- Ionosphere and magnetosphere dynamics
- Solar and Space Plasma Dynamics
- Meteorological Phenomena and Simulations
- Geomagnetism and Paleomagnetism Studies
- Earthquake Detection and Analysis
- Climate variability and models
- Magnetic confinement fusion research
- Oceanographic and Atmospheric Processes
- Tropical and Extratropical Cyclones Research
- Geophysics and Gravity Measurements
- Ocean Waves and Remote Sensing
- GNSS positioning and interference
- Fault Detection and Control Systems
- Target Tracking and Data Fusion in Sensor Networks
- Geophysical and Geoelectrical Methods
- Fusion materials and technologies
- Statistical Methods and Bayesian Inference
- Advanced Statistical Methods and Models
- Image and Signal Denoising Methods
- Cyclone Separators and Fluid Dynamics
- Superconducting Materials and Applications
- Fluid Dynamics and Turbulent Flows
- Aerosol Filtration and Electrostatic Precipitation
- Inertial Sensor and Navigation
- Hydrological Forecasting Using AI
The Institute of Statistical Mathematics
2013-2024
Research Organization of Information and Systems
2005-2024
The Graduate University for Advanced Studies, SOKENDAI
2016-2023
Japan Science and Technology Agency
2007-2016
Centre de Recherche en Économie et Statistique
2006
Centre for Research in Engineering Surface Technology
2006
Kyoto University
1999-2002
Grantmakers for Effective Organizations
1998
Abstract. A new filtering technique for sequential data assimilation, the merging particle filter (MPF), is proposed. The MPF devised to avoid degeneration problem, which inevitable in (PF), without prohibitive computational cost. In addition, it applicable cases a nonlinear relationship exists between state and observed where application of ensemble Kalman (EnKF) not effectual. MPF, procedure performed based on sampling forecast as PF. However, unlike PF, each member filtered generated by...
The source regions of region‐0 (R0), region‐1 (R1), and region‐2 (R2) field‐aligned currents (FACs) were statistically determined using DMSP particle precipitation magnetometer data. Each FAC sheet originates from more than one region in the magnetosphere, depending on latitude magnetic local time (MLT). R2 mostly central plasma (CPS) boundary (BPS) morning CPS, BPS, inner magnetosphere afternoon, all which are closed field lines. Near noon, some may originate low‐latitude layer (LLBL), is...
The present study examines seasonal variations of large‐scale field‐aligned current (FAC) systems in terms the dipole tilt and clock angles. Magnetic field measurements from DMSP F7 F12‐F15 satellites are used. This data set consists a total ∼185,000 auroral oval crossings, out which ∼121,000 crossings were selected for analysis. Focus is placed on latitude at demarcation between region 2 (R2) 1 (R1) currents intensities these currents. It found that dayside FAC moves poleward equatorward...
Abstract Magnetic fusion plasmas, which are complex systems comprising numerous interacting elements, have large uncertainties. Therefore, future reactors require prediction-based advanced control with an adaptive system model and estimation robust to uncertainties in the observations. To address this challenge, we introduced a approach based on data assimilation (DA), describes adaptation state probability distribution. The first implementation of DA-based was achieved at Large Helical...
[1] The present study statistically addresses the dependence of large-scale field-aligned currents (R1/R2) and particle precipitation in premidnight sector on solar illumination at ionosphere. energy flux average precipitating electrons ions are examined for downward R2 upward R1 currents. Results summarized as follows: (1) current density is larger dark hemisphere, whereas unclear. (2) For electron higher hemisphere. (3) currents, a similar preference hemisphere found flux, which, however,...
We present statistical observations demonstrating the action of broadband electromagnetic waves or structures in heating magnetosheath ions Earth's and across magnetopause low‐latitude boundary layer. Through use a generalized multicomponent fitting routine to full 3‐D distribution function, we distinguish from magnetospheric diffuse energetic ion populations through these regions. From this approach, find that energy density is well correlated with spectral field fluctuations on gyro‐radius...
We develop a Bayesian technique for estimating the parameters in observation‐noise covariance matrix R t ensemble data assimilation. design posterior distribution by using ensemble‐approximated likelihood and Wishart prior present an iterative algorithm parameter estimation. The temporal smoothness of can be controlled adequate choice two distribution, S number degrees freedom ν . estimated maximizing marginal likelihood. formalism handle cases which points or positions varies with time,...
The present study statistically compares large‐scale field‐aligned currents (FACs) under sunlit and dark ionospheric conditions. A total of ∼74,000 auroral oval crossings are selected from magnetic field measurements the DMSP F7 F12 to F15 satellites. For dayside FAC it is reconfirmed that both current intensity density larger in illuminated events than unilluminated events. As for nightside FAC, contrast, a few important features become clear first time, which can be summarized as follows:...
Abstract. To further our understanding of the solar wind entry across magnetopause under northward IMF, we perform a case study duskside Kelvin-Helmholtz (KH) vortex event on 24 March 1995. We have found that protons consist two separate (cold and hot) components in magnetosphere-like region inside KH vortical structure. The cold proton component occasionally consisted counter-streaming beams near current layer Low-energy bidirectional electron or flat-topped distribution functions direction...
Abstract We propose a method for estimating optimal error covariances in the context of sequential assimilation, including case where both system equation and observation are nonlinear. When is nonlinear, ensemble‐based filtering methods such as ensemble Kalman filter (EnKF) widely used to deal directly with nonlinearity. The present approach covariance optimization maximum likelihood estimation carried out by approximating mean. Specifically, approximated sample mean each member ensemble....
