American Statistical Association (ASA) Section on Statistics and the Environment (ENVR) Last Modified: 98-Mar-04

Dr. David Marker, senior statistician at Westat and member of theEPA-funded National Research Center for Statistics and the Environment,has been appointed to a 2-year term as the ENVR liaison officer. ENVRplans to use the position to maintain connections with environmentallyoriented statisticians in societies such as the InternationalEnvironmetrics Society, the International Statistical Institute,International Biometric Society, as well as to establish connections withother societies having an environmental focus. In addition, the liaisonofficer will explore the offering of section activities that are relevantfor non-statistical environmental professionals who use statistics aspart of their work.

The Section's proposal for a short course "Environmental Statisticswith ENVIRONMENTALSTATS" was accepted by the organizers of the Conferenceon Water Research Methods, Crete, Greece. Even though, we will not beable to run the course this time due to lack of travel funding forthe instructors, it is clear that Continuing Education (CE) is aneffective liaison activity with other organizations.So, we need to continue to look for such opportunities.We hope to be sponsoring a course in the upcoming Soils conference.We would like members to suggest meetings of this nature for CE activities,where we arelikely to meet scientists who do not normally attend the JSM.

Continuing Education seems to be one of the many activitiesto be benefited from the listserv of theSection (for listserv details, see the ENVR website at(http://www.pnl.gov/statenvi/). I am looking at the possibilityof setting up a Hypernews node in the Section web page to facilitate discussionson specific issues on CE. Since funding becomes a key issue in offeringcourses in various conferences, it may be time to consider offeringsome of our courses as "virtual seminars".

Please forward suggestions and ideas on the Continuing Educationthread to Nagaraj K. Neerchal, Continuing Education Coordinator, atnagaraj@math.umbc.edu.

The National Research Center for Statistics and the Environment (NRCSE,http://www.stat.washington.edu/NCES/) held a workshop, "CombiningInformation From Programs That Monitor Ecological and Natural Resources,"on November 21-22, 1997, at the University of Washington, Seattle, WA. Theobjective of this small workshop was to set an agenda for research (i.e.,directing attention to important problems where solutions are feasible andalso identifying problems that are not likely to be solved). Participantsin the workshop were Joe Sedransk, Tony Olsen, Loveday Conquest, PeterGuttorp, Jim Karr, Nick Chrisman, Mark Handcock, Mark Kaiser, Phil Larsen,Adrian Raftery, Hans Schreuder, Denis White, Steve Rathbun, Gary Oehlert,Abdel El-Shaarawi, and Dale Zimmerman.

Background information for the workshop was provided by two papers"Statistical Issues for Monitoring Ecological National Resources in theU.S.," by Olsen, et al. (to appear in Environmental Monitoring andAssessment) and "Integrating the Nation'sEnvironmental Monitoring and Research Networks and Programs: A ProposedFramework," a governmental report (March 1997) written by theEnvironmental Monitoring Team of the Committee on Environment and NaturalResources (CENR), National Science and Technology Council (available athttp://www.epa.gov/cludygxb/pubs.html). Participants presented informationon specific monitoring programs where a clear need for combininginformation existed. Other presentations described statisticalmethodologies that had been used in specific situations to combineinformation.

Currently, the information needed for effective monitoring of ecologicaland natural resources is obtained from many programs. These include (a)national, probability based sample surveys conducted on a regular schedule,(b) regional and local surveys, and (c) research-based intensive sitestudies. The population about which inferences can sensibly be made isusually clear for a study of type (a), but not necessarily for (b) or (c).For example, in (b) or (c) the selection of sites may be based onconvenience. Alternatively, sites that are thought to be "representative"may be chosen, but the criteria for selection are not specified. The scopeof any program in (a) - (c) is limited. For example, some surveys in (a)may be limited to non-Federal land, while others are carried out only onFederal land.

It is unlikely that a single, probability-based sample survey of the statusand/or trend of ecological and natural resources will be established.Hence, it is important to discern whether or not one can combineinformation from several of the programs (e.g., probability-based samplesurveys, intensive-site analyses, non-probability-based program) to providebetter inferences. The statistical problems are formidable because theprograms do not make observations at the same geographical sites, and theset of variables measured in one program will be different from thosemeasured in another program.

The principal question addressed at the workshop is the feasibility ofcombining information obtained by different programs. A corollary isinvestigating the possibility of modifying some of the existing programs tofacilitate the successful pooling of data. Workshop participants committedto prepare a brief article that discusses the research issues in combiningnatural resource data. Some participants are initiating research to pursuecombining information from specific monitoring programs.

For information about the workshop contact Tony Olsen(tolsen@mail.cor.epa.gov) or Joe Sedransk (jxs123@po.cwru.edu)