A-7 
1. INTRODUCTION 
In June 2000, Chesapeake Bay Program (CBP) partners adopted the Chesapeake 
2000 agreement (http://www.chesapeakebay.net/agreement.htm), a strategic plan 
that calls for defining the water quality conditions necessary to protect aquatic living 
resources. These water quality conditions are being defined through the development 
of Chesapeake Bay specific water quality criteria for dissolved oxygen, water clarity, 
and chlorophyll_a to be implemented as state water quality standards by 2005. One 
element of the newly defined standards is an assessment tool that addresses the 
spatial and temporal variability of these water quality measures in establishing 
compliance. This tool has become known as the Cumulative Frequency Diagram 
(CFD). 
The (CFD) was first proposed as an assessment tool by Paul Jacobson, of Langhei 
Ecology (www.LangheiEcology.com). At that time Dr. Jacobson was consulting 
with the Chesapeake Bay Program as a member of the Tidal Monitoring Network 
Redesign Team. Within this group, the CFD concept gained immediate recognition 
and support as a novel approach that permitted independent modeling of the time and 
space dimensions of the continuous domain that underlies Chesapeake Bay water 
quality parameters. In addition, because preparation of the CFD uses spatial inter¬ 
polation, the approach can allow integration of data collected on different spatial 
scales such as fixed station data and cruise track data. 
While the benefits of the CFD approach has been recognized (U.S. EPA 2003) and 
the the CBP has begun implementation of the approach for certain water quality 
parameters and segments of the Chesapeake Bay, investigations of the statistical 
properties revealed that the underlying shape parameters of the CFD were sensitive 
not only to rates of compliance but also to sampling design elements such as sample 
density. The novelty of the approach coupled with concerns about its statistical 
validity motivated the Chesapeake Bay Program to request that its Scientific and 
Technical Advisory Committee (http://www.chesapeake.org/stac/) empanel a group 
with expertise in criteria assessment, spatial data interpolation, and statistics to 
assess the scientific defensibility of the CFD. Here we report the findings of this 
panel. 
The primary goal of this panel is to provide an initial scientific review of the CFD 
compliance approach. This review addresses a wide range of issues including: bias 
and statistical rigor, uncertainty, practical implementation issues, and formulation of 
reference curves. Because of the novelty of the CFD approach, the panel has endeav¬ 
ored to research and explain the properties of the CFD and spatial modeling upon 
which the CFD approach depends to provide a basis for this evaluation. These activ¬ 
ities are beyond the scope of the typical review. However, because so little is known 
about the CFD, it was necessary to expand the knowledge base. 
The report is organized into 7 sections. In Section 2 of this report we present the 
CFD approach as a series of steps, each of which needs to be considered carefully in 
evaluating its statistical properties. Spatial interpolation is a critical but the most 
statistically nuanced step in the CFD approach. Spatial interpolation of water quality 
data in the CBP has to date received little statistical review. In Section 3 we evaluate 
appendix a 
The Cumulative Frequency Diagram Method for Determining Water Quality Attainment 
