CN103018736B - Satellite-borne remote sensor radiation calibration method based on atmospheric parameter remote sensing retrieval - Google Patents

Abstract

The invention discloses a satellite-borne remote sensor radiation calibration method based on atmospheric parameter remote sensing retrieval. The method includes nine steps. A surface reflectance is obtained through ground synchronous actual measurement during satellite crossing or historical data, imaging time and observation geometry parameters are obtained through remote sensing data head files, atmospheric parameters during imaging of synchronous crossing meteorological satellite sensors or related load retrieval remote sensors are utilized, the entrance pupil radiance of the remote sensors is calculated by an atmospheric radiation transmission model according to retrieval results of the atmospheric parameters, a calibration coefficient is calculated through a radiation calibration model, and thereby the on-orbit radiation calibration for a satellite-borne remote sensor is achieved. The satellite-borne remote sensor radiation calibration method based on the atmospheric parameter remote sensing retrieval is a pixel-level calibration method and has the advantages that the accuracy is high, the costs are low, simultaneously, the high frequency calibration can be achieved, historical remote sensing data can be calibrated, and the method has a wide application prospect to remote sensing data processing methods and application technical fields.

Description

A kind of satellite borne sensor Calibration Method based on atmospheric parameter remote-sensing inversion

Technical field

The present invention relates to a kind of satellite borne sensor Calibration Method based on atmospheric parameter remote-sensing inversion, belong to Remote Sensing Data Processing method and applied technical field, it satellite borne sensor in-orbit radiation calibration technical research and remotely-sensed data quantification application aspect significant.

Background technology

In order to meet the demand of remote sensing satellite quantification application, must carry out radiation calibration to satellite borne sensor, be observed ground object target feature and Changing Pattern thereof to guarantee that the data of remote sensor acquisition can reflect truly.By radiation calibration, the measured value (count value or magnitude of voltage) of remote sensor output is converted to absolute radiation amount (spoke brightness or reflectivity).Only have DN value in remote sensing images is converted to radiant quantity, the data that could obtain different location, different time and dissimilar remote sensor are carried out quantitative comparison and integrated application, to meet global resources and Study of Environmental Changes needs.Radiation calibration its objective is digital quantization by setting up remote sensor output DN value and its quantitative relationship between radiation value in corresponding visual field, be typically expressed as:

L=gain×DN+offset

Wherein L is the brightness of remote sensor entrance pupil spoke, the digital quantization output valve that DN is remote sensor, and gain and offset are respectively gain coefficient and the side-play amount of remote sensor, and these two amounts are commonly referred to as the calibration coefficient of remote sensor.

Radiation calibration is exactly gain and the side-play amount that will determine in calibration model.Conventional Calibration Method can be divided into calibration and three kinds of methods of alternative calibration on Laboratory Calibration, star.Distinct methods is being brought into play different effects in remote sensor different phase of living in respectively, but each method also all exists himself advantage and limitation.After satellite launch, remote sensor can carry out on star calibration or select to substitute calibration field substituting calibration for remote sensor continuous renewal calibration coefficient by robot scaling equipment on star in orbit.

On star, calibration adopts built-in calibrated radiation source, solar radiation source, moon radiation source or other radiation source on star to coordinate scaling system on the star carrying on satellite regularly to carry out radiation calibration to remote sensor.But on star, scaling system is affected by self optics, electronics, the physical construction of outer space radiation environment, scaling system, and performance can decay in time, affects calibration precision.On the star of the U.S., France, calibration technology is relatively ripe, maintains the leading position in the world.At present Chinese spaceborne remote sensor does not still possess and improves scaling system on reliable star.

Substituting calibrating method adopts underlying surface (land, lake or ocean) based on strict qualification as radiant correction field, utilize remote sensor entrance pupil spoke brightness when synchronously planar survey and atmospheric radiation transmission analog satellite pass by, realize the absolute radiometric calibration of remote sensor according to the statistical relationship between itself and remote sensing images gray-scale value.The method is subject to generally approving in the world after being proposed by the people such as the U.S. Slater eighties in last century, and China has set up Dunhuang calibration and Qinghai Lake radiant correction field the nineties in last century.Alternative calibrating method is also that current Chinese spaceborne remote sensor upgrades the main calibrating method that calibration coefficient adopts in orbit.But substitute calibration and consume a large amount of human and material resources and financial resources, calibration cost is high, can not meet high frequency time calibration demand.The present invention is exactly in order to solve satellite borne sensor high frequency time a kind of new technique of radiation calibration in-orbit.

