Characterising Spatial and Temporal Ionospheric Variability with a Network of Oblique Angle-of-arrival and Doppler Ionosondes

Abstract

Ionospheric variability exists on a broad range of scales, and routinely impacts skywave propagation modes of high frequency radio waves, to the detriment of radar and communication systems. In order to better understand the electron density structures associated with such variability at mid-latitudes, a network of oblique angle-of-arrival (AoA) and Doppler ionosondes were installed in central and northern Australia as part of the ELOISE campaign in 2015. This thesis analyses observations from the ELOISE AoA ionosondes, with a focus on characterising the influence of medium- to large- scale gradients and signatures of travelling ionospheric disturbances (TIDs). Following an overview of the experiment, the design and calibration of the new ionosonde system is described. With multi-channel receivers connected to each element of two twin-arm arrays, a total of eleven AoA paths of between 900 and 2700 km were collected, including nine with interleaved Doppler measurements using a special channel scattering function (CSF) capability. On-board signal processing was developed to perform real-time clear channel evaluation and CSF scheduling, and generate the AoA ionograms and delay-Doppler images with fitted electron density profiles. In further offline analysis, peak detection and mode classification was carried out, to support reflection point mapping and tilt estimation. Significant testing and validation of the new ionosonde before and after the experiment revealed AoA uncertainties on the scale of 0.2–0.5° in bearing and 0.4–0.9° in elevation. Having identified a low-elevation bias, models of tropospheric refraction and antenna mutual coupling effects were considered as possible correction strategies, but ultimately an empirical approach based on aggregated ionospheric returns was implemented. Small-scale (intra-dwell) ionospheric variability also has the potential to compromise results, through unresolved multi-mode mixing, and this has been investigated using a combination of spatial and temporal variability metrics derived from the CSF data. The analysis of large quantities of F2 peak data shows persistent diurnal patterns in the oblique AoA observables that are also well-captured by a conventional data-assimilative ionospheric model, even without the benefit of AoA and Doppler inputs. Furthermore, Doppler measurements are reproduced remarkably well using just the midpoint fitted profiles. A statistical study has quantified the level of consistency between observations and model, to provide greater confidence in the results. Many of the geophysical features can be interpreted as ionospheric gradients, as evident in the tilt estimates, and horizontally moving structures such as TIDs, using a form of Doppler-based drift analysis. While signatures of TIDs vary considerably, two simple wave-like perturbation models have been evaluated to help classify quasi-periodic behaviour in the AoA observations, as well as understand the directional filtering effect imposed by the path geometry. In some cases, a set of TID parameters can be determined by eye, but in others automatic parameter inversion techniques may be more viable. Two such techniques were implemented but results using both real and synthetic data demonstrated some significant limitations. Finally, attempts to relate TID signatures across multiple paths shows promise, but there still appears to be a strong dependence on path geometry that is difficult to eliminate.