|
 |
|
|
|
 |
National Research Program on Design-Based/Model-Assisted
Survey Methodology for Aquatic Resources
|
| |
| EPA Grant Number: R829096 |
| Title: National Research Program on Design-Based/Model-Assisted
Survey Methodology for Aquatic Resources |
| Investigators:
Stevens,
Don
L.
,
Herlihy,
Alan
T.
,
Urquhart,
N.
Scott
,
Lesser,
Virginia
,
Hughes,
Robert
,
Munoz-Hernandez,
Breda
,
Opsomer,
Jean
,
Courbois,
Jean-Yves
,
Conquest,
Loveday
,
Smith,
Ruben
,
Sifneos,
Jean
,
Murtaugh,
Paul
,
Gitelman,
Alix
I.
,
Breidt,
F.
Jay
|
| Institution:
Oregon State University
,
Colorado State University
,
Iowa State University
,
University of Washington
|
| EPA Project Officer:
Moore,
James
C.
|
| Project Period:
October 15, 2001
through
October 14, 2005
|
| Project Amount:
$2,989,884
|
| RFA:
Research Program on Statistical Survey Design and Analysis for Aquatic Resources (2001)
|
| Research Category:
Aquatic Ecosystem
|
Description:
Our objectives for this Program are to extend existing EMAP survey design and analysis methodology to enhance the use of auxiliary information; to develop design-based model -assisted methodology compatible with EMAP's multi-tiered studies; to implement these techniques on example data sets; to provide expert survey design and analysis assistance to States and Tribes; and to transfer that statistical expertise to States and Tribes through a combination of supervised application of statistical tools and structured distance learning techniques.
Approach:
Our approach will be to initiate statistical research on several well-defined
topics in model-assisted estimation: non-parametric model-assisted design-based
estimation, estimation and prediction using empirical orthogonal functions on
probability survey data, model-assisted imputation, model-assisted survey
design, and analysis of multi-year panel studies. We have made preliminary
contacts with some State and Tribal agencies who have a connection to EMAP. We
will strengthen and expand those contacts to include all identifiable State and
Tribal agencies who are using (or plan to use) EMAP, REMAP, or EMAP-like surveys
of aquatic resources. We will assist those agencies in applying existing survey
design and analysis methods, and, in particular, investigate and extend
techniques for incorporating auxiliary information in the design. In
collaboration with those agencies and other State and Tribal organizations, we
will identify further statistical research needs to be addressed by our
statistical methodology research teams.
Three state agencies (Oregon Department of Fish and Wildlife (ODFW),
California State Water Resources Control Board (CSWRCB), and the San Francisco
Estuary Institute(SFEI)) which have existing or planned aquatic monitoring
programs that use EMAP's sampling methodology have a special role. These
agencies will identify areas where our statistical expertise will be especially
beneficial and act as "laboratories" for testing distance learning methods. They
will provide guidance on the areas where they feel they need more knowledge; we
will structure the content and delivery method to meet those needs; and they
will provide feedback and critique.
The interaction between these three agencies and this Program has the
potential to develop into model archetypes for state and local level monitoring
programs. The three programs represent a microcosm that spans almost all of the
circumstances that state or local aquatic monitoring efforts are likely to
encounter. ODFW initiated a survey of Coho salmon in Oregon streams in 1998
using a rotating panel design and EMAP's spatially-balanced sample selection
procedure. The sampling is expected to continue indefinitely using the same
design, with the objective of estimating both annual escapement of coho salmon,
and trends in the escapement. The sample has nested subsamples of juvenile
counts and habitat quality, with landscape level information available on
several variables, such as vegetation cover, stream order, and basin size. Both
spatial and temporal correlation are present. In addition, there is substantial
non-response, which may be related to such information as landholder type and
size of holding. Thus, the data set provides fertile ground for developing
model-assisted, design-based estimation techniques, such as trend estimation,
imputation of missing data, small-area estimation, and a variety of
model-assisted status estimation techniques. This is a critical, high-visibility
issue in the Pacific Northwest because it involves a commercially and
recreationally important species that is listed as threatened by the federal
government.
The SFEI is presently concluding a re-design of the San Francisco Estuary
Regional Monitoring Program for Trace Substances. While the design has not been
finalized, it will almost certainly consist of an EMAP-like probability sample,
coupled with more intensive samples at selected sites. SFEI is now initiating
the design of a Bay Area Wetlands Regional Monitoring Program. This design
effort will provide us with the opportunity to investigate some model-assisted
design approaches on a relatively small scale. The regional monitoring conducted
by SFEI is an example of a state-supported estuarine monitoring program
encompassing the estuary itself and its associated wetlands.
The third
archetype program is the Surface Waters Ambient Monitoring Program (SWAMP) being
initiated by CSWRCB. This will be a state-wide program with multiple objectives,
which include an EMAP-like assessment, e.g., an estimate of the proportion of
the resource in nominal or degraded condition. This is the first state-funded
project of this scale that may use an EMAP-like sampling design. Its success may encourage other states to follow.
Expected Results:
We expect to accomplish five major goals with the Program. The first is to extend design-based statistical methodology to cover the unique circumstances encountered in EMAP. The second is to make both existing and newly-developed model-assisted design-based statistical tools more accessible to State and Tribe personnel, both the aquatic scientists and managers of aquatic resources. The third is to expand the pool of personnel in the States and Tribes who have both understanding of and experience in using the statistical tools. The fourth is to develop a cadre of statisticians with the experience and expertise to collaborate on monitoring aquatic resources. Finally, we will develop three archetypes of rigorous probability-based, state or local monitoring programs, along with archetype design-based, model-assisted analyses. The existence of these archetypes will benefit EMAP's efforts to build state, tribal, and local infrastructure to monitor the condition of the Nation's aquatic resources.
Supplemental Keywords:
public policy, decision making, community-based, monitoring, risk assessment, watersheds, streams, rivers, estuaries.
, Economic, Social, & Behavioral Science Research Program, Ecosystem Protection/Environmental Exposure & Risk, RFA, Scientific Discipline, Water, Applied Math & Statistics, Aquatic Ecosystem, Aquatic Ecosystems & Estuarine Research, Ecological Effects - Environmental Exposure & Risk, Ecological Risk Assessment, Economics & Decision Making, Ecosystem Protection, Ecosystem/Assessment/Indicators, Engineering, Chemistry, & Physics, Environmental Monitoring, Monitoring/Modeling, Urban and Regional Planning, decision-making, Bayesian approach, EMAP, aquatic, aquatic resources, community-based, community-based approach, decision making, empirical orthogonal functions, environmental data, environmental decision making, environmental decision-making, estuaries, landscapes, model assisted estimation, model-based analysis, modeling, monitoring, public policy, risk assessment, rivers, spatial analysis, statistical methodology, statistical tools, surface water, survey
|
| |
Relevant Websites:
http://www.orst.edu/dept/statistics/epa_program/index.htm 
|
| |
| Progress and Final Reports:
2002 Progress Report
|
|