Summary of the invention

The object of this invention is to provide a kind of satellite borne sensor Calibration Method based on atmospheric parameter remote-sensing inversion, it overcomes, and the alternative calibrating method of satellite borne sensor is expensive, the limitation of the low frequency, be a kind of high precision, low cost, can realize the calibrating method of high frequency time, Pixel-level, and the method can realize the radiation calibration to historical remotely-sensed data.

Technical solution of the present invention is: the ground synchronous of passing by via satellite actual measurement or historical data are obtained Reflectivity for Growing Season, obtain imaging time and observation geometry parameter by remotely-sensed data header file, atmospheric parameter when the weather satellite sensor that utilization is synchronously passed by or the imaging of associated load inverting remote sensor, utilize atmospheric radiation transmission to calculate the brightness of remote sensor entrance pupil spoke according to atmospheric parameter inversion result, and calculate calibration coefficient by radiation calibration model, the radiation calibration in-orbit of realizing satellite borne sensor, its concrete steps are as follows:

Step 1: select the remote sensing image data of ceiling unlimited, in image, comprise high reflectance earth's surface and antiradar reflectivity earth's surface, right

Image carries out the pre-service such as the rejecting of bad wave band and noise reduction, as target image undetermined;

Step 2: obtain target imaging region Reflectivity for Growing Season data;

Step 3: read remote sensor imaging time and observation geometry parameter from remotely-sensed data header file;

Near step 4: select different types of even atural object as region of interest (ROI) remote sensing image to be calibrated center (sub-satellite point);

Step 5: choose space distribution synchronous or the accurate satellite remote sensing date inverting study area atmosphere vapour post content that synchronously passes by;

Step 6: choose space distribution synchronous or the accurate satellite remote sensing date inverting study area Determination of Aerosol Optical that synchronously passes by;

Step 7: choose space distribution synchronous or the accurate satellite remote sensing date inverting study area concentration of ozone column that synchronously passes by;

Step 8: it is smelly that the atmosphere vapour post content that the Reflectivity for Growing Season obtaining according to step 2, the imaging time that step 3 obtains and observation geometry parameter, step 5 obtain, Determination of Aerosol Optical, the step 7 that step 6 obtains obtain

The information such as oxygen concentration, utilize atmospheric radiation transmission MODTRAN to calculate in ROI remote sensor entrance pupil spoke bright

Degree;

Step 9: at different-waveband respectively by DN value and the atmospheric radiation transmission meter of ROI each pixel in image to be calibrated

Linear fit is carried out in the entrance pupil spoke brightness of calculating, and obtains the calibration coefficient of the each wave band of remote sensor, thereby realizes spaceborne

The radiation calibration in-orbit of remote sensor.

Wherein, " selecting the remote sensor image data of ceiling unlimited; comprise high reflectance earth's surface and antiradar reflectivity earth's surface in image; image is carried out to the pre-service such as the rejecting of bad wave band and noise reduction; as target image undetermined " described in step 1, require to comprise in data large area atural object type homogeneous, underlying surface region that reflection characteristic is stable, wherein high reflectance earth's surface type can be desert or Gobi desert etc., and antiradar reflectivity earth's surface type can be dense vegetation etc.

Wherein, " the obtaining target imaging region Reflectivity for Growing Season data " described in step 2, for the stable underlying surface ground table section of reflection characteristic, Reflectivity for Growing Season variation over a period to come can be ignored, and these data can utilize historical measured data to replace; For the unsettled region of reflection characteristic, can utilize pass by simultaneous ground-based measurements reflectivity data or utilize remotely-sensed data that other has higher radiation calibration precision inverting obtains study area Reflectivity for Growing Season through atmospheric correction of satellite.

