[ASA Logo]

American Statistical Association (ASA)
Section on Statistics and the Environment (ENVR)

Last Modified: 2000-Aug-12
Send comments, corrections, additions to:
ENVR Web Editor or ENVR Webmaster		 [ENVR Logo]

AmStat Section News, June 2000
Section on Statistics and the Environment (ENVR)


View from the Chair
Barry D. Nussbaum, 2000 ENVR Chair

I sincerely hope by now you are busy making plans to attend the Joint Statistical Meetings in Indianapolis August 13-17, 2000. As you can see from the article below the program committee has a full schedule of informative sessions and interesting activities. I would especially like to remind you of two points that Kathy Ensor makes. First, there is something important everyday, so make your arrangements for the entire program including Thursday, August 17th. Second, the JSM is an outstanding opportunity to meet colleagues in fields aligned with yours. This occurs at the section meeting Monday evening, at the roundtables Wednesday, and throughout the week by meeting and greeting in the hallways. I want to thank Kathy for her excellent efforts in organizing this year's program and look forward to chatting with each and every one of you in August.

JSM 2000 - Plan to Stay Through Thursday Morning!
Katherine Bennett Ensor, 2000 ENVR Program Chair

If you are making travel plans to attend JSM 2000 this August in Indianapolis, plan to stay through the Thursday morning sessions! We conclude the program for the Section on Statistics and the Environment with two extremely strong sessions: Session 263 organized by Montserrat Fuentes, Spatial Production of Environmental Processes (8:30 on Thursday) followed by a Regular Contributed Session on Spatial Analysis and Methodologies for Environmental Problem Solving (Session 285 at 10:30 on Thursday). Of particular note in this latter session is the talk by our student award winner Deepak Agarwal, Investigating Tropical Deforestation Using Two Stage Spatially Misaligned Regression. Congratulations Deepak for your excellent entry! Session 285 represents a cohesive and strong set of papers covering many aspects of spatial modeling.

Business Meeting and Mixer: Of interest to Section members will be the ASA sponsored poster session on Sunday from 6 to 8 pm. Also, don't forget to leave time for the open business meeting of the Section that will be held at 6:00pm on Monday. Immediately after the open business meeting is the annual mixer. The business meeting and mixer are a perfect opportunity to meet other statisticians interested in environmental research. Another opportunity for informal interaction on an environmental topic is attending a roundtable luncheon.

Roundtable Luncheons: The Section is hosting five roundtable luncheons on Wednesday. The hosts are Montserrat Fuentes, Peter Guttorp, Anthony Olsen, Ta-Hsin Li and Stephan Sain. Topics of the luncheons include discussions of spatial statistics, model assessment, survey design, global climate change and environmental justice, respectively. The roundtable luncheons provide a unique opportunity for focused interaction between individuals with a wide range of experience and interests. Personally, I have always found the roundtable luncheons a worthwhile experience.

The Program: The Section is the primary sponsor of eight sessions and the co-sponsor of an additional five sessions. As environmental researchers are aware, data collected in many environmental settings is either spatially or temporally correlated, or both. The analysis of such data is a primary theme to the JSM 2000 program. This topic appears in all of the five invited and topic contributed sessions now listed:

The talks within these sessions cover many aspects of spatial and spatial-temporal modeling and environmental data analysis. In Session 64 we learn: 1) which spatial data analysis utilities should be included in Geographical Information Systems; 2) the importance of spatial-temporal modeling when inferring chlorophyll levels from remote sensing data; 3) how to combine multiple sources of data available for spatio-temporal problems with application to snow water equivalent data; and 4) methods for spatial interpolation of nonstationary processes. Session 99 covers such innovative topics as: 1) spherical wavelets as a method of analyzing scattered spherical data; 2) varying-coefficient functional data analysis as an approach for studying the time-evolution in the shape of the vertical profile curves of measurements of atmospheric dynamics and chemical constituents; 3) nonparametric estimation of the spatio-temporal correlation structure with application to sea surface temperatures; and 4) spatial-temporal modeling of ozone readings over Mexico City. The third session, Session 112 highlights the importance of assessment of deterministic models for 1) exposure, 2) air pollution modeling and includes 3) a discussion of a comprehensive approach to uncertainty for deterministic and hybrid models of ecological and environmental processes. A look at climate changes brings us to Session 236 in which we explore 1) methods of detecting faint signals in a sparsely measured system such as greenhouse warming, 2) learn how multiresolution bases can assist with capturing nonstationary structure and 3) how to simulate non-Gaussian random fields by clipping. This simulation method provides a means of evaluating variability of functions from satellite images. The fifth session, Session 263 begins with a presentation of 1) two new systems for modeling, estimating and predicting multivariate spatio-temporal processes followed by 2) an approach to combining multiple sources of information for purpose of improving understanding of the biological integrity of stream and river systems in the United States Mid-Atlantic Region. This second talk provides a different perspective to the same topic of presentation 3 in Session 64. We next learn of a Bayesian approach to spatial prediction with applications to annual averages of fine particulate data from the national monitoring network in the conterminous U.S. The last presentation in this session addresses the important issue of formulating models with spatial temporal dependence as models with spatially varying temporal dependence or temporally varying spatial dependence.

The regular contributed sessions form an amazingly cohesive set. In Session 148 we will learn about the National Survey of Lead and Allergen in Homes (Marker et al), sampling methods and sampling design as well as new methodologies in multivariate receptor modeling. Session 223, focuses on detecting trends in water quality, the degradation of ecosystems, mapping of forest inventories, real-time forecasting of daily streamflows as well as variations in El Nino Southern Oscillation. Following the general theme of spatial methods, in Session 285 we learn about spatial periodograms, finite populations in spatial data, non-stationary spatial processes, and Bayesian approaches for model averaging, the analysis of radon exposure data and the hierarchical analysis of spatial count data. Session 285 also features our Student Award Winning paper as previously noted. The three regular contributed sessions are:

In addition to the above eight sessions, the Section is co-sponsoring five sessions, namely:

The first of the above five sessions focuses again on the difficult problem of spatial modeling; the last addresses the important question of environmental justice. Session 29 involves new methodologies but includes examples as diverse as radioactive contaminant measurements and stream flow measurements. In Session 144, researchers present new methodologies that may prove useful in environmental applications. And finally, in Session 172 we will learn about Visualizing Abandoned Hazardous Waste Sites in the United States (by Carolyn Offutt) and Mapping and Modeling 1-Meter Multispectral Imagery Data (James Shiren)

As you can see, through the significant contributions of Section members and other researchers, JSM 2000 proves to be an exciting conference for statisticians interested in environmental research. If you have ideas for JSM 2001, contact Timothy G. Gregoire (timothy.gregoire@yale.edu), Program Chair-Elect for the Section on Statistics and the Environment. Make sure to check next month's ENVR Amstat News column, too, for further details about next year's JSM and Tim's solicitation of invited papers and posters.

[Return t ASA] [Return to ENVR]