| Climate and environment are the basis of human survival.The long-term global observation capability of atmospheric satellite remote sensing makes it have unique technical advantages in global climate and environmental monitoring.The Directional Polarimetric Camera(DPC)of GF-5 satellite increases the detection ability of multi-angle polarization information on the basis of traditional spectral dimension detection.The expansion of DPC information dimension improves the detection ability of atmosphere and environment.However,with the advantages of information acquisition,the ultra-wide angle and time-sharing imaging technology adopted by DPC also faces some new problems,such as lens polarization,stray light,Photo Response Non-Uniformity and earth curvature in orbit observation.These problems reduce the polarization imaging remote sensing data quality in varying degrees,which affects the retrieval ability of atmospheric elements such as clouds and aerosols.In order to ensure the application efficiency of atmospheric polarization remote sensing instrument,it is necessary to eliminate or reduce the influence of these factors on the quality of polarization remote sensing data by means of correction and validation.DPC is the first ultra-wide angle polarization imaging instrument in China aiming at business operation.In order to ensure the satellite remote sensing quality of the payload,this paper studies the data correction method,system design for business operation and on orbit validation in large field of view,which adapt to DPC quantitative remote sensing application:1)Accurate mathematical model and system integrated data correction algorithm are the guarantee of data accuracy and the premise of business operation.The large field of view transmission polarization radiation response model of DPC is derived,and the data correction process of modularization is designed.According to the optical information transmission process,the detector calibration module and optical system calibration module are developed to avoid error source aliasing.Considering the influence of temperature change on the spectral response of DPC,a set of general electro-optical parameters testing system of the image sensor with temperature control function is developed to study the calibration method of the detector.A detector temperature control method comprised of thermoelectric cooler based on Peltier effect and low-temp circulator is proposed,the temperature control precision is better thaną0.15?.In view of the on orbit working mode of DPC based on integral time adjustment,the non-uniformity correction method of detector based on multi parameters compensation is designed.After a series of preprocessing of the image including removal of darkness signal,removal of smearing effect and temperature compensation,the low-frequency unbalanced response difference of the image surface is eliminated,and the high-frequency difference of the field pixel is effectively suppressed.After the systematic correction of the detector,the radiation measurement error caused by temperature fluctuation is less than 0.1%,and the non-uniformity of single frame data is reduced from 1.141%to 0.513%.Through a large number of laboratory measurement experiments,the calibration parameter data are obtained,and the calibration and correction module effectively controlled the measurement error of DPC:the stray light radiation of high reflectivity cloud in near-infrared band up to 80%increment to the surrounding water was effectively corrected;the radiation measurement error caused by polarization was reduced to less than 0.34%;the field of view difference of time-sharing measurement of the filter and polarizer detection channel was effectively compensated by optical wedge,and the deviation was less than 0.1 pixel.After data correction,the DPC radiation measurement error is less than 5%,and the polarization measurement error is less than 2%which is verified by adjustable polarization light source and field sky light.The data correction method meets the requirements of instrument measurement accuracy.2)Before the on orbit operation of the instrument,the development of data correction ground system and DPC are carried out simultaneously.As a new research system,Due to the lack of effective reference and on orbit practical data support,firstly,the DPC on orbit data simulation method is used to carry out the system development.After on orbit operation,the system is verified by the actual on orbit data.Based on the data simulation methods such as orbit simulation,vector radiation transmission,photoelectric conversion and spatial projection,the characteristics of DPC multi-angle imaging data are studied to realize spatial matching of multi-angle and multi-spectral data with different spatial resolution.Based on this,the data format and interface of DPC are designed.The hierarchical compressed data format is used to effectively manage remote sensing data.In order to adapt to the characteristics of atmospheric change and meet the requirements of timeliness,a calculation program based on distributed processing is designed to realize the operation of DPC data correction.The program optimizes the convolution operation of stray light correction and the iterative operation of geometric correction.In a variety of computing environments,the data correction calculation can meet the timeliness requirements within 1 hour.During the on orbit test,the data correction data of 1609 orbits were pushed to complete the operation test of data correction ground system.Considering that the number of successive DPC multi-angle observations increased from 9 to 17?34,the data correction program adopts modular and extensible design.3)During the on orbit test,aiming at the timeliness requirements of multi-angle and large field of view validation,the on orbit validation method is studied,and the on orbit validation of key parameters of polarization correction is realized based on cloud polarization.During on orbit test,the compensation effect of wedge is evaluated by the image feature point algorithm with positioning accuracy of 0.02 pixels,which meets the design goal of less than 0.1 pixel deviations.Because the traditional single point statistical method cannot meet the timeliness requirements of large field of view validation,a fast validation method of DPC polarization radiation is studied.The desert field reflectance method based on bidirectional reflectance fitting was used to test the radiation information of multi-angle observation.Through large-scale ocean sun-glint,the fast evaluation of polarization detection performance in large field of view in complex cloudy environment is realized.The results show that the on orbit performance of DPC is consistent with the laboratory,the data correction method is effective,and the validation method meets the timeliness requirements of large field of view validation.In order to monitor the changes of DPC key parameters,the desert site was used to test the effect of multi-angle relative radiation correction.The relative radiation response change was less than 4%,and the pixel response non-uniformity was corrected effectively.Based on the polarization characteristics of cloud,the DPC shows the ability of using polarization information to identify cloud phase state.Continuous multi angle observation shows that the polarization reflectance of water cloud is consistent.The relative transmittance change of polarization parameter is less than 0.2%,and the change of polarizing degree is less than 0.01.The detection method of key parameters provides technical reference for on orbit calibration. |
PDF Full Text Request