Wherein, " reading remote sensor imaging time and observation geometry parameter from remotely-sensed data header file " described in step 3, solar zenith angle and position angle, observation zenith angle and position angle when its imaging time comprises remote sensor imaging time, study area longitude and latitude scope, sensor observed altitude, imaging with observation geometry parameter.

Wherein, near " remote sensing image to be calibrated center (sub-satellite point) select different types of even atural object as region of interest (ROI) " described in step 4 selects ROI can reduce the error causing due to the non-lambert's characteristic in earth's surface and sensor large visual field near sub-satellite point.

Wherein, " the choosing space distribution synchronous or the accurate satellite remote sensing date inverting study area atmosphere vapour post content that synchronously passes by " described in step 5, its algorithm is for utilizing near infrared channels reflectance inversion method atmosphere vapour post content, and concrete computation process is as follows:

$\mathrm{\τ}\left({\mathrm{\λ}}_k\right)=\frac{\mathrm{\ρ}\left({\mathrm{\λ}}_k\right)}{m_k\mathrm{\ρ}\left({\mathrm{\λ}}_i\right)\mathrm{\τ}\left({\mathrm{\λ}}_i\right)+n_k\mathrm{\ρ}\left({\mathrm{\λ}}_j\right)\mathrm{\τ}\left({\mathrm{\λ}}_j\right)}$

Wherein τ is atmospheric transmittance, and λ is wavelength, and k is near infrared water vapor absorption wave band, and i, j are atmospheric window wave band, and ρ is Surface reflectance, m _kand n _kfor denominator multinomial coefficient:

$m_k=\frac{{\mathrm{\λ}}_k-{\mathrm{\λ}}_j}{{\mathrm{\λ}}_i-{\mathrm{\λ}}_j}$

$n_k=\frac{{\mathrm{\λ}}_i-{\mathrm{\λ}}_k}{{\mathrm{\λ}}_i-{\mathrm{\λ}}_j}$

The pass of single channel atmosphere vapour post content and atmospheric transmittance is:

$w={\left(\frac{\mathrm{\α}-\ln \mathrm{\τ}}{\mathrm{\β}}\right)}^2$

Wherein w is steam post content, and α, β are empirical constant, for compound underlying surface α=0.02, β=0.651.

Each absorption bands is tried to achieve and is asked weighted mean value can obtain final Water Vapor Content result after steam post content to be:

K wk of w=∑

Wherein weight is:

$f_i=\frac{{\mathrm{\η}}_i}{\mathrm{\Σ}{\mathrm{\η}}_i}$

${\mathrm{\η}}_i=\frac{{\mathrm{\Δ\τ}}_i}{\mathrm{\Δw}}$

Wherein Δ τ is the difference of the corresponding transmitance of each wave band, and Δ w is the difference of Water Vapor Content maximal value and minimum value.

Wherein, " the choosing space distribution synchronous or the accurate satellite remote sensing date inverting study area Determination of Aerosol Optical that synchronously passes by " described in step 6, its method is for utilizing dark pixel inversion method Determination of Aerosol Optical, and concrete computation process is as follows:

NDVI=(L _nir-L _red)/(L _nir+L _red)

Wherein, NDVI is normalized differential vegetation index, L _nirfor the brightness of near-infrared band spoke, L _redfor the brightness of red spectral band spoke.Judge dark pixel by NDVI>0.6.

Relation between aerosol optical depth, Scattering Phase Function and atmospheric path radiation can be expressed as:

${\mathrm{\ρ}}_{\mathrm{path}}({\mathrm{\θ}}_v,{\mathrm{\θ}}_s,\mathrm{\φ})={\mathrm{\ρ}}_m({\mathrm{\θ}}_v,{\mathrm{\θ}}_s,\mathrm{\φ})+\frac{{\mathrm{\ω}}_0{\mathrm{\τ}}_0{P}_{\mathrm{path}}({\mathrm{\θ}}_v,{\mathrm{\θ}}_s,\mathrm{\φ})}{4\cos {\mathrm{\θ}}_v\cos {\mathrm{\θ}}_s}$

Wherein ρ _pathfor total atmospheric path radiation, ρ _mfor the journey radiation that molecular scattering causes, ω ₀for single scattering albedo, τ ₀for aerosol optical depth, P _pathfor aerosol scattering phase function, θ _vfor observation zenith angle, θ _sfor solar zenith angle, φ is observed azimuth.

