Back to the main page

Tropical Indian and Pacific SST (40°-280°E,20°S-20°N)



EOF analysis (with Reynold OI monthly SST anomaly data for observation)

  • EOF patterns
  • PC1
  • PC2
  • PC3
  • Sample outcome of a multiple regression with 3 PCs



    Amplitude of POAMA SST (with HadISST and POAMA 3 month running mean SST anomaly data)



    Statistical correction for POAMA SST

  • Mean error field of POAMA SST
  • EOF analysis on the error field: LT2 (eof 1-5) , LT2 (eof 6-10) , LT8 (eof 1-5) , LT8 (eof 6-10)

  • Correcting POAMA's SST, using a multiple regression model (before cross-validation)
  • Correcting POAMA's SST, using a multiple regression model (after cross-validation):
      + Corr.Coeff.
      + Ratio of STD
      + MSE Skill Score
      + CC , RMSE, and SS as functions of leadtime
      + Examples of using skillful predictors for Method III
  • Calibration of POAMA's SST , using EOF, SVD and projection of POAMA sst onto observed EOF patterns at LT 2 and LT 8


    Comparision with the output from a linear inverse model (LIM)



    ( For reminding myself )