Presenter: Dr. Hans Ngodock, Naval Research Laboratory 

Cost: Free!

When: Friday, April 22, 11:30 a.m.

Where: Zoom/Dalton J. Woods Auditorium; Energy, Coast and Environment Bldg. 

Despite its great practical importance, the theory of inverse problems remains poorly known. Indeed, physics was built by solving direct problems: from a given model to the corresponding measurable data. The opposite approach or the solution of an ‘inverse’ problem is not as straightforward: from observed data to model parameters. It is, however, essential in many fields. To determine the properties of the studied system from the observed data, i.e., to solve an inverse problem, is an otherwise difficult operation because it proceeds opposite to the normal process of experimental science. In mathematics it is, for example, the famous problem of the drum: find its form knowing the sound that it produces. One has then to search the coefficients of a differential equation or of a set of partial differential equations, knowing only the spectrum of the eigenvalues. This problem is often encountered when one wants to determine the properties of a medium that is inaccessible to direct observation, from observations made outside the medium. This is obviously the case in internal geophysics. This presentation focuses on the variational approach to solving inverse problems with an application to oceanography.