Dr. Rajib Paul, WMU professor of statistics, recently was invited by The International Environmetrics Society to talk in a session titled “Applications of Spatial Statistics to Environmental Science.”
The meeting, which took place in Hyderabad, India from Jan. 3-6 2012, aimed to encourage the implementation of statistical methods in environmental engineering, monitoring and care.
Paul’s talk, which drew from research performed in collaboration with scientists from the Commonwealth Scientific and Industrial Research Organization (CSIRO) in Australia, reflects on his three months of research in Australia in summer 2009. Titled “Reduced Rank Spatial Model With Temporal Confounding Effects,” Paul says the talk discussed “novel statistical methods to model complex processes in ‘big’ science.” Such processes, like measuring ocean salinity and temperature, have a very large global impact indeed.
Paul’s project deals with the mathematical analysis of the nonstationarity of processes such as ocean salinity and temperature. Previous studies, he notes, have neglected key factors (such as spatial nonstationarity) in measuring these processes; they also have not properly accommodated the fact that data is collected from flowing water, which requires a great deal of attention.
In order to collect data on the biogeochemical properties of different ocean depths, Paul’s team used an ocean glider (a small, autonomous subterranean vehicle, which collects data on ocean biogeochemical properties), which was deployed off the coast of southern Tunisia. The glider gathered information from several different ocean layers—a very time-consuming and expensive process, though less so than more common ship-based techniques.
Once in possession of the valuable data, Paul’s attention was focused on using it to create a reliable and inclusive statistical model for future use in environmental and ocean sciences.