An accurate estimation of precipitation intensity can only be achieved if the particle type of precipitation is identified. In this study, we describe the construction of a system based on neural networks that classifies precipitation particles by type. The system uses Doppler spectra obtained by a vertical pointing Doppler radar as input data and the type of precipitated particle, based on ground observations, as a teacher signal. The reliability of the classification system was evaluated by the system's F‐measure after a period time. We show that the F‐measure of the classification system can be 0.915, and we confirm that the system classifies precipitation particle types satisfactorily until 0.9–49.1 min after estimation.
Reconstruction of reliable plasmaspheric electron density profiles is important for understanding physical processes in the plasmasphere. This paper proposes a technique that can be applied to correct the plasmaspheric electron density profiles using ray tracing by scrutinizing dispersion analyses of lightning whistlers. The Global Core Plasma Model and the International Reference Ionosphere were introduced as a reference electron density profile. Modifications of this electron density profile were then proposed to satisfy the dispersion characteristics of lightning whistlers measured by satellites in the plasmasphere. We first introduce two kinds of functions to modify the electron density: constant and linear, the linear function is more adequate. We applied our method to two lightning whistler events on 14 August 2017, measured by the Plasma Wave Experiment/Waveform Capture aboard the Arase satellite, and analyzed the dispersion of the observed lightning whistlers. We show how the density modification affects the delay time of the ray path and satisfies the dispersion characteristics under the appropriate adjustments.
Plasma wave direction determined by propagation characteristics and remote sensing technology provides important information for understanding not only the local plasma environment but also the global features of space plasma. The wave distribution function method, a direction finding method, derives the directional distribution of wave energy density using a priori information such as the propagation mode, plasma density, and geomagnetic field intensity. The Markov random field model includes a noise integration kernel that corresponds to the noise used for the estimation, which improves the robustness of the model. However, the effectiveness of this noise integration kernel was only verified for the case where the electromagnetic field sensor noise levels were equal. The noise levels of the electromagnetic field sensors on board scientific satellites often change due to the degradation of the sensors during long‐term instrument operation. Therefore, we proposed two design methods for a noise integration kernel to improve the estimation accuracy when the noise levels among electromagnetic field sensors are different. In proposed model 1, the noise integration kernel was designed for each sensor, and in proposed model 2, the noise integration kernel was designed with the information that the noise level ratio between the sensors was known. The simulation results showed that proposed model 2 improved the misalignment of the peak in the direction of arrival when the noise levels were different regardless of whether the sensor was parallel or perpendicular to the external magnetic field.
The characteristics of plasma waves propagating in the Earth's plasmasphere and magnetosphere are important in terms of understanding the plasma environment in which these waves are excited and propagate. The propagation direction is a critical in propagation analysis of ELF/VLF band waves. The Means, eigenvector, amplitude, and SVD methods have been proposed previously to estimate the direction of arrival of a single wave, and the wave distribution function (WDF) method, which solves the ill‐posed problem, has been proposed to estimate the propagation direction of multiple arriving waves. In this paper, we propose a stochastically extended WDF method. Conventional WDF methods are based on maximum posterior probability estimation; however, the proposed method uses a Markov chain Monte Carlo method. In the proposed method, the estimated solution is obtained based on the sampling approximation by obtaining multiple samples following the posterior distribution. In addition, the WDF is discretized and quantized to improve calculation stability and efficiency. From the sampled WDF, the mean value of the WDF and 95% credible interval can be calculated without arbitrary parameters. The proposed method is more objective than conventional methods and can be used effectively as a basis for performance evaluation of the WDF method.
Random waves can be described as the sum of numerous plane waves, and stochastic processes describe their properties. Various methods have been used to widely investigate the propagation characteristics of electromagnetic waves in magnetized cold plasma based on a single‐plane wave approximation. On the other hand, the properties of random waves are difficult to analyze using these methods. Instead, a framework of the wave distribution function (WDF) should be employed. In this study, we provide an explicit expression for an integration kernel in a magnetized cold plasma used for the WDF method. We show that the kernel can be approximated in the case of ultralow frequency/very low frequency (ULF/VLF) parts of whistler‐mode waves with quasi‐parallel propagation. We also propose a method for estimating a WDF representing a directional distribution of the wave energy density based on the principle of maximum entropy using three‐component spectral matrix data of the magnetic field. Based on the insights obtained from the proposed method, we define a quantity called “sharpness,” which provides spreading of wave normal angles. The sharpness is particularly effective in showing the spread of the wave normal angles for random non‐single‐plane waves. Compared with the conventional methods which evaluate the propagation properties (such as planarity), the “sharpness” exhibited a low calculation load and can be implemented easily for onboard processing.
This study proposes a wave distribution function (WDF) estimation method based on minimizing the Bregman divergence in the function space. The proposed method is entirely data‐driven and requires no hyperparameter tuning during the estimation process. It can be regarded as an extension of the maximum entropy method. Whitening the spectral matrix was found to significantly enhance the estimation accuracy. By replacing the likelihood or loss function of the spectral matrix with its whitened counterpart, the estimated solutions were significantly improved. This method enables the precise estimation of WDFs from electromagnetic field observations, establishing a technical foundation for diagnosing space plasma environments and forecasting space weather. To address the uncertainty arising from the limited snapshots, we derive the posterior distribution of the spectral matrix. The Bregman divergence‐based framework ensures stable and interpretable estimation under both physical and statistical constraints. Numerical experiments confirmed that whitening markedly improved both accuracy and robustness in realistic random wave scenarios.