Virtual Observatory in the Geospace Environment Studies

Tatsuki Ogino and STEL MembersSolar-Terrestrial Environment Laboratory,

Nagoya Universityand

Group for Creative Research Project

CODATA-20 Beijing, Oct 23-25, 2006

Satell ite anomalies relatedto Space Weather are on therise due to the increased useof composite materialsinstead of metal, and smaller,faster chip designs. The threeprinciple anomaly types are:surface charging, internalcharging, and Single EventEffects (SEE).

Satell ite anomalies relatedto Space Weather are on therise due to the increased useof composite materialsinstead of metal, and smaller,faster chip designs. The threeprinciple anomaly types are:surface charging, internalcharging, and Single EventEffects (SEE).

The Sun’s magnetic field is propagated farbeyond our planetary system by the solar wind.The interact ion between this interplanetarymagnetic field (IMF) and Earth’s ownmagnetic field is a significant component ofSpace Weather.

The Sun’s magnetic field is propagated farbeyond our planetary system by the solar wind.The interact ion between this interplanetarymagnetic field (IMF) and Earth’s ownmagnetic field is a significant component ofSpace Weather.

The Ionosphere is a layer of the Earth’supper atmosphere that contains free electronsand ions produced by solar UV radiation.Disruptions of the ionosphere cansignificantly affect radar, and radiocommunication.

The Ionosphere is a layer of the Earth’supper atmosphere that contains free electronsand ions produced by solar UV radiation.Disruptions of the ionosphere cansignificantly affect radar, and radiocommunication.

The Sun powers both the space weather and theterrestrial weather machines. Solar events, i.e.,explosions of charged particles and the dynamicsof magnetic and electric fields, cause hugechanges in the near-Earth space environment.Satelli tes and communication signals musttraverse this electric space.

The Sun powers both the space weather and theterrestrial weather machines. Solar events, i.e.,explosions of charged particles and the dynamicsof magnetic and electric fields, cause hugechanges in the near-Earth space environment.Satelli tes and communication signals musttraverse this electric space.

Solar-Terrestrial Environment

FlaresFlares

Solar WindSolar Wind

MagnetosphereMagnetosphere

Y. K ami de

Interplaneta ry Magnetic Field

X-raysX-rays

IonosphereIonosphere

Coronal Mass EjectionsCoronal Mass Ejections

PolarGeosynchronousGeosynchronous

Regional Coupling in Sun-Earth System

1. From solar surface to interplanetary space

Solar flare to CME (Coronal Mass Ejection)

2. Solar wind-magnetosphere-ionosphere interactionCoupling between the CME and magnetosphere

3. Interstellar wind-heliosphere interaction Long term variation and solar cycle

SunSolar Wind

IMFMagnetosphere Ionosphere

AtmosphereThermosphere

Solar Physics

Solar wind-magnetosphere interaction

M-I Coupling

Thermosphere Model

GSFC/NASANRL, UCLA, MichiganNagoya, NICT, Kyusyu

(KRM, AMIE)

(NCAR et al)

Solar wind-magnetosphere-ionosphere-thermosphere interaction

Physical Model of Regional Coupling

Flare to CME

CME to Earth

Unified Model of Outer and Inner Magnetospheres and Ionosphere

40 Re

16 Re

10 Re

1.0 Re

1.5 Re

Outer Magnetosphere

Inner Magnetosphere

Ionosphere

-160 Re

Multi-Scale in Space-TimeInhomogeneous Grid

Earth

Cross-Scale Coupling in Sun-Earth System

(macro-, meso-, micro-scale coupling)1.Structure of magnetic reconnection (separation of

Ion-electron inertia scales) and particle acceleration

2. Dynamics of the high energy ring current particles

3. Variation of high energy particles in radiation belts

4. Particle acceleration/precipitation in aurora region

Magnetosphere: Cross scale coupling

磁気リコネクションセパラトリックスの構造

電子慣性スケールは重要?

電子電流ionyJ

electronyJ

流出

拡散領域

流入

電子慣性

イオン慣性 電子電流

イオン電流

Importance of magnetic reconnection for energy budgetCoupling between ion and electron inertia scales?

