Gu Xingfa

Radiometric cross-calibration of the CBERS-02 CCD camera with the TERRA MODIS

For the application of the CCD camera, the most important payload on CBERS-02, the key is to provide long-term stable radiometric calibration coefficients. Although the vicarious calibration had been proved successful, it had its limitations such as test site requirement and unsuitable for historical data. Cross-calibration is one of the alternative methods, but it needs synchro surface spectrum to achieve spectral band matching factors. Our effort is to probe the influences on these factors. Simulations with a lot of surface spectrum showed that the factors changed with the viewing geometry, atmospheric condition and surface targets. However, simulating with the same viewing geometry and atmospheric condition, the spectral band matching factors of the same or similar surface targets, spectrum acquired from different dates and different places would like to be consistent to each other within 1%–5%. Thus, the synchro measurement data can be substituted by the same or similar target from other source. Based on this method, using the MODIS as the reference, the cross-calibration was performed for CCD camera. The research demonstrated that the traditional method with single calibration site was inappropriate for CCD camera, since the offsets for its four spectral bands were not zeros. With four calibration sites, these offsets were obtained. And the camera was detected to degrade with dates based on four times of cross-calibrations.For the application of the CCD camera, the most important payload on CBERS-02, the key is to provide long-term stable radiometric calibration coefficients. Although the vicarious calibration had been proved successful, it had its limitations such as test site requirement and unsuitable for historical data. Cross-calibration is one of the alternative methods, but it needs synchro surface spectrum to achieve spectral band matching factors. Our effort is to probe the influences on these factors. Simulations with a lot of surface spectrum showed that the factors changed with the viewing geometry, atmospheric condition and surface targets. However, simulating with the same viewing geometry and atmospheric condition, the spectral band matching factors of the same or similar surface targets, spectrum acquired from different dates and different places would like to be consistent to each other within 1%–5%. Thus, the synchro measurement data can be substituted by the same or similar target from other source. Based on this method, using the MODIS as the reference, the cross-calibration was performed for CCD camera. The research demonstrated that the traditional method with single calibration site was inappropriate for CCD camera, since the offsets for its four spectral bands were not zeros. With four calibration sites, these offsets were obtained. And the camera was detected to degrade with dates based on four times of cross-calibrations.

In flight MTF monitoring and compensation for CCD camera on CBERS-02

In this article, the approach of simulating ideal tarp scene was proposed to determine the MTF for CCD camera on CBERS-02. The MTF acquired from this technique was compared to those from some common methods. MTFs achieved from different approaches were employed to compensate the CCD images based on three restoration algorithms: the iterative method, the Wiener filter and the modified inverse filter (MIF). The two-dimensional MTF was established with the value at frequency 0.5 for 45° directional MTF as the value at its four corners. This article addressed the variations of the restored CCD images with different techniques of constructing two-dimensional MTF, restoring algorithms and determining the in-flight MTFs. Results showed images restored with the MTF determined from the technique of simulating ideal tarp scene represent the real scene best. And the restoration technique based on the MIF would produce better restored images than the other two techniques. This research also demonstrated that the two-dimensional MTF used for restoration should be interpolated under the control of the MTF for 45° axis, rather than by multiplying the cross-orbit MTF with along-orbit MTF.In this article, the approach of simulating ideal tarp scene was proposed to determine the MTF for CCD camera on CBERS-02. The MTF acquired from this technique was compared to those from some common methods. MTFs achieved from different approaches were employed to compensate the CCD images based on three restoration algorithms: the iterative method, the Wiener filter and the modified inverse filter (MIF). The two-dimensional MTF was established with the value at frequency 0.5 for 45° directional MTF as the value at its four corners. This article addressed the variations of the restored CCD images with different techniques of constructing two-dimensional MTF, restoring algorithms and determining the in-flight MTFs. Results showed images restored with the MTF determined from the technique of simulating ideal tarp scene represent the real scene best. And the restoration technique based on the MIF would produce better restored images than the other two techniques. This research also demonstrated that the two-dimensional MTF used for restoration should be interpolated under the control of the MTF for 45° axis, rather than by multiplying the cross-orbit MTF with along-orbit MTF.

Biomass estimation and uncertainty analysis based on CBERS-02 CCD camera data and field measurement

The study site is located in Qianyanzhou experimental station and the surrounding area. Based on CBERS-02 satellite data and field measurement, we not only discussed the relationship between NDVI and biomass of two species of coniferous plantations, namely,Pinus massoniana Lamb andPinus elliottii Engelm, but also introduced the biomass models based on NDVI. The comparison between measured biomass in Qianyanzhou and biomass derived from CBERS-02 CCD data showed that it is feasible to estimate biomass based on NDVI. But its limitations cannot be ignored. This kind of model depends on the dominant vegetation species. There are some effect factors in estimating biomass based on NDVI. This paper analyzes these factors based on fine-resolution CBERS-02 CCD data and some conclusions are drawn: In Qianyanzhou, the area with good vegetation coverage, the nonlinearity of NDVI has little influence on scaling-up of NDVI. As a result of surface heterogeneity, scaling-up can cause NDVI within each pixel to change. Because scaling-up can cause pixel attribute to change, the applicability of biomass model is one of the sources of error in estimating biomass.

