No abstract

Abstract. An important component of ionospheric plasma irregularity studies in the Indian low latitudes involves the study of the plasma bubbles which produce intense scintillations of the transionospheric satellite signals. Many such plasma bubble induced (PBI) scintillation events were identified while recording 244 MHz signal from the geostationary satellite Fleetsat (73°E) at Delhi (28.6°N, 77.2°E) during March-April 1991. This type of scintillations represents changes in plasma processes. These scintillations are spectrally analyzed using an autoregressive (AR) scheme, which is equivalent to maximum entropy method of spectrum analysis, amenable to extracting optimum spectral content from short data lengths (20 -40 s). Each spectrum is assigned a level of detectability using the final prediction error (FPE) derived from the optimum filter order required to resolve the spectrum. Lower detectability together with a higher order filter indicate a higher level of coherence for the plasma irregularities (discrete structures). Consistent patterns for these scintillations emerge from the present analysis as follows: (1) the initial and final phases of a scintillation patch display quasiperiodic oscillations. Their corresponding spectra show dominant (Gaussian shaped) spectral features with detectability levels of )6 dB to )12 dB and requiring a higher order (>6) AR filter for their spectral resolution. These are most likely associated with discrete ''filament-like'' or ''sheet-like'' plasma structures that exist near the bubble walls. (2) Two main features of the scintillation spectra could be positively associated with the welldeveloped plasma bubble stage: (a) spectra displaying a power-law process with a single component spectral slope between 1.6 to 3.0. Generally such spectra are resolved with a 2nd order filter and have a 1 dB to 6 dB of detectability. (b) Spectra displaying a double slope, indicating an inner and an outer scale regime for the power-law irregularities. These spectra are resolved with higher order filters (>3 but <7) and possess detectability levels of )1 dB to 3 dB. These spectra display finer spectral changes, perhaps indicative of the nature of continuously evolving plasma irregularities. As an example, an analysis of a single scintillation patch is presented to highlight the geophysical significance of the present approach. Some important parameters used in the AR scheme of spectral analysis are given in the Appendix.

Abstract. An important component of ionospheric plasma irregularity studies in the Indian low latitudes involves the study of the plasma bubbles which produce intense scintillations of the transionospheric satellite signals. Many such plasma bubble induced (PBI) scintillation events were identified while recording 244 MHz signal from the geostationary satellite Fleetsat (73°E) at Delhi (28.6°N, 77.2°E) during March-April 1991. This type of scintillations represents changes in plasma processes. These scintillations are spectrally analyzed using an autoregressive (AR) scheme, which is equivalent to maximum entropy method of spectrum analysis, amenable to extracting optimum spectral content from short data lengths (20 -40 s). Each spectrum is assigned a level of detectability using the final prediction error (FPE) derived from the optimum filter order required to resolve the spectrum. Lower detectability together with a higher order filter indicate a higher level of coherence for the plasma irregularities (discrete structures). Consistent patterns for these scintillations emerge from the present analysis as follows: (1) the initial and final phases of a scintillation patch display quasiperiodic oscillations. Their corresponding spectra show dominant (Gaussian shaped) spectral features with detectability levels of )6 dB to )12 dB and requiring a higher order (>6) AR filter for their spectral resolution. These are most likely associated with discrete ''filament-like'' or ''sheet-like'' plasma structures that exist near the bubble walls. (2) Two main features of the scintillation spectra could be positively associated with the welldeveloped plasma bubble stage: (a) spectra displaying a power-law process with a single component spectral slope between 1.6 to 3.0. Generally such spectra are resolved with a 2nd order filter and have a 1 dB to 6 dB of detectability. (b) Spectra displaying a double slope, indicating an inner and an outer scale regime for the power-law irregularities. These spectra are resolved with higher order filters (>3 but <7) and possess detectability levels of )1 dB to 3 dB. These spectra display finer spectral changes, perhaps indicative of the nature of continuously evolving plasma irregularities. As an example, an analysis of a single scintillation patch is presented to highlight the geophysical significance of the present approach. Some important parameters used in the AR scheme of spectral analysis are given in the Appendix.