A-58 
review with the goal of implementing the best available technology for this 
aspect of criteria assessment. One of the first efforts under this task is a study 
of the 3-D variance stucture of the data to be interpolated. A short term option 
is to implement the optimal 2-D interpolator in layers as is done with the 
current IDW interpolator. 
5. High Density Temporal Data. As currently formulated, assessment for most 
of the open-waters of the Bay are based on “snapshots” in time of the spatial 
extent of criteria exceedence estimated via interpolation. Data collected for use 
in interpolation are actually spaced over multiple days due to the large expanse 
over which sampling must be conducted. It is clear that technology is becoming 
available that will produce high density data in both space and time. Interpola¬ 
tion should accommodate data that are collected densely in space. However, it 
is unclear how the CFD process will accommodate data that are high density in 
time. Further work is needed to evaluate methods to fully utilize the temporally 
intensive data that is currently being collected. 
The panel discussed several mechanisms for the CBP to make progress on chal¬ 
lenging tasks ahead (Table 7.1). We recommend that a review panel oversee the tasks 
over the next 3-5 year time frame. This panel would periodically review trials and 
other products conducted by individual external scientists (academic scientists or 
consultants) and existing teams of CBP scientists (e.g., the Criteria Assessment 
Protocols (CAP) workgroup). Tasks 1 and 2 are most immediate and critical and we 
recommend that these tasks by contracted out to external scientists, exploiting state- 
of-the-art approaches and knowledge. Task 3 could be conducted through CAP or 
other group of CBP scientists. Task 4 and 5 are less immediate but again will require 
substantial expertise and innovation and may be most efficiently accomplished by 
scientific expertise outside the immediate CBP community. 
REFERENCES 
Christensen OF, Diggle PJ, Ribeiro PJ. 2001. Analysing positive-valued spatial data: the 
transformed Gaussian model. In: Monestiez P, Allard D, and Froidevaux, editors. GeoENV 
III - Geostatistics for environmental applications. Quantitative Geology and Geostatislics. 
Dordrecht (Netherlands): Kluwer Academic Publishers. 11:287-298. 
Christman MC. 2006. The characterization and incorporation of uncertainty in fisheries 
management, In Fisheries Ecosystem Planning for Chesapeake Bay. Bethesda (MD): Amer¬ 
ican Fisheries Society. (In press). 
Cressie N. 1989. The Many Faces of Spatial Prediction. Mathematical Geology 1:163- 176. 
Cressie N. 1991. Statistics for Spatial Data. New York: Wiley. 928 p. 
Curriero FC. 2006. On the Use of Non-Euclidean Distance Measures in Geostatistics. Math¬ 
ematical Geology (in press). 
Daubert v. Merrell Dow Pharmaceuticals. Inc. 1993. 509 U. S. Supreme Court. 579. 
Deutsch CV. 1984. Kriging with Strings of Data. Mathematical Geology. 26:623-638. 
Diggle PJ, Tawn JA, Moyeed RA. 1998. Model Based Geostatistics (with Discussion). 
Applied Statistics 47:299-350. 
Diggle PJ, Ribeiro PJ. 2006. Model-based Geostatistics. New York: Springer. 230 p. 
appendix a 
The Cumulative Frequency Diagram Method for Determining Water Quality Attainment 
