Derivation of Phenological Information from Remotely Sensed Imagery for Improved Regional Climate Modeling

• Activity Number: 519
• Type: Contributed
• Date/Time: Wednesday, August 1, 2012 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistics and the Environment
• Author(s): Jing Wang*+ and Nathan Moore and Jiaguo Qi and Lijian Yang and Jennifer Olson and Nathan Torbick and Jianjun Ge
• Companies: University of Illinois at Chicago and Michigan State University and Michigan State University and Soochow University/Michigan State University and Michigan State University and Applied Geosolutions and Oklahoma State University
• Keywords: confidence bands ; polynomial spline ; RAMS model ; leaf area index

Abstract

Phenological information reflecting seasonal changes in vegetation is an important input variable in climate models such as the Regional Atmospheric Modeling System (RAMS). It varies not only among different vegetation types but also with geographic locations (latitude and longitude). In the current version of RAMS, phenologies are treated as a simple sine function that is solely related to the day of year and latitude, in spite of major seasonal variability in precipitation and temperature. The sine curves of phenology are far different from the reality in many parts of the globe and, therefore, derivation of more representative phenological information would improve regional climate simulations. In this study, advanced spline techniques and remote sensing observations were used to develop a set of phenological functions for all land covers in the East Africa, and subsequently used in the RAMS model simulation analysis. The results show that the spline technique can effectively be used to characterize the phenological properties of most land cover types and the use of remotely sensed phenological information in regional climate simulations resulted in much more realistic climate conditions of the East Africa region. These spline phenologies are specifically needed for future climate projections when no remote sensing data are available.