We present a new quantitative theory for the rate of change of total electron content index (ROTI) by noting its straightforward relationship to the phase structure function of ionospheric turbulence. The theory provides the dependence of ROTI on the sampling interval, satellite motion, propagation geometry, and the spectral shape, strength, anisotropy, and drift of the irregularities. We also present useful approximations to the full theory that elucidate the principal dependencies. We show, for example, that ROTI varies with the effective scan velocity (Veff) to the power ν − 1/2, where 2ν + 1 is the spectral index of the 3‐D irregularities. This dependence on Veff persists in the ratio of ROTI to the intensity scintillation index (S4), which thus depends on the particular viewing geometry for each satellite. While irregularity strength cancels when forming this ratio, the dependence on Fresnel scale remains. We validate the theory by comparing predictions of irregularity strength (CkL) and S4 from 1‐Hz ROTI measurements with calculations derived from 20‐Hz intensity samples collected at Ascension Island. The theory explains the observations well, except when the scan is directed nearly along the field‐aligned irregularities (within 20° of meridional). For cross‐field scans the standard deviations of the error in predicting S4 from 1‐Hz ROTI were 0.05, 0.11, and 0.14 for weak, moderate, and strong scintillation, respectively. These encouraging results suggest the possibility of using dense networks of inexpensive total electron content monitors to provide quantitative diagnostics of ionospheric scintillation and the irregularities that cause them.
We extend the power law phase screen theory for ionospheric scintillation to account for the case where the refractive index irregularities follow a two‐component inverse power law spectrum. The two‐component model includes, as special cases, an unmodified power law and a modified power law with spectral break that may assume the role of an outer scale, intermediate break scale, or inner scale. As such, it provides a framework for investigating the effects of a spectral break on the scintillation statistics. Using this spectral model, we solve the fourth moment equation governing intensity variations following propagation through two‐dimensional field‐aligned irregularities in the ionosphere. A specific normalization is invoked that exploits self‐similar properties of the structure to achieve a universal scaling, such that different combinations of perturbation strength, propagation distance, and frequency produce the same results. The numerical algorithm is validated using new theoretical predictions for the behavior of the scintillation index and intensity correlation length under strong scatter conditions. A series of numerical experiments are conducted to investigate the morphologies of the intensity spectrum, scintillation index, and intensity correlation length as functions of the spectral indices and strength of scatter; retrieve phase screen parameters from intensity scintillation observations; explore the relative contributions to the scintillation due to large‐ and small‐scale ionospheric structures; and quantify the conditions under which a general spectral break will influence the scintillation statistics.
Previous authors have reported on the morphology of GPS scintillations and irregularity zonal drift during the 2002 Conjugate Point Equatorial Experiment (COPEX) in Brazil. In this paper, we characterize the turbulent ionospheric medium that produced these scintillations. Using 10 Hz GPS carrier-to-noise measurements at Boa Vista (2.9°N, 60.7°W), Alta Floresta (9.9°S, 56.1°W), and Campo Grande (20.5°S, 54.7°W), we report on the variation of turbulent intensity, phase spectral index, and irregularity zonal drift as a function of latitude and local time for the evening of 1-2 November 2002. The method of analysis is new and, unlike analytical theories of scintillation based on the Born or Rytov approximations, it is valid when the scintillation index saturates due to multiple-scatter effects. Our principal findings are that (1) the strength of turbulence tended to be largest near the crests of the equatorial anomaly and at early postsunset local times, (2) the turbulent intensity was generally stronger and lasted two hours longer at Campo Grande than at Boa Vista, (3) the phase spectral index was similar at the three stations but increased from 2.5 to 4.5 with local time, and (4) our estimates of zonal irregularity drift are consistent with those provided by the spaced-receiver technique.