A Segment-Based Seawater Temperature Model by Using the Relative Gradient Method

A segment-based model for estimating the vertical profiles of temperature from the sea surface temperature has been developed.Using the Argo Climatology dataset(2005-2012)and the Argo float profiles,the vertical structure of the seawater temperature in the Northwestern Pacific Ocean was analyzed by the Relative Gradient method.The parameters of each layer were obtained including the mixed layer depth,temperature gradient in the mixed layer,upper thermocline depth,lower thermocline depth,deep ocean depth and coefficients for the fitting equations.The CTD,XBT data from World Ocean Database 09 were used for verifying the model.The results showed that this model performs well,especially in the upper ocean(400 m).The estimated temperature had a root mean square error(RMSE)of 0.778℃in all the layers and 0.494℃in the upper 400 m.