Consider that earth's surface lambert's characteristic has:

${\mathrm{\ρ}}^{*}={\mathrm{\ρ}}_m+\frac{{\mathrm{\ω}}_0{\mathrm{\τ}}_0{P}_{\mathrm{path}}}{4\cos {\mathrm{\θ}}_v\cos {\mathrm{\θ}}_s}+\frac{F_d\·T\·{\mathrm{\ρ}}^{\′}}{1-S{\mathrm{\ρ}}^{\′}}$

Wherein ρ ^*for apparent reflectance, the Reflectivity for Growing Season that ρ ' is lambertian reflection characteristics, F _dfor ground surface reflectance is 0 o'clock normalization downward radiation flux, T is the total percent of pass of atmosphere, and S is atmospheric backscatter ratio.

Determine aerosol type according to the earth's surface coverage condition in imageable target area, utilize atmospheric radiation transmission, set up the look-up table between the parameters such as atmospheric envelope top apparent reflectance, aerosol optical depth, Reflectivity for Growing Season, solar zenith angle and position angle, observation zenith angle and position angle.Obtain red, blue wave band aerosol optical depth by look-up table, ask the two mean value to obtain 550nm aerosol optical depth, and utilize the method for contiguous pixel interpolation to obtain the aerosol optical depth on bright earth's surface in ROI.

Wherein " choosing space distribution synchronous or the accurate satellite remote sensing date inverting study area concentration of ozone column that synchronously passes by " described in step 7, its concrete computation process is as follows:

Only considering that under atmospheric molecule Rayleigh scattering and Ozone Absorption condition, single scattering calculating formula can be expressed as:

$I_{\mathrm{ss}}=F_{\mathrm{\λ}}\frac{{\mathrm{\β}}_{\mathrm{\λ}}^{*}P\left(\mathrm{\Θ}\right)}{4\mathrm{\π}}{\∫}_0^{\mathrm{Ps}}\exp [-S_x\left(p\right){\mathrm{\α}}_{\mathrm{\λ}}^{*}X\left(p\right)-S_p\left(p\right){\mathrm{\β}}_{\mathrm{\λ}}^{*}p]\mathrm{dp}$

Wherein, F _λfor af at wavelength lambda solar irradiance, for the effective rayleigh scattering coefficient of unit atmospheric pressure, Rayleigh scattering phase function when P (Θ) is scatter angle theta, for Ozone Absorption coefficient, S _x(p) be the oblique journey ozone quality that the radiation of p barosphere is passed, S _p(p) be oblique journey air quality, X (p) is the above ozone post of p barosphere content, P _ssurperficial air pressure.

Multiple Scattering and reflected radiation I _msrjust drill calculating, be by calculating respectively passage built-up radiation I _mwith single scattering radiation I _ss, utilize the two to ask difference operation to obtain I _msr, be R in effective albedo ^*, effective atmosphere P ^*equivalent lambert's body, ultraviolet back scattering built-up radiation computing formula is:

$I_m=I_0+\frac{R^{*}T}{1-{R^{*}S}_b}$

Wherein, I ₀for calculating parameter, T is the ratio that back scattering radiation enters satellite sensor, S _bfor hemispherical reflectance, these three parameters are all x (p), θ ₀, temperature profile T (p) and effective atmosphere P ^*function.Determine in observation condition situation, effectively albedo R ^*with effective atmosphere P ^*all determine, can calculate I _mand I _ss, Multiple Scattering and reflected radiation I _msrfor:

I _msr=I _m-I _ss

So just can set up the Multiple Scattering and the reflected radiation look-up table that calculate for fast.

Ozone profile iterative computation Rodgers method formula table is shown:

$x_{n+1}=x_0+S_X{K}_n^T{(K_n{S}_X{K}_n^T+S_e)}^{-1}[(y_m-y_n)-K_n(x_0-x_n)]$

Wherein, x _n+1for n concentration of ozone column inverting iterative computation result, x ₀with x ₁be respectively priori concentration of ozone column and estimate initial value, S _xfor priori concentration of ozone column covariance matrix, S _efor Instrument observation covariance matrix, y _mfor the large pneumatic jack observation vector of instrument, y _niterative computation obtains observation vector for n time, K _nfor the nuclear matrix of n iterative computation.

