Which area should be preferentially observed to improve the vertical thermal structure prediction: A case study focusing on complex topography areas in the South China Sea.

Abstract In this case study, the rapidly growing initial errors in vertical thermal structure (VTS) prediction associated with complex topography areas in the South China Sea were captured by the Regional Ocean Modeling System (ROMS) in combination with the Conditional Nonlinear Optimal Perturbation (CNOP) approach. Two types of CNOP-based rapidly growing initial errors with opposing patterns were concentrated both vertically within the thermocline range and horizontally along the coast. The associated preferential observation area was further determined by computing the vertically integrated total energy of these errors over the entire depth range. The results demonstrated that the additional observations should be prioritized in areas with significant horizontal depth gradients, particularly in steep slope zones, in order to improve the VTS prediction in complex topography areas in the South China Sea. To confirm the effect of observation suggestions, relevant observation system simulation experiments (OSSEs) and observation system experiments (OSEs) will be carried out in the future.