Please use this identifier to cite or link to this item:
Title: Artificial neural network (ANN) for modelling earth’s magnetic field belonging to solar minimum observed at a low latitude station Alibag
Authors: Unnikrishnan, K
Keywords: Artificial neural network;Earth’s magnetic field;Solar minimum;Geomagnetic daily variation model
Issue Date: Jun-2012
Publisher: NISCAIR-CSIR, India
PACS No.: 84.35.+i; 91.25.Rt; 96.60.qd
Abstract: Artificial neural networks (ANNs) are well suited to environmental modelling as they are nonlinear, relatively insensitive to data noise, and perform reasonably well when limited data are available. By using solar flux (F10.7), day of the year, local time, and Ap as input, an appropriate ANN has been developed to model north-south component (X component) of earth’s magnetic field belonging to solar minimum period for Alibag (18.6°N, 72.87°E, geomagnetic latitude 10.37°N), a low latitude station of Indian sub-continent. For training the network three months, namely February, June, and September, were selected which represent three seasons winter, summer, and equinox, respectively of 2007 and 2008. Based on this analysis, it is observed that ANN model with 10 hidden neurons has good performance for 500 iterations. For testing the efficiency of ANN, hourly values of input and north-south component (X component) of earth’s magnetic field observed during January, October 2007 and May 2008 were used. To confirm the functional aspects of this model, similar investigations were carried out for other periods, January, October 2008, and May 2007. In this study, for the first time, artificial neural networks (ANNs) are utilised to develop a geomagnetic daily variation model for an Indian sub-continent station, Alibag.
Page(s): 359-366
ISSN: 0975-105X (Online); 0367-8393 (Print)
Appears in Collections:IJRSP Vol.41(3) [June 2012]

Files in This Item:
File Description SizeFormat 
IJRSP 41(3) 359-366.pdf555.04 kBAdobe PDFView/Open

Items in NOPR are protected by copyright, with all rights reserved, unless otherwise indicated.