Comparison of the influence factors onNDVI for CCD camera and WFI imager on CBERS-02

CCD and WFI are payloads in the visible light and near infrared bands on CBERS satellites and they have not been applied widely, especially WFI. This article mainly probes into the potential of these two sensors’ application in the vegetation monitoring and analyses the influences of radiometric calibration, atmosphere conditions and sun-observing geometry, etc. onNDVI obtained from these two sensors respectively. In addition it analyses the effect of the spectral response difference of red and near infrared spectral bands in CCD and WFI sensors onNDVI. Results indicate that radiometric calibration is the most important factor onNDVI, which cannot be applied to vegetation monitoring without radiometric calibration. Results also demonstrate that near surfaceNDVI is different greatly from that of TOA which change with atmosphere conditions. The study shows thatNDVI is also affected by the non-lambertian character of surface and the change of the atmospheric path with the observing geometry. AndNDVI of WFI is higher than that of CCD. They are very different because of their spectral diffirence, but they have a good linear relevant relationship.

Study of thin cloud removal method for CBERS-02 image

The China Brazil Earth Resources Satellite (CBERS) ended the dependence on foreign satellites. Along with the series of CBERS and the enhancement of satellite applications, it is necessary to search correlative data processing methods. Because cloud appears in the atmosphere, an image with clear sky conditions is often hard to obtain. How to remove cloud is an important sector of increasing image quality. An improved homomorphism filtering method is adopted to remove thin cloud qualitatively at first, and all bands are selected to classify pixels. Based on the fact that visible band images contain more aerosol effects than infrared bands, visible bands are used to determine clear or hazy regions and the reflectances of different surface classes in clear regions are calculated. At last the mean reflectance of the same classes in clear and hazy regions is matched to remove cloud. The image visual effect is improved; meanwhile, the ground reflectance is retrieved to meet the needs of remote sensing quantification.The China Brazil Earth Resources Satellite (CBERS) ended the dependence on foreign satellites. Along with the series of CBERS and the enhancement of satellite applications, it is necessary to search correlative data processing methods. Because cloud appears in the atmosphere, an image with clear sky conditions is often hard to obtain. How to remove cloud is an important sector of increasing image quality. An improved homomorphism filtering method is adopted to remove thin cloud qualitatively at first, and all bands are selected to classify pixels. Based on the fact that visible band images contain more aerosol effects than infrared bands, visible bands are used to determine clear or hazy regions and the reflectances of different surface classes in clear regions are calculated. At last the mean reflectance of the same classes in clear and hazy regions is matched to remove cloud. The image visual effect is improved; meanwhile, the ground reflectance is retrieved to meet the needs of remote sensing quantification.

Absolute radiometric calibration of CBERS-02 IRMSS thermal band

Based on the laboratory calibration before launch of CBERS-02 IRMSS thermal infrared channel, the onboard blackbody calibration, the radiometric crosscalibration against TERRA MODIS corresponding channel and the in-flight field calibration at Lake Qinghai: water surface radiometric calibration test site of China on Aug. 17, 2004 are carried out in this research. When making onboard blackbody calibration of CBERS-02 IRMSS, it is necessary to understand inside structures and light path in IRMSS camera and receive and process calibration blackbody signals at normal and high temperatures. Onboard blackbody calibration results of each detector are very important to relative radiometric calibration of images processing and absolute radiometric calibration of each detector. In-flight field calibration mainly contains field experiments, obtaining synchronous images, spectrum marching and MODTRAN radiative transmission computation. Radiometric cross-calibration of CBERS-02 IRMSS thermal band against TERRA MODIS selected 6 times synchronous images of two sensors passing through Lake Qinghai and Lake Taihu from August to December, 2004 to compute the cross calibration data. Combining the in-flight field calibration and radiometric cross calibration data, the absolute radiometric calibration coefficients are calculated by the linear regression method. This new calibration model can obviously control the radiometric calibration uncertainties aroused by the single point method used in the in-flight calibration of CBERS-02 IRMSS, this kind of sensors cannot receive the cosmic background radiation. In this research, the radiometric calibration coefficients obtained through the linear regression model are 8.53 (gain, unit: DN/W m−2sr−1μm−1) and 44.92 (offset, unit: DN).Based on the laboratory calibration before launch of CBERS-02 IRMSS thermal infrared channel, the onboard blackbody calibration, the radiometric crosscalibration against TERRA MODIS corresponding channel and the in-flight field calibration at Lake Qinghai: water surface radiometric calibration test site of China on Aug. 17, 2004 are carried out in this research. When making onboard blackbody calibration of CBERS-02 IRMSS, it is necessary to understand inside structures and light path in IRMSS camera and receive and process calibration blackbody signals at normal and high temperatures. Onboard blackbody calibration results of each detector are very important to relative radiometric calibration of images processing and absolute radiometric calibration of each detector. In-flight field calibration mainly contains field experiments, obtaining synchronous images, spectrum marching and MODTRAN radiative transmission computation. Radiometric cross-calibration of CBERS-02 IRMSS thermal band against TERRA MODIS selected 6 times synchronous images of two sensors passing through Lake Qinghai and Lake Taihu from August to December, 2004 to compute the cross calibration data. Combining the in-flight field calibration and radiometric cross calibration data, the absolute radiometric calibration coefficients are calculated by the linear regression method. This new calibration model can obviously control the radiometric calibration uncertainties aroused by the single point method used in the in-flight calibration of CBERS-02 IRMSS, this kind of sensors cannot receive the cosmic background radiation. In this research, the radiometric calibration coefficients obtained through the linear regression model are 8.53 (gain, unit: DN/W m−2sr−1μm−1) and 44.92 (offset, unit: DN).

