A-8 
alternative geostatistical methods as they pertain to the CFD approach. The CFD 
approach is an attainment procedure, which depends upon statistical comparison 
between attainment and reference curves. In Section 4, we present alternative types 
of references curves and discuss statistical properties of each. In Section 5 the statis¬ 
tical properties of CFD curves (applicable to both attainment and reference curves) 
is elucidated through a series of conditional simulation trials. 
In addition to this primary charge, the panel is sensitive to the fact that the CFD will 
be employed in the enforcement of water quality standards. Use as a regulatory tool 
imposes a standard of credibility, which we review in Section 6. We use here “best 
available science” and “best science” criteria to evaluate the overall validity and 
feasibility of the CFD approach, following guidelines established by the American 
Fisheries Society and Estuarine Research Federation (Sullivan et al. 2006). These 
follow other similar criteria (e.g., The Daubert Criteria (Daubert v. Merrell Dow 
Pharmaceuticals, Inc., 1993) and include: 
1. A clear statement of objective 
2. A conceptual model, which is a framework for characterizing systems, sating 
assumptions, making predictions, and testing hypotheses. 
3. A good experimental design and a standardized method for collecting data. 
4. Statistical rigor and sound logic for analysis and interpretation. 
5. Clear documentation of methods, results, and conclusions 
6. Peer review. 
The panel has made progress in better understanding statistical properties of the 
CFD approach and overall, we recommend it as a feasible approach and one that 
qualifies under most criteria for best available science. Still, we believe that our 
efforts should only represent the beginning of a longer term effort to (1) Use simu¬ 
lations and other means to support statistical comparisons of CFD curves; and (2) 
Support the CBP’s efforts to model water quality data with sufficient rigor in both 
spatial and temporal dimensions. Research and implementation recommendations 
follow in Section 7. 
2.0 BACKGROUND 
2.1 THE CFD ASSESSMENT APPROACH 
The water quality criteria assessment methodology currently proposed by the E.P.A. 
Chesapeake Bay Program (CBP) involves the use of a Cumulative Frequency 
Diagram (CFD) curve. This curve is represented in a two dimensional plane of 
percent time and percent space. This document briefly discusses the reasoning that 
lead to the development of this assessment tool. The proposed algorithm for esti¬ 
mating the CFD is given and illustrated with small data sets. Some properties and 
unresolved issues regarding the use of the CFD are briefly discussed. In Section 5, 
simulation studies explore in greater specificity the multiple issues related to error 
and bias in the CFD approach. 
appendix a 
The Cumulative Frequency Diagram Method for Determining Water Quality Attainment 
