Ionosphere Data Assimilation: Problems Associated with Missing Physics
Physics-based data assimilation models can be used for a wide range of applications in space physics, but as with all models the data assimilation models have limitations. The limitations can be associated with the data, the assimilation technique, or the background physics-based model. Here, we focused on the ionosphere and on elucidating the problems associated with missing physics in the background ionosphere model. The study was conducted with the Global Assimilation of Ionospheric Measurements-Gauss-Markov (GAIM-GM) physics-based data assimilation model. Simulations relevant to the low and middle latitude ionosphere were conducted in order to show how missing physics in the background ionosphere model affects the reconstructions. The low-latitude simulation involved the presence of equatorial plasma bubbles and a background physics-based ionosphere model that does not self-consistently describe bubbles. This problem, coupled with insufficient data, led to a Gauss-Markov reconstruction that contained a broad region of relatively low nighttime Total Electron Density (TEC) values instead of the individual plasma bubbles. The implications of plasma bubbles for reconstructions with the GAIM-FP (Full Physics) data assimilation model, where the electric fields and neutral winds are determined self-consistently, are noted. The mid-latitude simulation involved a Gauss-Markov ionosphere reconstruction for a geomagnetic storm where a Storm Enhanced Density (SED) appeared across the United States. Again, the background physics-based model (Ionosphere Forecast Model) could not produce the SED feature because this model does not take account of penetrating electric fields. Nevertheless, in this case there were sufficient data to overcome the deficiency in the background ionosphere model and the Gauss-Markov data assimilation reconstruction successfully described the SED feature and surrounding ionosphere.
KeywordsGeomagnetic Storm Neutral Wind Ionosphere Model Ensemble Kalman Filter Plasma Bubble
This research was supported by Office of Naval Research grant N00014-09-1-0292 to Utah State University.
- McDonald SE, Su Basu, Basu S, Groves KM, Valladares CE, Scherliess L, Thompson DC, Schunk RW, Sojka JJ, Zhu L (2006) Extreme longitudinal variability of plasma structuring in the equatorial ionosphere on a magnetically quiet equinoctial day. Radio Sci 41:RS6S24. doi: 10.1029/2005RS003366CrossRefGoogle Scholar
- Scherliess L, Thompson DC, Schunk RW (2010) Data assimilation models: a ‘new’ tool for ionospheric science and applications. In: Liu W, Fujimoto M (eds) The dynamic magnetosphere, IAGA. Springer, NY (in press)Google Scholar
- Schunk RW, Sojka JJ, Eccles JV (1997) Expanded capabilities for the ionospheric forecast model, Final Report, AFRL-VS-HA-TR-98-0001, 1–142Google Scholar
- Schunk RW et al (2005a) An operational data assimilation model of the global ionosphere. In: Goodman JM (ed) Proceedings of 2005 ionospheric effects symposium. JMG Associates Ltd, Alexandria, VA, pp 512–518Google Scholar