Siddhartha Nandy

Post date: Sep 3, 2013 3:33:50 PM

Title:

A VARIABLE SELECTION TECHNIQUE FOR DETECTING CLIMATE CHANGE ATTRIBUTION

Abstract: There is a considerable interest of developing sound quantitative and computationally feasible statistical tools suitable for analyzing spectral radiance change data to monitor the climate change and attributions. In this context, we developed a variable selection technique, specifically adaptive group LASSO type of selection, followed by a semi-parametric additive model. We validate the method by simulation studies and applying to a simulated Climate Absolute Radiance and Refractivity Observatory(CLARREO) - like measurements. Theoretical validity will also be discussed.