It is generally acknowledged Multiple Scattering and reflected radiation I _msrand I _ssbetween linear, can obtain:

I _msr=a(θ ₀,Ω)+b(θ ₀,Ω)I _m

Wherein, a and b depend on solar zenith angle θ ₀parameter with total amount of ozone Ω.

Set up the quick calculating look-up table between each parameter by atmospheric radiation transmission, obtain I by interpolation calculation _msr.Can inverting obtain concentration of ozone column like this.

Wherein " utilizing atmospheric radiation transmission MODTRAN to calculate remote sensor entrance pupil spoke brightness in ROI " described in step 8, its concrete grammar is: select suitable Atmospheric models according to imaging time and observation geometry parameter, the Reflectivity for Growing Season obtaining according to step 2, the atmosphere vapour post content that the imaging time of step 3 acquisition obtains with observation geometry parameter, step 5, the Determination of Aerosol Optical that step 6 obtains, concentration of ozone column, the calculating ROI region corresponding remote sensor entrance pupil spoke brightness that step 7 obtains.

Wherein " respectively the entrance pupil spoke brightness of the DN value of ROI each pixel in image to be calibrated and atmospheric radiation transmission calculating being carried out to linear fit at different-waveband " described in step 9, its concrete grammar is:

L=gain×DN+offset

Wherein L is the brightness of remote sensor entrance pupil spoke, and gain is gain, and DN is the digital count value of remote sensor imaging, and offset is side-play amount.The brightness of entrance pupil spoke and the digital quantization output DN value linear fit that utilize the simulation of step 8 result, can calculate i wave band calibration coefficient:

L _i=gain _i×DN _i+offset _i

The present invention's advantage is compared with prior art:

(1) the present invention is by the satellite borne sensor Pixel-level Calibration Method in-orbit based under the support of atmospheric parameter remote-sensing inversion, can obtain the atmosphere parameter of each pixel and corresponding entrance pupil spoke brightness, make radiation calibration be accurate to Pixel-level, and view picture remote sensing image is applied unified atmospheric parameter in conventional calibrating method, do not consider the otherness of study area atmospheric parameter with spatial variations.The present invention is more accurate for remote sensor entrance pupil spoke brightness simulation, therefore can obtain higher calibration precision.

(2) by the satellite borne sensor Pixel-level Calibration Method in-orbit based under the support of atmospheric parameter remote-sensing inversion, atmospheric parameter when the accurate scaled remote sensor imaging of inverting, has overcome atmospheric parameter simultaneous ground-based measurements in alternative calibrating method and has tested a large amount of human and material resources consumption and calibration cycle length and the representative poor problem of single-point atmospheric parameter measurement brought.Realize high precision, high frequency time, satellite borne sensor radiation calibration in-orbit cheaply, made up the deficiency in existing alternative Calibration Method.

(3) utilize the satellite borne sensor Pixel-level Calibration Method in-orbit based under the support of atmospheric parameter remote-sensing inversion, data selection regional extent is wide, and the time is flexible, can realize the historical data calibration to remote sensor.

Brief description of the drawings

Fig. 1 is techniqueflow block diagram of the present invention.

Embodiment

For a kind of satellite borne sensor Calibration Method based on atmospheric parameter remote-sensing inversion that explanation the present invention relates to better, environment and disaster monitoring forecast moonlet imaging spectrometer (HJ1A/HSI) is carried out to radiation calibration in-orbit.See Fig. 1, a kind of satellite borne sensor Calibration Method based on atmospheric parameter remote-sensing inversion of the present invention, specific implementation method is as follows:

Step 1: choose HJ/HSI dunhuang area on August 21st, 2010,4: 43: 26 imaging datas, choose the stable underlying surface region of earth's surface large area atural object type homogeneous in view data, reflection characteristic as target area;

Step 2: obtain dunhuang area Reflectivity for Growing Season data summer in 2010;

Step 3: obtain imaging time and observation geometry parameter according to HJ/HSI header file;

