Speaker: Yingchao Zhou, Graduate Student, Department of Statistics, Iowa State University

Title: Alaska Permanent Snow And Ice Classification

Abstract : Classifying permanent snow and ice points based on satellite images can be challenging. Using the NRI 2007 survey samples in Alaska as ground truth, we combine two remote sensing datasets MODIS and ABoVE with glacier inventory dataset, and train random forest predictor to classify whether a location is permanent snow and ice. In this talk, I will show the datasets and the procedure of filling the missing data in MODIS and ABoVE as well as model training in detail. Future plan of this project will be discussed.