 |
Center for Nonlinear Studies |
Langmuir probes are the instrument of choice to determine basic parameters such as the density and temperature in many laboratory and space plasma experiments. This choice is generally dictated by the relative simplicity with which probe characteristics can be interpreted on the basis of analytic expressions based on different approximations such as the Orbital Motion Limited. While fast and convenient for inferring plasma parameters in real time, the models on which these approximate analytic expressions are based, cannot account for several physical effects of importance in actual experimental conditions, such as drifts, magnetic fields, multiple ion species, nearby physical objects capable of deflecting or obstructing incoming particles, photo-electron emission, and more. A better interpretation of sensor measurements could be achieved by replacing approximate analytic expressions with sensor responses obtained from detailed kinetic simulations capable of accounting these many physical effects under realistic conditions of geometry and space environment conditions. Unfortunately such kinetic simulations require considerable computing resources, which renders them inapplicable to real-time interpretation of measurements. A solution to this predicament consists of i) computing sensor responses under a variety of expected space or lab plasma conditions, ii) constructing a solution library, or table of low level (LB1) sensor response corresponding to different plasma conditions, and iii) applying a suitable regression technique to infer plasma parameters from actual measured sensor responses. In this talk I present a possible approach to carrying out this objective with preliminary results obtained with synthetic data. The challenges of constructing solution libraries adapted to a specific experimental setup or satellite, and possible regression strategies are also discussed.
Host: Gian Luca Delzano