Journal of Atmospheric and Environmental Optics, Volume. 19, Issue 2, 232(2024)

Inversion of aerosol optical depth in Guiyang City based on LightGBM

PU Lilan and ZHANG Xianyun*
Author Affiliations
  • College of Mining, Guizhou University, Guiyang 550025, China
  • show less
    Figures & Tables(9)
    Land use type of Guiyang City
    Model learning curve
    The Importance of the features
    AOD prediction values and MODIS AOD trand
    AOD prediction error. (a) Urban; (b) non-urban
    Distribution of the prediction error. (a) Urban; (b) non-urban
    • Table 1. The parameters of GF-4 PMS sensor

      View table
      View in Article

      Table 1. The parameters of GF-4 PMS sensor

      BandSpectral range/μmResolution ratio/mRevisit time/s
      Pan0.45-0.905020
      Blue0.45-0.5250
      Green0.52-0.6050
      Red0.63-0.6950
      Near-infrared (NIR)0.76-0.9050
    • Table 2. The modeling data

      View table
      View in Article

      Table 2. The modeling data

      VariableAbbreviationsSpatial resolutionUnitSource
      Aerosol optical depthAOD3 km-MOD04_3K
      GF-4 surface reflectanceGF4SR50 m-GF-4 PMS
      Solar zenith angleSOZ-(°)GF-4 PMS
      Satellite zenith angleSAZ-(°)GF-4 PMS
      Relative azimuth angleREA-(°)GF-4 PMS
      ElevationEL-mSRTM DEM
      Normalized difference vegetation indexNDVI--GF-4 PMS
    • Table 3. The optimized model parameters

      View table
      View in Article

      Table 3. The optimized model parameters

      Max_depthLearning_rateNum_leavesLambda_l1Lambda_l2Num_boost_round
      70.0336754.04648.66282220
    Tools

    Get Citation

    Copy Citation Text

    Lilan PU, Xianyun ZHANG. Inversion of aerosol optical depth in Guiyang City based on LightGBM[J]. Journal of Atmospheric and Environmental Optics, 2024, 19(2): 232

    Download Citation

    EndNote(RIS)BibTexPlain Text
    Save article for my favorites
    Paper Information

    Category:

    Received: Oct. 20, 2022

    Accepted: --

    Published Online: Jun. 24, 2024

    The Author Email: Xianyun ZHANG (mec.xyzhang@gzu.edu.cn)

    DOI:10.3969/j.issn.1673-6141.2024.02.009

    Topics