A-19 
instance, to compute a model cell value at 5m deep, monitoring data at 5m deep are 
preferred. If fewer than n (typically 4) monitoring data values are found at the 
preferred depth, the depth window is widened to search up to d (normally +/-2m) 
meters above and below the preferred depth, with the window being widened in 0.5m 
increments until n monitoring values have been found for the computation. The 
smallest acceptable n value is selectable by the user. If fewer than n values are 
located, a missing value (normally a -9) is calculated for that cell. A second search 
radius filter is implemented to limit the horizontal distance of monitoring data from 
the cell being computed. Data points outside the radius selected by the user 
(normally 25,000m) are excluded from calculation. This filter is included so that 
only data that are near the location being interpolated are used. 
In this version of the Interpolator, Segment and Region filters have been added. 
Segments are geographic limits for the interpolator model. For instance, the Main 
Bay is composed of 8 segments (CB1TF, CB20H, ...,CB8PH). The tributaries are 
composed of 77 additional segments, using the CBP 2003 segmentation. These 
segments divide the Bay into geographic areas that have somewhat homogeneous 
environmental conditions. This segmentation also provides a means for reporting 
results on a segment basis, which can show more localized changes compared to the 
whole Bay ecosystem. 
Segment and bathymetry information use by the interpolator is stored in auxiliary 
files. Segment information allows the interpolator to report results on a segment 
basis which can show more localized changes compared to the whole Bay 
ecosystem. These segment and bathymetry files have been created for the main bay 
and all of the larger tributaries. The CBP segmentation scheme was replicated in 
these files by partitioning and coding the interpolator cells that fall within each 
segment. 
The interpolator also identifies the geographic boundary that limits which moni¬ 
toring station data are included in interpolation for a given segment through a region 
file. Use of data regions ensures that the interpolator does not “reach across land” to 
obtain data from an adjacent river which would give erroneous results. By using data 
regions, each segment of cells can be computed from their individual subset of moni¬ 
toring data. Each adjacent data region should overlap by some amount so that there 
is a continuous gradient, and not a seam, across segment boundaries. 
Current Implementation of CFD 
The Chesapeake Bay Program has initiated implementation of the CFD as an assess¬ 
ment tool. The Criteria Assessment Protocols (CAP) workgroup was formed in the 
fall of 2005 to develop detailed procedures for implementing criteria assessment. 
This workgroup has developed and implemented procedures that use the CFD 
process and conducted a CFD evaluation of dissolved oxygen for many designated 
assessment units. 
The CFD methodology was first applied in the Chesapeake Bay for the most recent 
listing cycle which was completed in the Spring of 2006 and was based on data 
collected over the period 2002 through 2004. CFDs were developed and utilized 
appendix a 
The Cumulative Frequency Diagram Method for Determining Water Quality Attainment 
