1 Introduction

An international joint research projectcalled the International Geophysical Year(IGY) was conducted from 1957 to 1958. Forthe IGY, worldwide observation networks anddata centers called the World Data Center(WDC) were developed to study the ionos-phere, geomagnetic field, the sun, cosmicrays, and other topics of interest.

During this IGY period, a 24-hour forecastof geomagnetic storms called Space WorldInterval (SWI) was exchanged among partici-pating institutions for joint international obser-vation of the sun and the earth’s environment.This SWI is one origin of space weather infor-mation service provided by the InternationalSpace Environment Service (ISES). In Japan,the Radio Research Laboratories (today’sNational Institute of Information and Commu-nications Technology) conducted the SWI. Atthe time of the IGY, there was no means ofcommunications capable of sending largequantities of data immediately; thus, observa-tion information was encoded and sent to dif-ferent sites by telex. Each site then convertedthe information into Morse code and broad-casted it by shortwave. Figure 1 shows an

example of encoded observation data. Theseactivities were later inherited to form theInternational URSIgram and World Days Ser-vice (IUWDS). Figure 2 shows the informa-tion exchange in international collaboration inthe IUWDS.

Among the events held during the days ofthe IGY were humankind’s first artificialsatellite launched by the Soviet Union (today’sRussian Republic), the launch of Sputnik 1,and the discovery of the radiation belt byExplorer 1, the U.S. artificial satellite. Morethan half a century has passed since that time,

485WATARI Shinichi

4-1-3 Space Weather Forecast Using Real-time Data

WATARI Shinichi

In recent years, advances in telecommunications technology have made it possible to col-lect space-based and ground-based observation data needed for space weather forecasts innear real-time. Real-time data also make it possible to recognize the space environment condi-tions through continuous monitoring and the nowcast of space weather for issuing necessarywarnings about space storms. Moreover, many other applications are conceivable, such asnumerical predictions using real-time data as inputs in a prediction model. This paper describesthe roles of real-time space weather monitoring.

KeywordsReal-time monitoring, Space weather, International Space Environment Service,International Geophysical Year

Fig.1 Example of encoded observationdata

486 Journal of the National Institute of Information and Communications Technology Vol.56 Nos.1-4 2009

and numerous artificial satellites are nowbeing used for communications, broadcasting,positioning, meteorological observation, andother purposes. Through satellite disruptionsexperienced due to the effects of the spaceenvironment, people recognize the need topredict the space environment (known as“space weather”) that affects manmade tech-nological systems in space and on the ground.The IUWDS was renamed the InternationalSpace Environment Service (ISES) in 1996 asan international organization for such spaceweather forecast. And now, people around theworld are conducting research on spaceweather［1］–［3］.

The rapid spread of the fast Internet inrecent years has made it possible to obtainlarge quantities of observation data fromground-based observatories and satellites onan almost real-time basis. This in turn hasmade it possible to proceed with joint observa-tions while monitoring observation dataobtained from those ground-based observato-ries and satellites, and has expanded the rangeof possible joint observations, such as flexiblymodifying the object of observation accordingto the object’s activity. In space weather, it hasbecome possible to continuously monitor the

space environment by using real-time data,thereby recognizing that status. Another thingmade possible is automatically detecting thearrival of solar energetic particles, interplane-tary shocks, and other hazardous phenomena,for which warnings are sent to users. Real-time data from satellites and ground-basedobservation networks are monitored at theJapanese Space Weather Forecast Center ofthe National Institute of Information andCommunications Technology (NICT) as indi-cated in Fig. 3 and, in case a space stormoccurs, a warning is sent to the users. Suchreal-time data can be used not only to recog-nize the space environment conditions but alsoas inputs for a prediction model. The use of aprediction model will hopefully increase theprecision of prediction. NICT has beenattempting to predict the geomagnetic Dstindex with a neural network by using real-timesolar wind data［4］. The simulation model ofthe earth’s magnetosphere using magnetohy-drodynamic code is calculated in a near real-time manner with real-time solar wind data.NICT has been developing real-time simula-tion models of the ionosphere and of the sunand solar wind. The simulation model of theionosphere is calculated by using the results of

