TY - JOUR
T1 - 基于稳定场地的太阳反射波段基准传递定标及其不确定性评估
AU - Hu, Qi
AU - He, Yuqing
AU - Xu, Na
AU - He, Xingwei
AU - Wang, Ling
AU - Wang, Qian
AU - Hu, Xiuqing
AU - Hu, Bin
AU - Xu, Hanlie
N1 - Publisher Copyright:
© 2024 Science Press. All rights reserved.
PY - 2024
Y1 - 2024
N2 - The stability of Pseudo-Invariant Calibration Sites (PICS) contributes significantly to the improvement in calibration accuracy. The number of PICS is increasing as the work continues to advance. Therefore, the frequency of cross calibrations based on dessert sites has been significantly increased. Establishing a generic site-based cross calibration and uncertainty analysis method is necessary to confirm calibration uncertainties for different sites. Our study aims to improve the overall accuracy of satellite remote sensor observations by developing a cross calibration method over desert sites. In this study, a cross calibration and uncertainty assessment scheme aiming at solar bands is described, and the optimal matching scheme of the cross calibration is given by sensitivity analysis of the uncertainty. With image data of Libya sites from MODIS and MERSI-II, the main uncertainty contributors are found in the geometric, temporal, spatial, and spectral domains. For the four aspects, the uncertainty analysis model is independently constructed using the atmospheric radiative transfer model and the bi-directional reflectance distribution function. The sensitivity of each matching condition to the effect of uncertainty is multiplied and simulated by Monte Carlo method. The geometric and atmospheric distribution patterns of satellite matching data are summarized, which is conducted through statistically analyzing the matching data of MODIS and MERSI-II over Libya sites in 2020. The probability distribution density of the matching condition is used as the input condition, and the discrete distribution of the relative deviation of Top-Of-Atmosphere (TOA) reflectance is obtained by the uncertainty analysis model. The standard deviation of the distribution of relative deviations of TOA reflectance is statistically considered the standard uncertainty. After independent analysis of each factor of uncertainty, the total uncertainty is obtained by the Root-Sum-Squared method. The total uncertainty of each channel could be controlled under 1.5% (at) when the difference in sensor zenith angles between the two remote sensors is than ±7°, the difference between the solar zenith angles is less than ±6°, the aerosol thickness is less than 0.39, and the uniformity of the observation site is less than 0.02. The results between the MODIS reflectance and the digital number recorded by MERSI reveal a good linear relationship. This cross calibration result also has an accuracy in the range of 0.5%—1.5% for each band compared with operational calibrations. Although we only applied the algorithm to MERSI-II as a demonstration, our algorithm is applicable to other sensors with few modifications.
AB - The stability of Pseudo-Invariant Calibration Sites (PICS) contributes significantly to the improvement in calibration accuracy. The number of PICS is increasing as the work continues to advance. Therefore, the frequency of cross calibrations based on dessert sites has been significantly increased. Establishing a generic site-based cross calibration and uncertainty analysis method is necessary to confirm calibration uncertainties for different sites. Our study aims to improve the overall accuracy of satellite remote sensor observations by developing a cross calibration method over desert sites. In this study, a cross calibration and uncertainty assessment scheme aiming at solar bands is described, and the optimal matching scheme of the cross calibration is given by sensitivity analysis of the uncertainty. With image data of Libya sites from MODIS and MERSI-II, the main uncertainty contributors are found in the geometric, temporal, spatial, and spectral domains. For the four aspects, the uncertainty analysis model is independently constructed using the atmospheric radiative transfer model and the bi-directional reflectance distribution function. The sensitivity of each matching condition to the effect of uncertainty is multiplied and simulated by Monte Carlo method. The geometric and atmospheric distribution patterns of satellite matching data are summarized, which is conducted through statistically analyzing the matching data of MODIS and MERSI-II over Libya sites in 2020. The probability distribution density of the matching condition is used as the input condition, and the discrete distribution of the relative deviation of Top-Of-Atmosphere (TOA) reflectance is obtained by the uncertainty analysis model. The standard deviation of the distribution of relative deviations of TOA reflectance is statistically considered the standard uncertainty. After independent analysis of each factor of uncertainty, the total uncertainty is obtained by the Root-Sum-Squared method. The total uncertainty of each channel could be controlled under 1.5% (at) when the difference in sensor zenith angles between the two remote sensors is than ±7°, the difference between the solar zenith angles is less than ±6°, the aerosol thickness is less than 0.39, and the uniformity of the observation site is less than 0.02. The results between the MODIS reflectance and the digital number recorded by MERSI reveal a good linear relationship. This cross calibration result also has an accuracy in the range of 0.5%—1.5% for each band compared with operational calibrations. Although we only applied the algorithm to MERSI-II as a demonstration, our algorithm is applicable to other sensors with few modifications.
KW - Monte Carlo method
KW - cross calibration
KW - medium resolution imaging spectroradiometer
KW - reflection band
KW - remote sensing
KW - uncertainty
UR - http://www.scopus.com/inward/record.url?scp=85202292394&partnerID=8YFLogxK
U2 - 10.11834/jrs.20232095
DO - 10.11834/jrs.20232095
M3 - 文章
AN - SCOPUS:85202292394
SN - 1007-4619
VL - 28
SP - 2045
EP - 2061
JO - National Remote Sensing Bulletin
JF - National Remote Sensing Bulletin
IS - 8
ER -