Estimation of Argo sea subsurface temperature based on a thermal parametric model

Argo has become an important part of the global ocean observation system. While, due to the lack of sea surface measurements, application of the Argo data still has some limitations. Based on a simple thermal parametric model, the relationship between sea surface temperature and subsurface temperature was constructed in this paper by the thermal parameters calculated from Argo profile data and WOA09 climatic data. And then sea surface temperature was estimated by Argo subsurface data in the Pacific Ocean. Among them, the thermal parameters were calculated by the maximum angle method. The novel objective method produced the more accurate thermal parameters than those by the iterative technique used in previous studies, such as mixed layer depth, thermocline temperature gradient, thermocline bottom depth and so on. Compared with the SST extrapolated from the traditional method and retrieved from satellite SST,the RMSEs between the estimated Argo surface temperature and the observed GTSPP or Argo NST have been significantly reduced; the correlation coefficient between estimation result and in situ observation was larger as well. The Argo surface temperature estimated by the thermal parametric model in the Pacific Ocean was theoretically proved to be reliable through correlation analysis. This study made up for the lacking of surface observation in Argo. It provided a new way toestablish complete Argo data sets.