Bayesian Filtering in Incoherent Scatter Plasma Parameter Fits

Abstract

Incoherent scatter (IS) radars are high-power, large-aperture radars that detect radio wave scattering from thermal fluctuations in the ionospheric plasma. Power spectral density of the scattered signal is a function of number density, temperature, bulk velocity, and ion-neutral collision frequency of a number of ion species and electrons (for example Swartz & Farley, 1979, and references therein). All these parameters cannot be fitted to the spectrum, and a commonly used approximation is the four-parameter fit of N e , T e , T i and V i. Equal temperatures and bulk velocities are assumed for all ion species, and the ion-neutral collision frequency and ion composition are taken from ionospheric models. In the F 1 region the four-parameter fit often produces incorrect temperatures (for example Blelly et al., 2010), because ion composition models are unreliable in the transition region from the E region molecular NO + and  2 O ions to the F 2 region atomic O +. Incorrect compositions bias the temperatures, because the IS spectrum is sensitive to the ratio T i /m i , where m i is the mean ion mass. This is known as the "temperature-ion composition ambiguity" (TICA; Martínez-Ledesma et al., 2019). Several authors have addressed the TICA problem by means of modeling the F 1 region temperature and ion composition profiles (

Precipitating Electron Energy Spectra and Auroral Power Estimation by Incoherent Scatter Radar With High Temporal Resolution

“…The BAFIM‐ELSPEC analysis method is applied to an auroral event containing three small substorms that occur in the pre‐midnight and post‐midnight sectors on 9 March 2016. The four‐parameter fits of N e , T e , T i , and V i to the E region EISCAT UHF ISR data were performed with 4 s/1.8 km resolutions by using the BAFIM (Virtanen et al., 2021). We find that N r is systematically smaller than N e in the E region when electron precipitation heats the electron gas above the ion temperature.…”

“…BAFIM was tuned so that the "effective" time and range resolutions of N e are very close to the time and range steps, while resolutions of the other plasma parameters are effectively coarser. Interested readers are referred to Table 1 of Virtanen et al (2021) for the values of the tuned analysis parameters and their physical meanings. For this particular study, however, we changed the electron density correlation length 𝐴𝐴 ( 𝑠𝑠 ℎ ) and process noise 𝐴𝐴 (𝑠𝑠 𝑡𝑡 ) scaling parameters to 0.1 and 1.0 ⋅ 10 12 m −3 s −1/2 , respectively.…”

“…The full‐profile analysis allows one to include prior information about the plasma parameter profiles, but it is also computationally heavier than the gated analysis. The BAFIM (Virtanen et al., 2021) is an extension module to GUISDAP, which allows one to include prior information about plasma parameter gradients in both range and time in the gated GUISDAP analysis. BAFIM thus extends the idea of full profile analysis to smoothness in both range and time, but without increasing the computational burden of the gated analysis.…”

“…The analysis method we use to calculate the differential electron energy flux from the EISCAT UHF ISR data consists of two steps. First, plasma parameters are fitted to the incoherent scatter data with high time and range resolutions using the combination of GUISDAP (Lehtinen & Huuskonen, 1996) and BAFIM (Virtanen et al., 2021). Second, the fitted electron density altitude profiles are inverted into differential energy fluxes of precipitating electrons using the ELSPEC software (Virtanen et al., 2018).…”

A Log-Gaussian Cox Process with Sequential Monte Carlo for Line Narrowing in Spectroscopy

“…It has the potential to be used for practically any spectroscopic measurement of electromagnetic phenomena. For example, ionospheric spectra measured by incoherent scatter radar [27], X-ray spectra used for industrial quality control [25], or hyper-spectral radiance measurements in satellite remote sensing [7]. Collection of these types of data is very time-consuming and expensive.…”