Physics of auroral phenomena : proceedings of the 33rd Annual seminar, Apatity, 02 - 05 March, 2010 / [ed.: A.G. Yahnin, A. A. Mochalov]. - Апатиты : Издательство Кольского научного центра РАН, 2011. - 206 с. : ил.

“Physics o fAuroral Phenomena”, Proc. XXXIIIAnnual Seminar, Apatity, pp. 17 - 20, 2011 Polar ©Kola Science Centre, Russian Academy of Science, 2011 GeoP^ys'cal W у Institute SOFTWARE-COMPUTER COMPLEX FOR SPACE WEATHER PREDICTING N.A. Barkhatov, S.E. Revunov (Nizhny Novgorod State Pedagogical University (NNSPU), Nizhny Novgorod, Russia) Abstract. Neural network software-computer complex to search for non-linear causal relationships in the problems of solar-terrestrial physics is developed. The main objective o f the program is to ensure neural network numerical experiments for search of non-linear connections between geomagnetic indices and near Earth space parameters. Work program algorithm of an artificial neural network (ANN) include back propagation Elman network, feed forward network, fuzzy logic network and Kohonen layer classification network. The interface of application allows to change neural network architecture (the number of hidden layers, the number o f neurons in the layer, the input and target data, and the number of cycles of training). 1. Review of existing computer complexes for space weather forecast One of the fundamental problems of solar-terrestrial physics is nonlinear laws of causc-effect relationships in system «solar wind - magnetosphere - ionosphere» is establish. Currently, the Internet has large number of web sites that provide information on geomagnetic conditions and the state of interplanetary space in real time runs. The best- known resource for today is the following: 1) World Data Center for Geomagnetism, Kyoto [http://swdcwww.kugi.kyoto-u.ac.jp] . On the web site the data on indices of geomagnetic activity, geomagnetic field is posted. 2) Support site for patrol spacecraft ACE [http://www.srl.caltech.edU/ACE/ASC/level2/index.h tml]. The resource provides information about the Interplanetary Magnetic Field (IMF) and Solar Wind Parameters (PSW). The description of the spacecraft instruments, data on the location of the spacecraft is posted. 3) Support site for patrol spacecraft WIND [http://pwg.gsfc.nasa.gov/windnrt] . The resource provides information about the IMF and PSW. The description of the spacecraft instruments, data on the location of the spacecraft is posted. 4) Support site for polar patrol spacecraft DMSP/POLAR [http://solar.uleth.ca/www/aurora.php] . The website contains the latest images of the polar regions o f , Earth, visible areas of the auroral activity. Provided related information. 5) Support site for spacecraft SOHO [http://sohowww.nascom.nasa.gov/data/latestimages. html]. The site contains links to the latest images of the Sun. Data provided in different formats - graphics, video, text. These resources provide real-time data with a refresh rate of 10-60 minutes. With access to the latest data and prediction/recovery model of any process that uses them in calculations, you can create software and hardware system (internet-machine) on the Internet the results o f this forecast published. Currently, most of existing internet machines predicting the dynamics of the global geomagnetic conditions described by the indices aimed, as a consequence of the interaction of Earth's magnetosphere with the interplanetary medium. Most recently -'nternet-machines Dst-index is predicted. Another area of intemet-machines with the restoration of maps of the distribution of energies for injected particles in the polar regions of Earth is directed. One more area in real-time data processing is to simulate the state of the magnetosphere - modeling geomagnetic field and distribution maps of the solar wind flow pressure on the magnetosphere is directed. Links to some existing intemet-machines are listed below. 1) Neural network restoration of Dst using real-time data N, V and Bz with the ACE spacecraft [http://www2.nict.go.jp/y/y223/sept/ace/nnw] . Advantages: providing data in digital form, providing the history of the forecast for 1 month. Disadvantages: the authors did not provide information about recovery algorithm, not provided description for ANN and comparison with the real Dst index, for example with Kyoto data [http://swdcwww.kugi.kyoto-u.ac.jp] 2) Recovery Dst index for real-time data from the ACE spacecraft based on the Temerin and Li 2006 model [http://lasp.colorado.edu] . Advantages: comparison of the model with actual Dst data from Kyoto [http://swdcwww.kugi.kyoto-u.ac.jp] . Providing data in digital form. Doing history prediction within 1 month. Providing links to real­ time data from the ACE spacecraft. Providing links to the description of the model for recovery Dst index. 3) Recovery Dst index for real-time data from the ACE spacecraft and the presentation of the current state o f geomagnetic conditions [http://sprg.ssl.berkeley.edu/dst_index/] . Advantages: providing a convenient form to evaluate the current geomagnetic conditions. Disadvantages: not provided information about the recovery algorithm. ^, H r ; , r0 --J;HH a 1 j обдастн.'.Я •,!: •• :• *