphere, according to the existing prior knowledge about the aerosol in the observation scene, select the initial values of the corresponding four parameter factors, and then obtain the initial effective transmittance of the entire atmosphere Transmittance model.