A*16 
of error propagation. Developing the statistical algorithms to quantify this uncer¬ 
tainty is challenging. 
Even if the uncertainty can be properly quantified, the issue of who gets the benefit 
of doubt due to this uncertainty is a difficult question to resolve. This is a broad 
sweeping issue regarding uncertainty in the regulatory process, not a problem 
specific to the CFD approach. None-the-less, it must be dealt with here as well as 
elsewhere. One option is to require that the assessment curve be significantly above 
the reference curve to establish noncompliance. This option protects the regulated 
party from being deemed out of compliance due to random effects, but if assessment 
CFD curves are not accurately determined, it could lead to poor protection of envi¬ 
ronmental health and designated uses. A second option is to require that the 
assessment curve be significantly below the reference curve to establish compliance. 
This results in strong protection of the environmental resource, but could lead to the 
regulated party implementing expensive management actions that are not necessary. 
Some compromise between these extremes is needed. The simplest compromise is 
to ignore variability and just compare the assessment curve to the reference curve. 
As long as unbiased estimation is implemented for both the assessment curve and the 
reference curve, this third option will result in roughly equal numbers of false posi¬ 
tive (declaring noncompliance when in fact compliance exists) and false negative 
(declaring compliance when in fact noncompliance exists) results. This offers a 
balanced approach, but there is no mechanism to motivate a reduction of these false 
positive and false negative errors. 
2.2 DATA AVAILABLE AND CURRENT METHODS 
OVERVIEW OF TYPES OF DATA AVAILABLE 
The Chesapeake Bay monitoring program routinely monitors 19 directly measured 
water quality paramenters at 49 stations in the mainstem Bay and 96 stations in the 
tidal tributaries. The Water Quality Monitoring Program began in June 1984 with 
stations sampled once each month during the colder late fall and winter months and 
twice each month in the warmer months. A refinement in 1995 reduced the number 
of mainstem monitoring cruises to 14 per year. “Special” cruises may be added to 
record unique weather events. The collecting organizations coordinate the sampling 
times of their respective stations, so that data for each sampling event, or “cruise”, 
represents a synoptic picture of the Bay at that point in time. At each station, a hydro- 
graphic profile is made (including water temperature, salinity, and dissolved oxygen) 
at approximately 1 to 2 meter intervals. Water samples for chemical analysis (e.g., 
nutrients and chlorophyll) are collected at the surface and bottom, and at two addi¬ 
tional depths depending on the existence and location of a pycnocline (region(s) of 
density discontinuity in the water column). Correlative data on sea state and climate 
are also collected. 
In addition, Chesapeake Bay Program partner organizations Maryland Department 
of Natural Resources and the Virginia Institute of Marine Science have recently 
begun monitoring using a technology known as data flow. DATAFLOW is a system 
of shipboard water quality probes that measure spatial position, water depth, water 
appendix a 
The Cumulative Frequency Diagram Method for Determining Watei Quality Attainment 