Step 4: evenly choose 50 sampled points as ROI in the stable region of underlying surface type of ground objects homogeneous, reflection characteristic in HJ/HSI image;

Step 5: choose Hyperion dunhuang area on August 21st, 2010, atmosphere vapour post content when 4 o'clock 19 minutes 36 seconds imaging data inversion imagings;

Step 6: choose Hyperion dunhuang area on August 21st, 2010, Determination of Aerosol Optical when 4 o'clock 19 minutes 36 seconds imaging data inversion imagings.Calculate red, blue wave band aerosol optical depth by look-up table, ask the two mean value to obtain 550nm aerosol optical depth, after interpolation, obtain bright earth's surface sampled point Determination of Aerosol Optical in ROI;

Step 7: the atmospheric ozone post content while choosing SCIAMACHY dunhuang area data inversion imaging on the 21st August in 2010;

Step 8: according to Reflectivity for Growing Season, imaging time and observation geometry parameter, steam post content, aerosol optical depth, ozone post content, selects middle latitude Atmospheric models in summer, utilizes atmospheric radiation transmission to calculate the entrance pupil spoke brightness in ROI;

Step 9: for different-waveband, each pixel DN value is carried out to linear fit with the brightness of corresponding entrance pupil spoke, each pixel in the same band is got to average and reduce stochastic error, calculate each wave band calibration coefficient:

L _i=gain×DN _i?。

Claims (10)

1. the satellite borne sensor Calibration Method based on atmospheric parameter remote-sensing inversion, is characterized in that: the method concrete steps are as follows:

Step 1: select the remote sensing image data of ceiling unlimited, comprise high reflectance earth's surface and antiradar reflectivity earth's surface in image, image is carried out to bad wave band and reject and noise reduction pre-service, as target image undetermined;

Step 2: obtain target imaging region Reflectivity for Growing Season data;

Step 3: read remote sensor imaging time and observation geometry parameter from remotely-sensed data header file;

Near step 4: select different types of even atural object as region of interest ROI remote sensing image to be calibrated center is sub-satellite point;

Step 5: choose space distribution synchronous or the accurate satellite remote sensing date inverting study area atmosphere vapour post content that synchronously passes by;

Step 6: choose space distribution synchronous or the accurate satellite remote sensing date inverting study area Determination of Aerosol Optical that synchronously passes by;

Step 7: choose space distribution synchronous or the accurate satellite remote sensing date inverting study area concentration of ozone column that synchronously passes by;

Step 8: the Reflectivity for Growing Season obtaining according to step 2, the imaging time that step 3 obtains and the atmosphere vapour post content of observation geometry parameter, step 5 acquisition, the Determination of Aerosol Optical that step 6 obtains, the concentration of ozone column information that step 7 obtains, utilize atmospheric radiation transmission MODTRAN to calculate remote sensor entrance pupil spoke brightness in ROI;

Step 9: respectively the entrance pupil spoke brightness of the DN value of ROI each pixel in image to be calibrated and atmospheric radiation transmission calculating is carried out to linear fit at different-waveband, obtain the calibration coefficient of the each wave band of remote sensor, thereby realize the radiation calibration in-orbit to satellite borne sensor, wherein, the digital quantization output valve that DN is remote sensor.

2. a kind of satellite borne sensor Calibration Method based on atmospheric parameter remote-sensing inversion according to claim 1, it is characterized in that: " the selecting the remote sensor image data of ceiling unlimited; comprise high reflectance earth's surface and antiradar reflectivity earth's surface in image; image is carried out to bad wave band and reject and noise reduction pre-service; as target image undetermined " described in step 1, require to comprise in data large area atural object type homogeneous, underlying surface region that reflection characteristic is stable, wherein high reflectance earth's surface type is desert or Gobi desert, and antiradar reflectivity earth's surface type is dense vegetation.

3. a kind of satellite borne sensor Calibration Method based on atmospheric parameter remote-sensing inversion according to claim 1, it is characterized in that: " the obtaining target imaging region Reflectivity for Growing Season data " described in step 2, for the stable underlying surface ground table section of reflection characteristic, Reflectivity for Growing Season variation is over a period to come ignored, and these data utilize historical measured data to replace; For the unsettled region of reflection characteristic, utilize pass by simultaneous ground-based measurements reflectivity data or utilize remotely-sensed data that other has higher radiation calibration precision inverting obtains study area Reflectivity for Growing Season through atmospheric correction of satellite.

