Natesh Pillai | Two Vignettes of Uncertainty Quantification Due to Measurement Error in Climate Science

ABSTRACT: In this talk, we will discuss uncertainty quantification on two different problems in climate science, connected by the common theme of measurement error. In the first half of the talk, we will revisit constructing sea surface temperatures in early 20th century using the data obtained from ship tracks. Careful examination of the data reveals new challenges that were not noticed before. Next, we will describe the problem of quantifying uncertainties for present day temperature from weather stations. Both problems serve as templates for the utility for data science methods in this field.BIO: Pillai is a professor in the Department of Statistics at Harvard University, having joined the university in 2010. .He was born and raised in Kerala, India, and earned his bachelor’s degree from the Indian Institute of Technology (IIT) Madras, Chennai. He did graduate studies at Duke University in Durham, North Carolina, and was awarded his Ph.D. in statistics in 2008. After Duke, I was a postdoctoral research fellow at the Center for Research in Statistical Methodology (CRiSM), University of Warwick, UK, during 2008-10.