Abstract We present an iterative algorithm for estimating the observation‐noise covariance matrix R t ensemble‐based data assimilation. The proposed enables us to estimate parameters in , number of which may become exceedingly large optimal is estimated based on maximum likelihood and can be described as expectation‐maximization (EM) algorithm. key constructing obtain optimality condition a closed form. method estimates adaptive current observation y . application coupled atmosphere–ocean...
Abstract The solar zenith angle (SZA) dependence of the conductance is studied and a simple theoretical form for Hall‐to‐Pedersen ratio developed, using peak plasma production height. European Incoherent Scatter (EISCAT) radar observations at Tromsø (67 MLAT) on 30 March 2012 were used to calculate conductance. daytime electric associated with created by extreme ultraviolet radiation incident neutral atmosphere Earth. However, it has been uncertain whether previous models are consistent...
Abstract Non-Gaussian probability density functions (PDFs) in convection initiation (CI) and development were investigated using a particle filter with storm-scale numerical prediction model an adaptive observation error estimator (NHM-RPF). An observing system simulation experiment (OSSE) was conducted 90-min assimilation period 1000 particles at 2-km grid spacing. Pseudosurface observations of potential temperature (PT), winds, water vapor (QV), pseudoradar rainwater (QR) the lower...
We have studied the outermost field‐aligned currents (FACs) in plasma sheet boundary layer (PSBL) with Geotail observations. The trajectory and data of enable us to conduct a more comprehensive systematic survey than former work based on ISEE 1 2 data. identified FACs variation dawn‐dusk magnetic field component calculated their intensities assumption geometry. A case study indicated that flow earthward side reconnection site while are directed tailward side. This tendency was confirmed...
Abstract In data assimilation, covariance matrices are introduced in order to prescribe the weights of initial state, model dynamics, and observation, suitable specification covariances is known be essential for obtaining sensible state estimates. The specified by sample converted according an assumed structure. Modelling structure consists regularization a constraint dynamic relationship. Regularization required converting singular into non‐singular covariance, removing spurious correlation...
[1] Variations in the dayside field-aligned current (FAC) density (J//), parallel potential drop (Δϕ//), peak precipitating electron energy (peak Ee), and flux (ɛ) as functions of solar wind (SW) interplanetary magnetic field (IMF) are investigated with Defense Meteorological Satellite Program observations a quasi-stationary low-latitude boundary layer (LLBL)–FAC coupling model. Region 1 (R1) J// responses to variations SW velocity (Vsw) (nsw) at 8–16 local time (MLT) suggest that R1 these...
Abstract. The present study statistically examines the dependence of intensities dayside (MLT=8-12h) large-scale field-aligned currents (FACs) on ionospheric conductance using summary data DMSP-F7 constructed by procedure Higuchi and Ohtani (2000). We have found that, in region, R1 R0 a higher correlation between conductivity FAC intensity than R2, suggesting that are driven more voltage-like source R2. This result is consistent with idea interaction solar wind open magnetospheric magnetic...
Abstract. In search for clues towards the understanding of cold plasma sheet formation under northward IMF, we study temperature anisotropy two-component protons in near dusk low-latitude boundary observed by Geotail spacecraft. The result from mixing component solar wind and hot magnetospheric origin, may be most eloquent evidence transport process across magnetopause. occasionally has a strong flank, sense depends on locations: parallel is enhanced tail flank while perpendicular dayside....
We report an application of the ensemble Kalman filter (EnKF) and smoother (EnKS) to intermediate coupled atmosphere-ocean model Zebiak Cane, into which sea surface height (SSH) anomaly observations by TOPEX/POSEIDON (T/P) altimetry are assimilated. Smoothed estimates 54,403 dimensional state obtained from 1,981 observational points with 2,048 members. While assimilated data SSH anomalies alone, estimated temperature (SST) reproduce primary temporal characteristics actual SST. The smoothed...
The relationship between the Region 2 field‐aligned currents and solar wind dynamic pressure is investigated using magnetic field data from Defense Meteorological Satellite Program‐F13 (DMSP‐F13). generation of associated with direction gradient in magnetosphere, relation to that thermal gradient. Past spacecraft observations have suggested spatial distribution magnetosphere varies pressure. Therefore, we can expect would depend on We compared current intensity at altitude ionosphere, as...
We present a technique for estimating ring current ion distributions and electric potential in the inner magnetosphere by incorporating data from High‐Energy Neutral Atom (HENA) imager on IMAGE satellite into kinetic model context of assimilation. Data assimilation is an approach which adjusts physics‐based according to differences with observations. perform using particle filter (PF) applicable high‐dimensional systems observations relatively low computational cost. In technique,...
Abstract. Through the effort to obtain clues toward understanding of transport cold plasma in near-Earth magnetotail under northward IMF, we find that two-component protons are observed midnight sheet (−10>XGSM>−30 RE, |YGSM| <10 RE) IMF by Geotail spacecraft. Since frequently on duskside during intervals but hardly dawnside, those found thought come from dusk flank. The proton component region occasionally has a parallel anisotropy, which resembles tail flank duskside. flows with...
We have developed a statistical downscaling method for generating probabilistic climate projections using multiple general circulation models (GCMs). A regression model was established so that the combination of coefficients GCMs reflects characteristics variation observations at each grid point. adopted elastic net penalty to estimate model, considering projection similarities. Using an observation system with high spatial resolution, we conducted statistically downscaled 20-km horizontal...
We obtained the distribution of X‐type magnetic neutral lines in magnetotail from X GSM = 0 to −210 R E with Geotail observations. For detection lines, we adopted simultaneous reversals field north south (south north) and ion bulk velocity earthward tailward (tailward earthward). In near‐tail region (0 ≥ > −30 ), occurrence rate reversal events had a peak at Y ≃ +9 was characterized by very asymmetric form direction. The appeared expand direction up 50 width −50 . This maintained further...