Abstract: The present disclosure discloses a method for reconstructing geostationary ocean color satellite data based on Data INterpolating Empirical Orthogonal Functions. The method includes steps of: 1) for an original ocean color remote sensing area, using concentration detection and edge detection to eliminate abnormal image elements; 2) in the Data INterpolating Empirical Orthogonal Functions, using Laplace smoothing filter to smoothly filter a time covariance matrix; 3) for temporal modes value obtained by the Data INterpolating Empirical Orthogonal Functions, first decomposing it into sub-temporal mode, then adopting empirical modal decomposition, and finally performing interpolation, to obtain an ocean color remote sensing reanalysis data set that maintains its original accuracy and full coverage of time and space.

Abstract: The present disclosure discloses a method for reconstructing geostationary ocean color satellite data based on Data INterpolating Empirical Orthogonal Functions. The method includes steps of: 1) for an original ocean color remote sensing area, using concentration detection and edge detection to eliminate abnormal image elements; 2) in the Data INterpolating Empirical Orthogonal Functions, using Laplace smoothing filter to smoothly filter a time covariance matrix; 3) for temporal modes value obtained by the Data INterpolating Empirical Orthogonal Functions, first decomposing it into sub-temporal mode, then adopting empirical modal decomposition, and finally performing interpolation, to obtain an ocean color remote sensing reanalysis data set that maintains its original accuracy and full coverage of time and space.