Fig.2 Information exchange by international collaboration in the IUWDS

487WATARI Shinichi

possible to issue early warnings to usersregarding the arrival of solar energetic parti-cles, occurrence of geomagnetic storms,enhancement of high-energy electron flux on ageosynchronous orbit, and other phenomenahazardous to technological systems. Anotherpossibility is detecting the occurrence of ahazardous space environment automaticallyand then issuing warnings about space storms.For example, Shinohara et al［8］. developed anautomatic detection system for Storm SuddenCommencement (SSC) based on real-timedata from NICT’s ground-based geomagneticobservation network. This system uses real-time data from the geomagnetic observationnetwork to automatically detect impulsivechanges in the geomagnetic field associatedwith the arrival of interplanetary shocks usingthe maximum values of amplitude, rise timeand temporal change, and then reports theresults of detection by email. Figure 4 is awebpage of space weather information show-ing the onset of geomagnetic storms, occur-rence of Dellinger phenomenon, arrival ofsolar energetic particles, enhancement of high-energy electron flux, and occurrence of a spo-radic E layer by automatic detection［9］. Nearreal-time data from the U.S. meteorological

the magnetosphere simulation model. Thesimulation model of the sun and solar winduses solar magnetic field data as input.

One effective way to collect real-time datafrom observation points without a well-estab-lished communication infrastructure is byusing artificial satellites. One ongoing pio-neering project for collecting real-time datausing artificial satellites is an internationalproject called the INTERMAGNET (Interna-tional Real-time Magnetic Observatory) thatwas launched at the end of the 1980s［6］. Tocollect geomagnetic data from observationpoints, the INTERMAGNET uses the commu-nication channels of meteorological satellites(GOES, Himawari, and Meteosat). NICTbecame a member of the INTERMAGNETand has collected geomagnetic data on a real-time basis from the Kakioka Magnetic Obser-vatory, the Memambetsu Magnetic Observato-ry, and other observatories using the meteoro-logical satellite Himawari.

This paper describes the usefulness ofreal-time data on space weather, includingexamples of NICT［7］.

2 Space storm warnings based onreal-time data

Real-time data enables contentious moni-toring of the space environment in order torecognize the conditions thereof. This makes it

Fig.3 NICT’s Japanese Space WeatherForecast Center that monitors real-time data from satellites andground-based observation networks

Fig.4 Webpage that provides spaceweather information

488 Journal of the National Institute of Information and Communications Technology Vol.56 Nos.1-4 2009

satellite GOES, solar X-rays, solar high-ener-gy particles, and high-energy electrons areused for the automatic detection of events.

3 Real-time data as inputs for thespace weather forecast model

Real-time data can be used as inputs for aprediction model. The use of a predictionmodel will hopefully increase the precisionof prediction. The NASA-launched ACE(Advanced Composition Explorer) spacecrafthas been observing solar wind at point L1where the sun and earth are balanced in gravi-ty upstream from the earth by approximately1.5 million km, and continuously sendingsolar wind data on a real-time basis since1997. NICT has been sharing the reception ofreal-time data from the ACE spacecraft incooperation with the National Oceanic andAtmospheric Administration/Space WeatherPrediction Center (NOAA/SWPC)［10］. Real-time solar wind data from the ACE spacecraftenables the arrival of solar wind disturbancesto be detected from approximately 30 minutesto one hour before reaching the earth. One

attempt to use real-time solar wind data is thereception of data from the International SunEarth Explorer 3 (ISEE-3) spacecraft conduct-ed from March 1980 to mid-1982 byNOAA/SEC (today’s NOAA/SWPC) with thecooperation of NASA for early warningsabout geomagnetic storms. At that time, datawas used mainly to monitor solar wind condi-tions［11］.

Several prediction systems have beendeveloped based on a neural network modelby using real-time solar wind data. Watanabe

Fig.5 Prediction of the Dst index basedon a neural network using real-timedata as inputs

Fig.6 Prediction of high-energy electronflux on a geosynchronous orbitbased on a neural network usingreal-time data as inputs

Fig.7 Results of real-time simulation of theearth’s magnetosphere based onmagnetohydrodynamic code usingreal-time solar wind data as inputs

489WATARI Shinichi

et al［3］. developed a prediction model for thegeomagnetic Dst index, an indicator of geo-magnetic storms, by using the Elman neuralnetwork. Watanabe et al［3］. used real-timesolar wind data (e.g., velocity, density, mag-netic field intensity, x, y, and z components ofthe magnetic field) as inputs. Figure 5 showsan example of Dst index predictions based onthis model. Watari et al［12］. developed a pre-diction model of high-energy electron fluxvariations on a geosynchronous orbit based ona neural network. An enhancement of high-energy electron flux on a geosynchronousorbit causes artificial satellite failure due todeep dielectric charging. Therefore, the pre-diction of such enhancements is necessary.High-energy electron flux on a geosynchro-nous orbit can be predicted 24 hours inadvance by using present high-energy electronflux observed by the GOES satellite and real-time solar wind data from the ACE spacecraft.Figure 6 shows an example of predicting high-energy electron flux on a geosynchronousorbit based on this model.

