M.S. candidate, Department of Atmospheric Sciences – University of Illinois
October 31, 2007
3:00 pm: Conversation and Cookies in Room 109 ASB
3:30 pm: Seminar in Room 144 Loomis Lab
Dynamic Global Vegetation Models (DGVMs) are relatively new in the field of atmospheric science. They will soon require validation techniques if they are to improve. However, the datasets for these models are very large in volume and variety. Principal Component Analysis is a statistical technique which is designed for such datasets, and whose purpose is to identify key variables and discard superfluous ones. PCA was performed month-by-month on three years of Ameriflux weather data from their Bondville, IL site. Retention techniques were then used to identify which variables were most important and how their importance changed throughout the year. This information can perhaps be taken into account when improving vegetation models and implementing new data measurement projects.
<< Back to Colloquia Schedule