An optimal parametrization framework for infrasonic tomography of the stratospheric winds using non-local sources. Issue 3 (5th December 2016)

• Record Type: Journal Article
• Authors: Blom, Philip S.; Marcillo, Omar E.
• Is Part Of: Geophysical journal international. Volume 208:Issue 3(2017:Mar.)
• Page Start: 1557
• Page End: 1566
• Publication Date: 2016-12-05
• Subjects: Inverse theory -- Wave propagation -- Acoustic properties; Geophysics -- Periodicals; 550
• DOI: 10.1093/gji/ggw449
• Languages: English
• ISSNs: 0956-540X

Abstract

Summary: A method is developed to apply acoustic tomography methods to a localized network of infrasound arrays with intention of monitoring the atmosphere state in the region around the network using non-local sources without requiring knowledge of the precise source location or non-local atmosphere state. Closely spaced arrays provide a means to estimate phase velocities of signals that can provide limiting bounds on certain characteristics of the atmosphere. Larger spacing between such clusters provide a means to estimate celerity from propagation times along multiple unique stratospherically or thermospherically ducted propagation paths and compute more precise estimates of the atmosphere state. In order to avoid the commonly encountered complex, multimodal distributions for parametric atmosphere descriptions and to maximize the computational efficiency of the method, an optimal parametrization framework is constructed. This framework identifies the ideal combination of parameters for tomography studies in specific regions of the atmosphere and statistical model selection analysis shows that high quality corrections to the middle atmosphere winds can be obtained using as few as three parameters. Comparison of the resulting estimates for synthetic data sets shows qualitative agreement between the middle atmosphere winds and those estimated from infrasonic traveltime observations.