Real-time solar wind data can also be usedas inputs for a numerical model. Figure 7shows the results of real-time simulation ofthe earth’s magnetosphere, as developed byNICT based on magnetohydrodynamic codeusing real-time solar wind data as inputs［4］.Results of this simulation are open to the pub-lic on a webpage. The geomagnetic indexcalled the auroral electrojet (AE) index—anindicator of aurora activity—is calculated byusing the results of this simulation［13］.

4 Roles of real-time data in spaceweather research

Real-time data has made it possible tocontinuously monitor data from space-basedand ground-based observations. This increasesthe flexibility of joint observations. For exam-ple, this has made it easier to flexibly changeobservation timing and the object of observa-tion according to the status of the object’sactivity, by such means as determining thetiming for launching an observation rocket.

The continuous monitoring of real-time dataopens up the possibility of discovering newphenomena from hitherto unrecognized real-time data streams. Moreover, the real-timeacquisition of data makes it possible to verifyideas by promptly referring to relevant datawhenever a new phenomenon is discovered.

5 Real-time space weather dataas materials for education andpublic outreach

Real-time space weather data are excellentmaterials for education and public outreach.Images of auroras and the sun provided on areal-time basis allow the general public tobecome more familiar with present spaceweather conditions. For example, the SOHO(Solar and Heliospheric Observatory) space-craft launched by ESA and NASA allows nearreal-time images of the sun taken by whitelight, extreme ultraviolet, and coronagraphs tobe made available to the public on the web-page shown in Fig. 8［14］. Thus, the generalpublic can access the webpage and enjoy thereal-time solar images.

Space weather sometimes affects our dailylives and could be the entry point for attract-ing public interest in space sciences. Forexample, when a strong geomagnetic stormoccurred in March 1989, a geomagnetically-

Fig.8 Webpage of near real-time solarimages from the SOHO spacecraft

490 Journal of the National Institute of Information and Communications Technology Vol.56 Nos.1-4 2009

induced current caused a major power black-out in Canada’s province of Quebec, affectingabout six million people［15］. Moreover, spacestorms disrupt communication satellites andbroadcasting satellites, even resulting in com-munication and broadcasting services beingoccasionally interrupted［16］.

6 Technology needed for real-time monitoring

The global deployment of an observationnetwork to collect real-time data requires care-ful consideration of how that data is to be col-lected, securing a power supply, providingmaintenance in case of equipment breakdown,and other matters. At unattended observationpoints, the recovery from equipment trouble isparticularly difficult. One must therefore con-sider building a highly reliable system,installing observation points at locations offer-ing easier access in case of trouble, and othermatters. Regarding the power supply, there arecases where no power supply is available or,even if available, a blackout may occur orvoltage may become unstable. It is also neces-sary to take measures against blackouts byusing an uninterruptible power supply, solarbatteries, wind generators or similar equip-ment, or to develop equipment operating atlow power consumption. When installing anobservation point at a location without a well-established communication infrastructure as ameans of collecting data, one conceivable ideafor collecting data on a real-time basis is touse artificial satellites. For example, the Iridi-um satellite system engages in mutual com-munications among 66 satellites that orbit theearth at an altitude of about 780 km, therebyachieving a global communication network byusing small terminals. The main problemposed by this system is its still high communi-

cation charges, but the use of the Iridium sys-tem enables the collection of data on analmost real-time basis from global ground-based observation networks. The NationalInstitute of Polar Research in Japan has devel-oped unattended geomagnetic observationequipment using the Iridium system for datatransmission, and has thus built a wide-rang-ing geomagnetic observation network inAntarctica［17］. As a ground-based communi-cation network, the network of mobile tele-phony is rapidly spreading, making real-timedata collection based on mobile telephonyconceivable. Moreover, given the higher per-formance and lower costs of wireless network-ing equipment, the use of a wireless networkbetween an observation point and the nearestaccess point of an Internet network makes itpossible to build a data collection network atrelatively low cost.

