Scientific talk by Mubasharah Omer (Factas)
Field Extrapolation for Inverse Magnetization Problems
Abstract:
Scanning microscopes such as the Super-conducting Quantum Interference Device (SQUID) are being increasingly utilized in the context of paleomagnetic studies. For magnetized rock samples, these microscopes measure the vertical component of the weak magnetic field, produced by the magnetization within the sample, on a plane above it. Unfortunately, these magnetometers do not directly provide information about the magnetization, which may offer a wealth of information regarding the origin and age of the magnetized sample itself. To retrieve this quantity from measured data of the magnetic field is an ill-posed inverse problem that is unstable and has non-unique solutions. Moreover, in utilization of already developed methods for solving inverse problems, the magnetometers pose serious shortcomings: the measured field data is spatially limited, and its measurements are polluted by noise. Thus, for better performance of inversion techniques, the ill-conditioning of such inverse problems must be reduced, for instance, by extending the range of the data set through extrapolation techniques to get a more complete, clean, and wide image of the magnetic field. In this effort, we explore one approach to extrapolate and simultaneously denoise the measured field data, thereby uniquely solving the inverse problem for a function related to the magnetization.
When: Monday, November 25 at 2pm
Where: Euler Violet
