Hierarchical Statistical Analysis of Global and Regional Environmental DataEPA Grant Number: R827257
Title: Hierarchical Statistical Analysis of Global and Regional Environmental Data
Investigators: Cressie, Noel A.C. , Berliner, Mark
Current Investigators: Cressie, Noel A.C. , Berliner, Mark , Wilke, Christopher K.
Institution: The Ohio State University
Current Institution: The Ohio State University , University of Missouri - Kansas City
EPA Project Officer: Louie, Nica
Project Period: September 1, 1998 through August 31, 2001 (Extended to March 31, 2003)
Project Amount: $325,000
RFA: Environmental Statistics (1998) RFA Text | Recipients Lists
Research Category: Environmental Statistics , Health , Ecosystems
Hierarchical spatio-temporal statistical models will be used to provide optimal change-of-resolution filtering of massive environmental data sets from polarorbiting satellites. Compromises between statistical efficiency and computational speed will be assessed. Data from field studies. which are expensive but measure the environmental process of interest directly, will be combined with cheaper, more plentiful satellite or meteorological data to obtain high-resolution maps of the process of interest.
The underlying approach to all three objectives is to use multivariate hierarchical statistical models as a framework for inference. To analyze massive data sets from polar-orbiting satellites, a spatial change-of-resolution model with temporal autoregressions at the coarsest scale will be used. Both Bayes and empirical Baves inference are to be implemented to gauge their effectiveness with massive data sets. Multivariate spatio-temporal dependencies are modeled through conditional distributional dependencies. Combining data can then be carried out, conceptually at least, in a statistically optimal manner.
Statistically optimal formulas and algorithms for filtering Earth Observing System (EOS) satellite data onto a regular global grid are expected. Recommendations for achieving a compromise between statistical efficiency and computational speed are also expected. Finally, we expect to complete an environmental study of a problem for which combining information from various sources is needed.
Improvements in Risk Assessment or Risk Management:
A primary goal of EOS is to measure, through a series of remote-sensing instruments and satellites. climate change. Global warming would impose risks of an uneven nature throughout the world (e.g., coastal regions would be inundated by rising sea levels and lower soil moisture could be expected in the Midwest USA). This research provides optimal filtering of the EOS data that is in turn used in General Circulation Models (GCMs), and from which risks of various kinds can be calculated. Health (ecosystem or human) data will be incorporated into the problem of combining information, so that ecosystem or human health risks can be assessed accurately and precisely.