University of Maryland, USA
Author earned his PhD degree in Solar Photvoltaics Engineering from IIT, Delhi. He has worked as a postdoctoral fellow at the EU Joint research Centre, Italy, and research scientists at the Queensland University of Technology, Australia on data analysis and quality issues of Electrical grid systems. Later he became interested in more science-focused field of satellite remote sensing retrievalsapplying physically deterministic methods. For a decade, he carried his research work in different countries at reputed academic institutes i.e. University of Toronto, University of Bonn and Dalhousie University. He reviewed articles for IEEE journals and has given invited talks different reputed organizations including IEEE society.
There are two distinct different schools of thoughts in parameter estimation: deterministic and stochastic. Deterministic inverse method assumes that there are true values of parameters, as opposed to all parameters in stochastic method are remaining as random variables. However, stochastic methods are dominating for operational satellite remote sensing retrievals for a long time, where many ambiguities are persistingin the reliability and quality of retrieved products. One of major drawbacks in prevalent stochastic inverse method is that errors are treated as definite information and use these as the input parameters. This is not the case for the deterministic method, thus ambiguities are inevitably low. First I will present the classical difference between two inverse methods in this talk. Later I will demonstrate a successful implementation of deterministic method in operational environment for near real time geophysical parameters estimation from satellite measurements. The various advantages using deterministic method over the stochastic method will be discussed subsequently.