An official website of the United States government
Abstract Data from a network of high‐frequency (HF) beacons deployed in Peru are used to estimate the regional ionospheric electron density in a volume. Pseudorange, accumulated carrier phase, and signal power measurements for each of the 36 ray paths provided by the network at a 1 min cadence are incorporated in the estimates. Additional data from the Jicamarca incoherent scatter radar, the Jicamarca sounder, and GPS receivers can also be incorporated. The electron density model is estimated as the solution to a global optimization problem that uses ray tracing in the forward model. The electron density is parametrized in terms of B‐splines in the horizontal direction and generalized Chapman functions or related functions in the vertical. Variational sensitivity analysis has been added to the method to allow for the utilization of the signal power observable which gives additional information about the morphology of the bottomside F region as well as absorption including absorption in the D and E regions. The goal of the effort is to provide contextual information for improving numerical forecasts of plasma interchange instabilities in the postsunset F region ionosphere associated with equatorial spread F (ESF). Data from two ESF campaigns are presented. In one experiment, the HF data revealed the presence of a large‐scale bottomside deformation that seems to have led to instability under otherwise inauspicious conditions. In another experiment, gradual variations in HF signal power were found to be related to the varying shape of the bottomside F layer.
more »
« less
This content will become publicly available on December 1, 2025
Ionospheric Radio Beacon Signal Analysis and Parameter Estimation Using Automatic Differentiation
Abstract Continuous wave signals from a network of high frequency (HF) beacons in Peru and other instruments are used to reconstruct the regional ionospheric electron number density in the volume surrounding the network. The continuous wave (CW) HF signals employ binary phase codes with pseudorandom noise (PRN) encoding, and the observables include propagation time or pseudorange, Doppler shift or beat carrier phase, and amplitude. A forward model based on geometric optics in an inhomogeneous, anisotropic, lossy plasma is used to relate plasma number density to the observables. Plasma number density is parametrized in terms of a modified Chapman profile in the vertical and biquintic B‐splines in the horizontal. Sensitivity analysis is required both to model the ray amplitudes and to solve the two‐point boundary problem for each ray. Sensitivity analysis is performed here using reverse‐mode automatic differentiation. In particular, we use an LLVM compiler (Clang), the corresponding OpenMP library, and the Enzyme Automatic Differentiation Framework plugin to compute the sensitivity (gradients) of ray endpoints with respect to their initial bearings. The resulting algorithm exhibits no performance penalty compared to variational sensitivity analysis and is far simpler to implement.
more »
« less
- Award ID(s):
- 2213849
- PAR ID:
- 10588967
- Publisher / Repository:
- Wiley
- Date Published:
- Journal Name:
- Journal of Geophysical Research: Machine Learning and Computation
- Volume:
- 1
- Issue:
- 4
- ISSN:
- 2993-5210
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Abstract Super Dual Auroral Radar Network (SuperDARN) radars operate in a coordinated but monostatic configuration whereby high‐frequency (HF) signals scattered from ionospheric density irregularities or from the surface of the Earth return to the transmitting radar where Doppler parameters are then acquired. A bistatic arrangement has been developed for SuperDARN radars in which HF signals transmitted from one radar are received and analyzed by another radar that is separated by a large distance (>1,000 km). This new capability was developed and tested on radars located in Oregon and Kansas. Numerous 3‐day bistatic campaigns have been conducted over a period extending from September 2019 through March 2020. During these campaigns three distinct bistatic propagation modes have been identified including a direct mode in which signals are transmitted and received through the radar side lobes. Direct mode signals propagate along the great‐circle arc connecting the two bistatic radar sites, reflecting from the ionosphere at bothEregion andFregion altitudes. Two additional modes are observed in which HF signals transmitted from one radar scatter from either ionospheric density irregularities or from the surface of the Earth before propagating to the bistatic receiving radar. Ray tracing simulations performed for examples of each mode show good agreement with observations and confirm our understanding of these three bistatic propagation modes. Bistatic campaigns continue to be scheduled in order to improve technical aspects of this new capability, to further explore the physical processes involved in the propagation and scattering of HF bistatic signals and to expand the coverage of ionospheric effects that is possible with SuperDARN.more » « less
