• FIEEE, FSPIE, FINNS
• Dean, College of Engineering
• Harris Professor of Electrical Engineering
• Florida Institute of Technology
• USA

Abstract:

Volcano eruptions emit infrasound signals that contain information pertaining to the intensity of the eruption, presence of ash emissions, and certain characteristics of the volcano itself. Knowledge of the eruption intensity can provide an estimate of the height of the ash column. This talk focuses on exploiting the infrasonic characteristics of volcanos by extracting unique cepstral-based features from the volcano’s infrasound signature. Two sets of results will be presented: (1) These volcano feature vectors are then used to train and test a neural-classifier that is developed to distinguish the ash generating eruptive activity from three volcanoes, namely, Mount St. Helens-Washington, USA, Tungurahua-Ecuador, and Kasatochi-Alaska, USA. The neural classifier that consists of three neural modules (one for each of the volcanos) is able to correctly distinguish the eruptive activity of each of the three volcanos with a correct classification rate (CCR) of more than 96%. (2) A separate set of cepstral features are extracted from time-domain infrasound signals and used to train, test and validate a neural classifier that can distinguish between a set of volcano plinian-type (in the infrasound frequency range 0.01 Hz to 0.1 Hz) eruptions (sometimes referred to as Vesuvian eruptions), and other infrasound signals that are Not-of-Interest (NOI) but have frequency content in the bandwidth of the plinian eruptions (that is, tsunamis, mountain associated waves (MAW), avalanches and bolides). In this case the neural architecture consists of two neural modules. One neural module is associated with the volcano activities for four different volcanos, namely, Lascar-Northern Chile, Augustine-Alaska, USA, Kasatochi-Alaska, USA, and Manam-Papua, New Guinea. That is, the neural module is responsible for classifying a plinian eruption event associated with any of the four volcanos. The second neural module is associated with classifying the collective activity of the NOI events previously mentioned. In this initial study, to obtain an understanding of the neural-classifier’s ability to distinguish between the volcanic events and other natural events, the NOI events are considered to be made up of only the MAW infrasound signals. For this case, the performance for the two-neural module classifier is greater than 90% CCR.

Biosketch: 

Dr. Ham received his B.S. (with distinction), M.S. and Ph.D. degrees in electrical engineering from Iowa State University in 1976, 1979 and 1980, respectively.

Dr. Ham has been a faculty member at Florida Tech since 1988. He is a Fellow of IEEE, SPIE and INNS, and a member of Eta Kappa Nu, Tau Beta Pi, Phi Kappa Phi and Sigma Xi. He is the past president of the International Neural Network Society (INNS) (2007–2008), and served on the INNS Board of Governors from 2009-2011. From 1977-1978 he worked for Shell Oil Company as a Geophysicist, and from 1980-1988 he was a Staff Engineer at Harris Corporation in Melbourne, Florida, where he worked: in the Systems Analysis Group (there he performed all of the error analysis for the control algorithms for the Hubble Space Telescope); and the Large Space Structures Controls Group (there he developed highly-robust control algorithms for flexible space structures).

Dr. Ham has over 100 technical publications and is author of the textbook: Principles of Neurocomputing for Science and Engineering, McGraw-Hill, 2001. His research interests include neural networks, tactical infrasound systems, adaptive signal processing, image processing, speech processing, and biosensors.