Inversion techniques for analyzing long pulse incoherent scatter radar data with a focus on F-region ion drift estimation

Incoherent scatter radar (ISR) probing of the topside F-region ionosphere with reduced electron densities residing at relatively large radar ranges requires the use of uncoded (i.e., narrow bandwidth) pulse transmissions of relatively long durations (typically of 0.5 ms or longer). In such uncoded long pulse (ULP) radar measurements, the range resolution is necessarily poor due to range-mixed scattered signals arriving from a large volume of the ionosphere illuminated uniformly by the transmitted pulse, rendering an accurate and range-specific inference of ionospheric state parameters such as composition, temperatures, and drift velocities of ionospheric charged species a very challenging data inversion problem. This thesis develops and applies inversion techniques to address this challenge in connection with long pulse ISR data collected at the Arecibo Observatory with the specific objective of obtaining accurate estimates of different drift velocities of ionic species residing in regions of the ionosphere with comparable concentrations — this is achieved using forward models and error covariance tensors derived and developed to invert the scattered signal autocorrelation function (ACF) and phase: In the first stage, low-computational-cost data inversion techniques were identified to efficiently retrieve height-localized estimates of species densities and temperatures, as well as ion drift velocities. Using the low-cost parameter estimates as initial guesses, a secondary high-resolution forward model of periodogram-averaged Doppler spectrograms was then developed for multi-height estimation of ionospheric density and temperature profiles with 30 km height sampling. These high-resolution full-profile inversion results were subsequently utilized in improving the drift velocity estimates of heavy and light ions co-existing on the topside F-region ionosphere using an ACF phase fitting method applied with a forward model of a hybrid phase estimator and the associated error covariance matrix derived for the first time in this work. The hybrid phase method provides a new pathway for obtaining the full covariance of the phase errors enabling a complete maximum likelihood estimation that improves drift estimates with accurate error analysis.

Name thesis:degree_name

Ph.D.

Level thesis:degree_level

Dissertation

Discipline thesis:degree_discipline

Electrical & Computer Engr

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2024

Author dc:creator

• Wang, Binghui

Contributors dc:contributor

• Kudeki, Erhan
• Jin, Jianming
• Waldrop, Lara
• Ilie, Raluca

• Incoherent Scatter Radar
• Remote Sensing

Statement dc:rights

• Copyright 2024 Binghui Wang

Language dc:language

eng, en

Handle dc:identifier

https://hdl.handle.net/2142/127276

OAI record oai:identifier

https://www.ideals.illinois.edu/oai-pmh?verb=GetRecord&metadataPrefix=etdms&identifier=oai%3Awww.ideals.illinois.edu%3A2142%2F127276