Fundmental Challenges in Space Weather of the Ionosphere Thermosphere System

In this paper we discuss some of the fundamental questions that underlie space weather in the ionosphere and thermosphere and how we can chart a path forward that combines, theory, first principles models, assimilative models, space experiments, and groundbased observations. We will suggest approaches to improve our understanding of the system. Certainly, the community has made significant progress since the dawn of the Space Age but there are fundamental assumptions that need to be revisited. When we speak of the scientific questions we should recognize that an important challenge for the scientific community is to convey to the users the limitations of our knowledge for the particular application the user has in mind. First, we should define what we mean by “space weather”. In some cases, it is taken to mean the effects of the space environment on human systems, with special emphasis on those effects that arise from the extrema in the environment. Many space weather users are only concerned with this aspect of the problem. In the scientific community, we consider space weather as the variability of the space environment (with a focus on the ionosphere and thermosphere or “i/T”) about the average conditions. The average conditions are assumed to be represented by climatological model such as MSIS, for the thermosphere, or IRI for the ionosphere. We note that the tacit assumption is that we have a meaningful measure of the variance inherent in the climatological model so that we can judge when “space weather” is occurring. The impact of this for the science community is that many fail to appreciate whether the variability of the system or its cause is understood or if it is predictable. This is also true for the user community. The I/T system is driven by external forcing (the coupling to the Sun and the magnetosphere) and the lower atmosphere (an impact that has only become fully realized in the last decade or so). The I/T system is also sensitive to initial conditions (i.e. the effects on previous activity). However, the parameters in the climatological models reflect only the most simplistic representation of these drivers (i.e. a solar flux index). First principles models are often “tuned” to reproduce these climatological values then they are assumed to capture the physics of the system while embodying many parameterizations of sub-gridscale processes and the external drivers. Note that the models require that they linearize the equations they solve. Spatial resolution is constrained by the mechanics of the solution. Note that “improving” the resolution of the models may not actually improve their ability to reproduce past observations. It certainly may have little if anything to do with the models’ ability to forecast the future state of the system. An example of this is the problem of ionospheric irregularities. We may be able to predict in a climatological sense when and where they are like to occur but first principles models cannot predict them. In fact, first principles models do not enable us to discover the seed mechanism for the Rayleigh-Taylor Instability that produces ionospheric bubbles. We will discuss how the use of assimilative data models and various kinds of data can enable a better understanding of the conditions at any one time. There are, however, limitations; the assimilative models have a very limited ability to predict the future state. This is a challenge that demonstrates the value of continued investment in first principles models and in understanding the limits of the predictability of chaotic systems.