Motivation behind the Development of Inversion Methods in Geophysics: Sean Twomey and his Influence (Invited Presentation)

In the early 1960s Sean Twomey was intrigued by the pioneering paper by Phillips on solutions to integral equations and the problem of unwanted oscillations found in numerical solutions. He explored how to replace a two-matrix solution to this problem with a single matrix, which had the additional attribute that he was not restricted to solving for the same number of ‘unknowns' as ‘measurements.' This led to exploring whether more measurements led to better solutions and eventually to studying the number of independent pieces of information in remote sensing, whereby increasing the number of measurements did not in and of itself reduce the uncertainty in the solutions or bring with it more ‘information'. I will describe this early pioneering work, its use by Twomey in solving for the size distribution of aerosol particles in a cloud chamber following filtering by nuclepore filters, and the extensions of this work today up to and including the widely used inversion of aerosol size distribution and refractive index in the AERONET system of ground-based sun and sky radiometers.