IONOSPHERIC ELECTRON DENSITY ANOMALIES DETECTEDBY BP ARTIFICIAL NEURAL NETWORK BEFORE 2008 M8.6 INDONESIA EARTHQUAKE
Ren Chengmian 1) ; and Xiong Jing 2,3)
1)Second Surveying and Mapping Institute of Guizhou Province,Guiyang 550004;2)Key Laboratory of Earthquake Geodesy,Institute of Seismology,CEA,Wuhan 430071;3)Wuhan Base of Institute of Crustal Dynamics,CEA,Wuhan 430071
Abstract:On the basis of the F2 layer peak electron density(NmF2) from University Corporation for Atmospheric Research(UCAR),U.S.A., we constructed a Back Propogation(BP) artificial neural network(ANN) into which inputting five parameters:DOY(day of year), local time(LT), longitude(LON), latitude(LAT) and solar activity index of F10.7(FLUX) and then using the Matlab ANN for building the local ionosphere model and made the simulation.Through comparing the model value with observations during the 2012 Indonesia earthquake period,it is found that around the epicenter a obvious NmF2 decrease appeared on day 16(26 March) before the earthquake, the rate of decline to 40%.
Ren Chengmian ,and Xiong Jing. IONOSPHERIC ELECTRON DENSITY ANOMALIES DETECTEDBY BP ARTIFICIAL NEURAL NETWORK BEFORE 2008 M8.6 INDONESIA EARTHQUAKE[J]. jgg, 2012, 32(Supp.): 28-31.