4. a kind of satellite borne sensor Calibration Method based on atmospheric parameter remote-sensing inversion according to claim 1, it is characterized in that: " the reading remote sensor imaging time and observation geometry parameter from remotely-sensed data header file " described in step 3, solar zenith angle and position angle, observation zenith angle and position angle when its imaging time comprises remote sensor imaging time, study area longitude and latitude scope, sensor observed altitude, imaging with observation geometry parameter.

5. a kind of satellite borne sensor Calibration Method based on atmospheric parameter remote-sensing inversion according to claim 1, it is characterized in that: near " the selecting different types of even atural object as region of interest ROI remote sensing image to be calibrated center is sub-satellite point " described in step 4, near sub-satellite point, select ROI can reduce the error causing due to the non-lambert's characteristic in earth's surface and sensor large visual field.

6. a kind of satellite borne sensor Calibration Method based on atmospheric parameter remote-sensing inversion according to claim 1, it is characterized in that: " the choosing space distribution synchronous or the accurate satellite remote sensing date inverting study area atmosphere vapour post content that synchronously passes by " described in step 5, its algorithm is for utilizing near infrared channels reflectance inversion method atmosphere vapour post content, and concrete computation process is as follows:

Wherein τ is atmospheric transmittance, and λ is wavelength, and k is near infrared water vapor absorption wave band, and i, j are atmospheric window wave band, and ρ is Surface reflectance, m _kand n _kfor denominator multinomial coefficient:

The pass of single channel atmosphere vapour post content and atmospheric transmittance is:

Wherein w is steam post content, and α, β are empirical constant, for compound underlying surface α=0.02, β=0.651;

Each absorption bands is tried to achieve and is asked weighted mean value can obtain final Water Vapor Content result after steam post content to be:

w＝∑ _if _iw _i

Wherein weight is:

Wherein △ τ _ifor the difference of the corresponding transmitance of each wave band, △ w is the difference of Water Vapor Content maximal value and minimum value.

7. a kind of satellite borne sensor Calibration Method based on atmospheric parameter remote-sensing inversion according to claim 1, it is characterized in that: " the choosing space distribution synchronous or the accurate satellite remote sensing date inverting study area Determination of Aerosol Optical that synchronously passes by " described in step 6, its method is for utilizing dark pixel inversion method Determination of Aerosol Optical, and concrete computation process is as follows:

NDVI＝(L _nir-L _red)/(L _nir+L _red)

Wherein, NDVI is normalized differential vegetation index, L _nirfor the brightness of near-infrared band spoke, L _redfor the brightness of red spectral band spoke, judge dark pixel by NDVI>0.6;

Relation table between aerosol optical depth, Scattering Phase Function and atmospheric path radiation is shown:

Wherein ρ _pathfor total atmospheric path radiation, ρ _mfor the journey radiation that molecular scattering causes, ω ₀for single scattering albedo, τ ₀for aerosol optical depth, P _pathfor aerosol scattering phase function, θ _vfor observation zenith angle, θ _sfor solar zenith angle, φ is observed azimuth;

Consider that earth's surface lambert's characteristic has:

Wherein ρ * is apparent reflectance, the Reflectivity for Growing Season that ρ ' is lambertian reflection characteristics, F _dfor ground surface reflectance is 0 o'clock normalization downward radiation flux, T is the total percent of pass of atmosphere, and S is atmospheric backscatter ratio;

Determine aerosol type according to the earth's surface coverage condition in imageable target area, utilize atmospheric radiation transmission, set up the look-up table between atmospheric envelope top apparent reflectance, aerosol optical depth, Reflectivity for Growing Season, solar zenith angle and position angle, observation zenith angle and position angle parameter, obtain red, blue wave band aerosol optical depth by look-up table, ask the two mean value to obtain 550nm aerosol optical depth, and utilize the method for contiguous pixel interpolation to obtain the aerosol optical depth on bright earth's surface in ROI.

