ISSN 1000-0526
CN 11-2282/P
Optimal Spectral Region Selection for FY-3D/HIRAS Spectral Calibration Accuracy Evaluation
Author:
Affiliation:

Chinese Academy of Meteorological Sciences, Beijing 100081; National Satellite Meteorological Centre, Beijing 100081; Key Laboratory of Radiometric Calibration and Validation for Environmental Satellites, CMA, Beijing 100081; State Key Laboratory of Severe Weather, Chinese Academy of Meteorological Sciences, Beijing 100081

Clc Number:

Fund Project:

  • Article
  • |
  • Figures
  • |
  • Metrics
  • |
  • Reference
  • |
  • Related
  • |
  • Cited by
  • |
  • Materials
  • |
  • Comments
    Abstract:

    Accurate evaluation and monitoring of spectral frequency accuracy is significant for an hyper-spectral infrared (IR) interferometer before data application. A “cross-correlation method” is commonly used to evaluate the accuracy of the spectral accuracy, which allows the maximum correlation, or minimum standard deviation condition, to be met between an observed and a simulated spectrum by shifting the observed spectrum. From the perspective of the time-consuming calculation, it is not necessary to detect the frequency offset in the entire spectrum, but to evaluate the accuracy of the spectral calibration with a part of the spectral region. To comprehensively evaluate the dependence of the “cross-correlation method” on spectral regions, a preliminary selection of spectral regions was carried out based on the sensitivity analysis of the simulated spectra (with only the off-axis effect of the instrument considered), and the optimal spectral region was selected based on the sensitivity analysis of the on-orbit data. The results based on simulated spectra showed that spectral region selection is sensitive to the evaluation method in long-wave (LW) and short-wave (SW) bands, but insensitive in mid-wave (MW) band. In addition, the sensitivity is rela-ted to the envelope characteristics and to the radiation energy of the absorption line in the spectral regions. The absolute errors for selecting different spectral regions can reach 3.05 and 3.35 ppm (1 ppm=10-6) in LW and SW bands, respectively. Selecting high-radiation-energy and stable atmospheric composition regions can effectively reduce the error introduced by the “cross-correlation method” . The results using on-orbit data of high-spectral-resolution infrared atmospheric sounder (HIRAS), which was firstly carried on the Fengyun 3D (FY-3D) satellite, showed that the instrument observation error and radiative transfer model simulation error are also needed to be considered in practical data applications to achieve the best spectral regions. Finally the best reference spectral regions for HIRAS spectral accuracy evaluation were obtained, located in 716-766 cm-1 for LW band, 1 〖KG-*5〗270-1 〖KG-*5〗320 cm-1 for MW band and 2 〖KG-*5〗159-2 〖KG-*5〗209 cm-1 for SW band, respectively. The mean spectral bias results based on the selected best reference spectra regions for HIRAS are all less than 2 ppm for the three bands, spectral bias standard deviations are less than 2 ppm for LW and MW bands, and about 4 ppm for SW band. The obtained best spectral regions are also applicable to the spectral frequency accuracy evaluation and long-term frequency monitoring of other infrared interference instruments.

    Reference
    Related
    Cited by
Get Citation
分享
Article Metrics
  • Abstract:
  • PDF:
  • HTML:
  • Cited by:
History
  • Received:April 10,2019
  • Revised:March 09,2020
  • Adopted:
  • Online: June 29,2020
  • Published:

WeChat

Mobile website