15 
Also plotted on this same axis in blue in Figure II-6 is a default reference curve 
based on 10 percent of the area of the percent space x percent time (the default refer¬ 
ence curve described previously and illustrated in Figure II-4). This evaluation was 
undertaken given an approach to deriving and assessing attainment of numerical 
chlorophyll a criteria is based largely on thresholds that should rarely be exceeded 
in healthy populations (e.g., the 90 th percentile). These two curves are very close in 
shape, further supporting the use of the default reference based on a 10 percent area 
under the curve. The EPA recommends use of the default reference curve, illustrated 
in Figure III-4 and defined by Equation 1, when an applicable biologically-based 
reference curve is not available. 
Figure 11-6 Comparison of hyperbolic curves based on 10 percent of area under the curve 
(blue) and 10 percent frequency exceedance (black). 
COMPARING ASSESSMENT AND REFERENCE CURVES 
Reference curves are more or less continuously defined while assessment curves 
have relatively few discrete measures. Biological reference curves can contain 
hundreds of points; the default reference curve has an infinite number of points. By 
contrast, curves for three-year assessments of summer (June-September) monthly 
means will have 12 data points with the curve defined by linear interpolation 
between neighboring points. For this reason, it is possible for portions of the assess¬ 
ment curve to be above the reference curve even without any measured point 
exceeding the reference curve. This situation becomes more comprehensible by 
chapter ii 
Refinements to the Chesapeake Bay Water Quality Criteria Assessment Methodology 