Structure of the magnetosphere and magnetic reconnection in the tail

Schematic diagram of the cross scale coupling between ions and electrons in the region of tail reconnection

Electron inertia scale

Ion inertia scale

Central region of reconnection

Geospace Study

GeospaceSun-earth system

CMEcoronal hale Solar wind-magnetosphere-

Ionosphere-thermosphere coupling

High energy particles of radiation beltsSolar wind-magnetosphereinteraction

Aurora

solar wind

Space Weather Map

Courtesy of NASA

electron

ionRing current

Inner magnetosphereouter belt

inner belt

Objectives of Virtual Laboratory

1.Real time accumulation of observation data and their mutual utilization

2.Produce new simulation/modeling data by physical model using observation data as input

3.Produce animation and 3D visualization (VRML, VR) data

4.Comparison and assimilation of simulation/modeling data with observations

5.Sharing of VR function through high speed network

・ Parallelization of 3D MHD code

3D array of (Nx,Ny,Nz)=(N,N,N)3D decomposition, (Px,Py,Pz)Number of cpu, 　 P=Px*Py*Pz

Supercomputing

・ From vector-parallel machine to scalar-parallel machine・ For high performance computation by scalar-parallel machine　　 Increase of the number of cpu in parallel computation　 Decrease of the amount of communication　 Increase of hit rate of cache

3. New Development of 3D MHD Code by Using MPI

Set up of environment

Initial condition

Read of previous data

Boundary condition

Data transfer (send+recv)

Computation of First Step

Data transfer (send+recv)

Computation of Second Step

Inner boundary condition

Upstream condition

Write of sampling data

End of simulation

Efficiency of Parallel Computation by Domain Decomposition

Com

putation Tim

e

（Ts

）

Number of Processors （ P）

1

101

102

103

104

10410310210111

101

102

103

104

Com

munication T

ime (T

c)

1-dimension

2-dimension

3-dimension

PTs

1

)1(1 PTc

)1(2 2

1

2 PTc

)1(3 3

1

3 PTc

To realize high performance computation by scalar-parallel supercomputer　　・ Longer inner do loop in vector-parallel machine, This brings to decrease of the hit rate of cache in scalar- parallel machine　　　　・ Decrease of amount of communication　　 Adoption of 2D and 3D domain decompositions

　・ Increase of the hit rate of cache　 Put related variables in nearer addresses

Four Key Functions of Supercomputer

　　　 Speed　　　 cpu 610 Gflops 12.48 Tflops

　 Main Memory 　　 　 Disc 　　 　 ＋　 Graphics　 1TB 11.5 TB 　　　　 17TB 50 TB 　　　　　 　 Animation, 3D Visualization

　　　　　 Network　　　　　 SuperSINET (10 Gbps)

VPP5000/64 (vector-parallel)HPC2500/1536 (scalar-parallel)

Supercomputer system of the Information Technology Center, Nagoya UniversityFujitsu PRIMEPOWER HPC2500/1536

21 Tflops11.5 TB

1975 1980 1985 1990 1995 2000103

104

105

106

107

108

109

1010

Year

To

tal G

rid N

umb

er,

Ng

2D and 3D MHD Simulation

○ ○

○

○○

○ ○ ○

○

○ ○

○

○

○○○

○

○○

○

○

○

○

○

○

○

○

●

●

●

●●

●

○○

○

△

△

△

▲ ▲ ▲

▲

▲

▲

▲

▲

△

3D MHD ○ ●

2D MHD△ ▲

ADENART×○○

○

○○

26*26*18

62*32*32

94*53*53

162*42*52

182*122*62

212*142*72

302*102*102

642*162*322

402*202*202

802*202*402

○

○△

18*18*12

××

○

○△

482*162*162

502*202*102

322*82*162

322*82*82

160*60*70

10 times/ 5yearsor

4 times/ 3 years

152*102*52182*62*62

1502*1002*11202240*560*1120

1002*1002*1120

1340*448*8961002*502*1120

802*202*672

2005

○

○

1024*1024*1024

2048*1024*1024

o 2048**3

Expansion of 2D and 3D MHD Simulations

To carry out Geospace Simulation

Use one of the largest supercomputers in the world

Improve numerical methods

Use efficient parallel computation methods

Understand well the simulation results by graphics

Virtual Reality (VR) System Using 3 Screens

Screen

Projector x 8

Gigabit Ethernet

Graphics computer (PC) x 8

Parallel control computer (PC)