Method study and uncertainty analysis of calibration coefficients validation based on the Inner Mongolia test site

Calibration coefficients validation is the foundation for ascertaining the sensor performance and carrying out the quantitative application. Based on the analysis of the differences between the calibration and validation, two calibration coefficients validation methods were introduced in this paper. Taking the HJ-1A satellite CCD1 camera as an example, the uncertainties of calibration coefficients validation were analyzed. The calibration coefficients validation errors were simulated based on the measured data at an Inner Mongolia test site. The result showed that in the large view angle, the ground directional reflectance variation and the atmospheric path variation were the main error sources in calibration coefficients validation. The ground directional reflectance correction and atmospheric observation angle normalization should be carried out to improve the validation accuracy of calibration coefficients.

Modeling diurnal variation of ground thermal radiance images using energy balance model and endmember composing technique

Modeling and analyzing dynamic changes of land thermal radiance scenes play an important role in thermal remote sensing. In this paper, the diurnal variation of ground surface thermal scene is mainly discussed. Firstly, based on the land surface energy balance equation, the diurnal variation of land surface temperatures (LSTs) over bare land covers were simulated by an analytical thermal model with second harmonic terms, and the diurnal LST variation of vegetation canopy was simulated using the Cupid model. Secondly, normalized difference vegetation index (NDVI), normalized difference water index (NDWI), and ratio resident-area index (RRI) were used to evaluate the endmember abundance of four land cover types including vegetation, bare soil, impervious and water area, which were calculated from IKONOS visible and near infrared (VNIR) bands. Finally, the thermal radiance scenes at various times and view angles were modeled based on the linear-energy-mixing hypothesis. The results showed that the simulated daily LST variations for vegetated and bare surfaces are correlated with the measured values with a maximum standard deviation of 2.7°C, that land thermal radiant textures with high-resolution are restored from the linear-energy-mixing method, and that the information abundance of the scene are related to the distribution of land cover, the imaging time, and the view angle.

Inversion and validation of leaf area index based on CBERDS02B image data in GuangXi province of China

CBERDS02B satellite has been successfully launched in September 2007, the target of this paper is to get the vegetation index from visible red-band, near-infrared band and the blue-band surface reflectance data of CBERDS02B satellite, through the empirical model of the relations between the vegetation index and LAI, and combined with the classification data to integrate the appropriate model, in order to get the regional leaf area index image in Binyang County of Nanning City in Guangxi Province of China. To make the operation more rapid and feasible, I decided to use an empirical model to obtain LAI, This method is simple and easy to calculate, more realizable, and suitable for remote sensing application. In this paper I use part of the measured data to validate a wide range of VI-LAI models. In order to identify the advantages and disadvantages of the various models, different plants use different types of vegetation model, I finally choose four VIs, such as SR, NDVI, SAVI, EVI, then combine these with the classification data to get the best mixed model so as to attain the leaf area index image of the research region. Then I use the other part of the measured data to get the validation of the mixed model. Ultimately I improve the overall accuracy of the model, and gain more accurate LAI images in the region.

The quantification of remote sensing

The paper firstly defined the remote sensing information quantification, analyzed the necessity of developing remote sensing quantification, figured out the application guidelines requirement, and pointed out the importance of quantification research. Then taking the remote sensing application research of CBERS-02 data quantification as the example, the paper described the whole quantification system of “remotely sensed digital signal-radiation information-field parameter inversion”. Finally the paper gave the prospect for the development trend of the quantitative remote sensing.