7 Conclusion

To attract attention in a timely manner inresponse to the arrival of a space storm thatmay affect our technological systems, such asby disrupting artificial satellites, it is neces-sary to monitor the space environment at alltimes. Recent advances in telecommunicationstechnology have made it possible to acquirespace environment data on a real-time basisfor monitoring the space environment. How-ever, when considering whether to build aglobal real-time observation network, onefinds that there still are regions where a com-munication network is difficult to secure, suchas the Arctic and Antarctic regions. As afuture challenge, hopes run high for the devel-opment of a highly reliable and relatively low-cost communication system that enables thecollection of real-time data.

References01 P. Song, H. Singer, and G. Siscoe (eds), “Space Weather,” Geophys. Monogr. Ser., Vol. 125, Washing-

ton, D.C., pp. 11–22, 2001.

491WATARI Shinichi

02 I.A. Daglis(ed.), “Effects of Space Weather on Technology Infrastructure,” NATO Science Series, Kluwer

Academic Publishers, Dordrecht, Netherlands, 2004.

03 S. Watari, “Impacts of space storms on technologies and space weather forecast (in japanese),”

J. Plasma Fusion Res., Vol. 82(11), pp. 739–744, 2006.

04 S. Watanabe, E. Sagawa, K. Ohtaka, and H. Shimazu, “Prediction of the Dst index from solar wind para-

meters by a neural network method,” J. Communications Research Laboratory, Vol. 49(4), pp. 69–85,

2002.

05 M. Den, T. Tanaka, S. Fujita, T. Obara, H. Shimazu, H. Amo, Y. Hayashi, E. Nakano, Y. Seo, K. Suehiro,

H. Takahara, and T. Takei, “Real-time Earth magnetosphere simulator with three-dimensional magneto-

hydrodynamic code,” Space Weather, Vol. 4, S06004, doi: 10.1029/2004SW000100, 2006.

06 http://www.intermagnet.org/Welcom_e.html

07 S. Watari, “Space weather and real-time monitoring,” Data Science Journal, Vol. 8, pp. S78–S84, 2009.

08 M. Shinohara, T. Kikuchi, and K. Nozaki, “Automatic realtime detection of sudden commencement of

geomagnetic storms,” Journal of National Institute of Information and Communications Technology, Vol.

52(3/4), pp. 197–205, 2005.

09 http://www.swc.nict.go.jp/

10 E. Sagawa, S. Watanabe, K. Ohtaka, and H. Shimazu, “Realtime data reception of ACE and IMAGE

satellites,” J. Com. Res. Lab., Vol. 49(4), pp. 55–67, 2002.

11 J. A. Joselyn, “Real-time prediction of global Geomagnetic activity,” in Solar Wind-Magnetosphere Cou-

pling, edited by Y. Kamide and A. Slavin, Tokyo, TERRAPUB, pp. 127–141, 1986.

12 S. Watari, M. Tokumitsu, K. Kitamura, and Y. Ishida, “Forecast of high-energy electron flux at geosyn-

chronous orbit using a neural network method (in Japanese),” IEICE Technical Report, SANE2007-83,

pp. 7–12, 2007.

13 K. Kitamura, H. Shimazu, S. Fujita, S. Watari, M. Kunitake, H. Shinagawa, and T. Tanaka, “Properties of

AE indices derived from real-time global simulation and their implications for solar wind-magnetosphere

coupling,” J. Geophys. Res., 113, A03S10, doi: 10.1029/2007JA012514, 2008.

14 http://sohowww.nascom.nasa.gov/data/realtime-images.html

15 D. H. Boteler, “Space Weather Effects on Power Systems,” in Space Weather, Geophys. Monogr. Ser.,

Vol.125, edited by P. Song, H. Singer, and G. Siscoe, AGU, Washington, D.C., pp. 347–352, 2001.

16 E. J. Daly, “Outlook on Space Weather Effects on Spacecraft,” G., in Effects of Space Weather on Tech-

nology Infrastructure, NATO Science Series, edited by I. A. Daglis, Kluwer Academic Publishers, Dor-

drecht, Netherlands, pp. 91–108, 2004.

17 A. Kadokura, H. Yamagisi, N. Sato, K. Nakano, and M. C. Rose, “Unmanned magnetometer network

observation in the 44th Japanese Antarctic Research Expedition: Initial results and an event study on

auroral substorm evolution,” Polar Science, Vol. 2(3), pp. 223–235, 2008.

WATARI Shinichi, Dr. Sci.

Research Manager, Space EnvironmentGroup, Applied ElectromagneticResearch Center

Solar-Terrestrial Physics, SpaceWeather