A-30 
Table 3.2. Comparison of the capabilities of methods available for interpreting data 
collected for Chesapeake Bay water-quality criteria assessment. 
Attributes 
Sample-based 
IDW 
Kriging 
Provides Spatial 
Prediction 
Yes 
Yes 
Yes 
Provides Prediction 
Uncertainty 
No 
not routine 
Yes 
Uncertainty for CFD 
No 
No 
Yes 
Deal with Anisotropy 
No 
Possible, but 
not routine 
Yes 
Can Include Cruise 
Track/Fly over 
No 
Yes 
Yes 
Feasibility of 3 
dimensional 
interpolations 
No 
Yes 
Possible, but not 
routine 
Feasibility of mainstem- 
tributary interpolations 
No 
Yes 
Possible 
Inclusion of covariates to 
improve prediction 
No 
No 
Yes 
Predictions of non-linear 
functions of predicted 
attainment surfaces 
P(y>c) 
No 
No 
Yes 
Level of Sophistication 
Lowest 
Low 
Very High 
Automation 
Yes 
Yes 
Possible, but not 
routine 
4.0 CFD REFERENCE CURVES 
There are several approaches to defining reference curves that are proposed for use 
in the CFD assessment methodology. One is a biologically based definition and other 
approaches are based on an arbitrary allowable frequency (see Section 2). Here we 
review these options in greater detail. 
4.1. BIOLOGICAL REFERENCE CURVES 
The idea behind biological reference curves is to identify regions of the Bay that 
have healthy biological indicators and are thus considered to be in attainment of their 
designated use. CFDs would be developed for these areas in the same way that CFDs 
would be developed elsewhere, but those curves developed for healthy areas would 
be considered “reference” curves. For example, healthy benthic IBI scores might be 
used as indicators of adequate bottom dissolved oxygen. 
The success of the CFD-based assessment will be dependent upon decision rules 
related to the biological reference curves. These curves represent desired segment- 
designated use water quality outcomes and reflect sources of acceptable natural 
appendix a 
The Cumulative Frequency Diagram Method for Determining Water Quality Attainment 