Conceptual Diagram of PLAGS VR SystemConceptual Diagram of PLAGS VR System

Virtual Reality (VR) System

1. Use VRML, AVS, OpenGL and other softwares　　　　 VRML: Everyone can use freely and everywhere International Standard　 2. CAVE System ~2 M$　　　 3. PC VR System (with large screen) ~20 K$

4. PC VR System (with usual screen) 2~6 K$

polarizing grasses, unaided VR, high speed Internet

Plasmoid ejection in magnetotail by VRML

magneticreconnection

plasmoid

earth

CAWSES Space Weather Database in Japan

Scientific Committee on Solar-Terrestrial Physics (SCOSTEP)

CAWSES program (2004－ 2008) The Climate And Weather of the Sun-Earth System Program

STEP program (1990－ 1997) S-RAMP program (1998－ 2002)

The CAWSES Space Weather Database in Japan has been begun in order to promote the international collaborative CAWSES research program.

Computer / Network

Collaborative Research Program by Using Supercomputer of the Information Technology Center, Nagoya University

From vector-parallel machine, Fujistu VPP5000/64 to scalar-parallel machine, Fujitsu PRIMEPOWER HPC2500/1536

We have developed new programs which efficiently work in the scalar-parallel machine.

GIGAbit Network and TV Conference System

On-line lecture program using a video teleconferencing system New TV conference system using TCP/IP of Internet

From Gigabit Network (2000-2003) to JGN2 of NICT (2004-2008)

"Common Usage of the Information of Geospace Environment Using High Speed Network" for 2004-2008

Near real-time exchange and common usage of information and data on the Geospace (Solar-Terrestrial) Environment

Simulation, animation movie, 3-dimensional visualization, virtual reality (VR), 4-dimensional movie

NICT Koganei

Nagoya U

Kyoto U Ehime U, IT Center

Kyusyu U

Yamanashi

JGNⅡ

APAN Abilene

SINET

Kanto-2

100Mbps

10Gbps

1Gbps

100Mbps1Gbps

1Gbps

USA, China, Korea

SuperSINET and JGN-2 (1-10 Gbps)

Virtual ObservatoryTo estimate physical quantities in the huge space by using cross scale coupling model and exchange the information via high speed network

A synthetic and integrated research project on “Space Weather Forecast”

One of the e-Science Projects (2006~ )“Geospace Virtual Laboratory/Virtual Organization”　 Solar-Terrestrial Environment Laboratory and Information Technology Center of Nagoya University (+ JAXA/ISAS, Ehime U and NICT)

Use NAREGI-Beta version grid middleware(1) Data grid Laboratory data, S-RAMP data, CAWSES data(2) Graphics grid and virtual reality grid(3) Computing grid scheduler for supercomputers heterogeneous computer link programs minimizing communication amount

Geospace Research Center

Solar-Terrestrial Environment Laboratory

Division 1

Atmospheric Environment

Division 2

Ionospheric and Magnetosphric Environment

Division 3

Heliospheric Environment

Division 4

Integrated Studies

Joint Research Project

theory

simulation data archivesdata analysis

integration

observations

Domestic research institutions

Foreign research institutions

Promotion, Cooperation andmanagement

Virtual Laboratory　(1) Provide and share the data of observations and modeling/simulations

(2) Use mutually huge data of graphics, animation and 3-dimensional visualization through high speed network (Super-SINET, JGN-2: 10Gbps to 40Gbps? in 2008)

(3) Share the 3D visualization data (VRML, VR) Establish a system by which people can watch 3D figures of geospace under their own remote control

Real Time, Animation, Virtual Reality (VR), Advanced IT