8. a kind of satellite borne sensor Calibration Method based on atmospheric parameter remote-sensing inversion according to claim 1, it is characterized in that: " the choosing space distribution synchronous or the accurate satellite remote sensing date inverting study area concentration of ozone column that synchronously passes by " described in step 7, its concrete computation process is as follows:

Only considering that under atmospheric molecule Rayleigh scattering and Ozone Absorption condition, single scattering calculating formula is expressed as:

Wherein, F _λfor af at wavelength lambda solar irradiance, for the effective rayleigh scattering coefficient of unit atmospheric pressure, Rayleigh scattering phase function when P (Θ) is scatter angle theta, for Ozone Absorption coefficient, S _x(p) be the oblique journey ozone quality that the radiation of p barosphere is passed, S _p(p) be oblique journey air quality, X (p) is the above concentration of ozone column content of p barosphere, P _ssurperficial air pressure;

Multiple Scattering and reflected radiation I _msrjust drill calculating, be by calculating respectively passage built-up radiation I _mwith single scattering radiation I _ss, utilize the two to ask difference operation to obtain I _msr, be equivalent lambert's body of R*, effective atmosphere P* in effective albedo, ultraviolet back scattering built-up radiation computing formula is:

Wherein, I ₀for calculating parameter, T is the ratio that back scattering radiation enters satellite sensor, S _bfor hemispherical reflectance, these three parameters are all the function of temperature profile T (p) and effective atmosphere P*; Determine in observation condition situation, effectively albedo R* and effective atmosphere P* determine, can calculate I _mand I _ss, Multiple Scattering and reflected radiation I _msrfor:

I _msr＝I _m-I _ss

So set up and be used for the Multiple Scattering and the reflected radiation look-up table that calculate fast;

Ozone profile iterative computation Rodgers method formula table is shown:

Wherein, x _n+1for n concentration of ozone column inverting iterative computation result, x ₀with x ₁be respectively priori concentration of ozone column and estimate initial value, S _xfor priori concentration of ozone column covariance matrix, S _efor Instrument observation covariance matrix, y _mfor the large pneumatic jack observation vector of instrument, y _niterative computation obtains observation vector for n time, K _nfor the nuclear matrix of n iterative computation;

It is generally acknowledged Multiple Scattering and reflected radiation I _msrwith built-up radiation I _mbetween linear:

I _msr＝a(θ ₀,Ω)+b(θ ₀,Ω)I _m

Wherein, a and b depend on solar zenith angle θ ₀parameter with total amount of ozone Ω;

Set up the quick calculating look-up table between each parameter by atmospheric radiation transmission, obtain I by interpolation calculation _msr, inverting obtains concentration of ozone column like this.

9. a kind of satellite borne sensor Calibration Method based on atmospheric parameter remote-sensing inversion according to claim 1, it is characterized in that: " the utilizing atmospheric radiation transmission MODTRAN to calculate remote sensor entrance pupil spoke brightness in ROI " described in step 8, its concrete grammar is: select suitable Atmospheric models according to imaging time and observation geometry parameter, the Reflectivity for Growing Season obtaining according to step 2, the imaging time that step 3 obtains and observation geometry parameter, the atmosphere vapour post content that step 5 obtains, the Determination of Aerosol Optical that step 6 obtains, the concentration of ozone column that step 7 obtains, calculate the corresponding remote sensor entrance pupil spoke brightness of ROI region.

10. a kind of satellite borne sensor Calibration Method based on atmospheric parameter remote-sensing inversion according to claim 1, it is characterized in that: " respectively the entrance pupil spoke brightness of the DN value of ROI each pixel in image to be calibrated and atmospheric radiation transmission calculating being carried out to linear fit at different-waveband " described in step 9, its concrete grammar is:

L＝gain×DN+offset

Wherein L is the brightness of remote sensor entrance pupil spoke, and gain is gain, the digital quantization output valve that DN is remote sensor, and offset is side-play amount, utilizes the brightness of entrance pupil spoke and the DN value linear fit of step 8 result simulation, calculates i wave band calibration coefficient:

L _i＝gain _i×DN _i+offset _i。