A point of view on semi-empirical thermospheric models

Abstract

Different types of models of the thermosphere, either with experimental characteristics or with theoretical characteristics, have been elaborated since the Space Age. These models describe the variation of different parameters (temperature, density, pressure, concentrations,…). A brief description of these models is given. As far as theoretical models are concerned, the conservation equations of energy, of momentum or mass and physico-chemical processes are taken into account. Experimental models are based upon empirical representation of variations of parameters at a reference altitude, assuming diffuse equilibrium for each component. Because this field of research is too large, this study deals mainly with the comparison between the experimental models. This comparison generally gives satisfactory agreement. However significant disagreements also clearly appear and typical examples are given. Several reasons can be put forward on the origin of such disagreements: indicators of sinks and sources of energy, distribution of data with respect to space-time and geophysical parameters, knowledge of some parameters such as the eddy diffusion coefficient, knowledge of interactions with other zones of the atmosphere (e.g. mesosphere, ionosphere, magnetosphere). This list is not exhaustive. Of course, new programs are developed (MAP, EISCAT, UARS, SPACELAB experiments), but this will take time to bear fruit and to gather data. Therefore, all existing data constitute a priceless bank of data over two solar cycles, even if some have been only partly recorded and prevent us from making a perfect model. On the other hand, a model can be very useful without being perfect. It has to be a clear conventional reference which gives a satisfactory mean of the behaviour of the thermosphere. It must be easy to use and also be a practical tool to test data or be compared with theoretical models to be developed. Finally, it has to permit one to solve partly some physical problems in assuming some parameters as known. Our opinion is that a new international experimental reference model based upon simplified physical concepts and upon new measured parameters should be very useful if an easily-accessible data bank is created at the same time.