-
SUMMARY The sensitivity of Rayleigh wave amplitude to Earth structure has applications to seismic tomography, both in cases where amplitude information is used to supplement phase velocity data to improve images of elastic parameters, and to correct amplitudes for local Earth structure in attenuation tomography. We review the theoretical basis of the ray theoretical approximation, in which the wave amplitudes are controlled by a combination of geometrical spreading and local changes in energy density due to Earth structure. We focus mainly on the latter effect, which we term the constant energy flux approximation. We investigate the ray theoretical basis for this approximation, test it against a full waveform simulation that verifies its accuracy and show how it can be used to compute the sensitivity of amplitude to elastic moduli and density. We investigate how perturbing these parameters in a set of simple Earth models affects Rayleigh wave amplitudes, and demonstrate that a slow velocity heterogeneity can cause either increased or reduced amplitudes, depending upon the depth of the heterogeneity and the observation frequency. Consequently, amplitude sensitivity can be either positive or negative, and its magnitude can vary significantly with frequency. Although an added complication, the very different behaviour of phase velocity and amplitudes to changes in Earth structure implies that the two types of data are complementary and suggest the effectiveness of using both in Rayleigh wave tomography.more » « less
-
Abstract Trans‐ionospheric high frequency (HF: 3–30 MHz) signals experience strong attenuation following a solar flare‐driven sudden ionospheric disturbance (SID). Solar flare‐driven HF absorption, referred to as short‐wave fadeout, is a well‐known impact of SIDs, but the initial Doppler frequency shift phenomena, also known as “Doppler flash” in the traveling radio wave is not well understood. This study seeks to advance our understanding of the initial impacts of solar flare‐driven SID using a physics‐based whole atmosphere model for a specific solar flare event. First, we demonstrate that the Doppler flash phenomenon observed by Super Dual Auroral Radar Network (SuperDARN) radars can be successfully reproduced using first‐principles based modeling. The output from the simulation is validated against SuperDARN line‐of‐sight Doppler velocity measurements. We then examine which region of the ionosphere, D, E, or F, makes the largest contribution to the Doppler flash. We also consider the relative contribution of change in refractive index through the ionospheric layers versus lowered reflection height. We find: (a) the model is able to reproduce radar observations with an root‐median‐squared‐error and a mean percentage error (δ) of 3.72 m/s and 0.67%, respectively; (b) the F‐region is the most significant contributor to the total Doppler flash (∼48%), 30% of which is contributed by the change in F‐region's refractive index, while the other ∼18% is due to change in ray reflection height. Our analysis shows lowering of the F‐region's ray reflection point is a secondary driver compared to the change in refractive index.more » « less
-
Abstract The Super Dual Auroral Radar Network (SuperDARN) is a network of High Frequency (HF) radars that are typically used for monitoring plasma convection in the Earth's ionosphere. A majority of SuperDARN backscatter can broadly be divided into three categories: (a) ionospheric scatter due to reflections from plasma irregularities in the E and F regions of the ionosphere, (b) ground scatter caused by reflections from the ground/sea surface following reflection in the ionosphere, and (c) backscatter from meteor trails left by meteoroids as they enter the Earth's atmosphere. Due to the complex nature of HF propagation and mid‐latitude electrodynamics, it is often not straightforward to distinguish between different modes of backscatter observed by SuperDARN. In this study, we present a new two‐stage machine learning algorithm for identifying different backscatter modes in SuperDARN data. In the first stage, a neural network that “mimics” ray‐tracing is used to predict the probability of ionospheric and ground scatter occurring at a given location along with parameters like the elevation angles, reflection heights etc. The inputs to the network include parameters that control HF propagation, such as signal frequency, season, UT time, and geomagnetic activity levels. In the second stage, the output probabilities from the neural network and actual SuperDARN data are clustered together to determine the category of the backscatter. Our model can distinguish between meteor scatter, 1/2 hop E‐/F‐region ionospheric as well as ground/sea scatter. We validate our model by comparing predicted elevation angles with those measured at a SuperDARN radar.more » « less
