Study On Sea Surface Temperature Retrieving Method Base On Infrared Remote Sensing

As an important physical quantity of the sea-air boundary layer,the sea surface temperature plays an important role not only in the whole biosphere,but also in many fields such as military,environment,meteorology and so on.Because of the special temperature of the sea surface,it is the most accurate and effective way to invert the non-contact infrared sensor.Based on the sea surface radiation transfer equation,the sea surface temperature is retrieved and the neural network is used to correct the inversion error.Finally,the inversion value of sea surface temperature is obtained.In this paper,the specificity of the sea surface temperature is discussed,and the basic theory of infrared remote sensing is studied.Based on this,the rationality of applying the radiant energy inversion temperature value is put forward.After studying the thermodynamic properties of the key ocean and the complex sea surface and atmospheric environment,the factors influencing the accuracy of the ocean surface temperature inversion and the specific influence are analyzed.The energy transfer process of the sea epidermis is studied and the transport process is expressed as the energy transfer equation of the sea surface temperature.The qualitative analysis is quantified into the form of theoretical publicity.Based on this,a method of sea surface temperature inversion based on radiation transfer equation is proposed.In order to explain the possible error of the inversion method,this paper presents the method of using neural network to correct the time error of the theoretical model after elaborating the specific point of neural network and its basic working mechanismFour important parameters in the temperature inversion model are analyzed and studied.The theoretical calculation models of the various parameters and the neglected physical processes in the calculation and the possible calculation errors of the parameters are presented.The reason of the change of the parameters and the effect of the parameter variation on the temperature inversion are analyzed.The sensitivity of the sea surface temperature inversion accuracy to the parameters is simulated by the theory and MATLAB simulation.On the basis of the field experiment,the sea surface temperature is multiplied by the inversion model,and compared with the measured value and the real value,the reason of the error is explained and different error correction methods are put forward for different error sources.The neural network algorithm is applied to the error correction of the inversionmodel.And the effect of the correction is evaluated,which proves the validity of neural network in the error correction of sea surface temperature inversion.