A-56 
Table 7.1. Research Tasks, examples of specific subtasks, and suggested time frame for 
continued CFD research. 
Task 
Schedule 
1. Effects of Sampling Design on CFD Results 
2006-2008 
(a) Continue simulation work to evaluate CFD bias reduction 
via conditional simulation. 
(b) Investigate conditional simulation for interpolation 
methods other than kriging - this may lead to more simulation work. 
(c) Implement and apply interpolation with condition 
simulation on CBP data. 
2. Statistical inference framework for the CFD 
2006-2008 
(a) Implement and evaluate confidence interval procedures. 
(b) Conduct confidence interval coverage experiments. 
(c) Investigate confidence interval methods for non-kriging 
interpolation methods. 
(c) Implement and evaluate confidence interval procedures. 
3. Choice of Interpolation Method 
2006-2008 
(a) continue to investigate other more nonparametric 
interpolation methods (e.g. loess and splines). 
(b) implement a file system and software utilizing the “best” 
interpolation for CBP data. 
(b) compare interpolations and CFD's based on IDW and 
“best” method. 
4. Three-Dimensional Interpolation 
2007-2009 
(a) Implement 2-D kriging in layers to compare to current 
approach of 2-D IDW in layers. 
(b) Conduct studies of 3-D anisotrophy in CBP data. 
(c) Investigate software for full 3-D interpolation. Examples 
of options include: custom IDW software, custom kriging software 
using GMS routines, custom kriging software using the R-package, or 
some other off the shelf product. 
5. High Density Temporal Data 
2008-2010 
(a) Develop methods to use these data to improve temporal 
aspect of CFD in current implementation. 
(b) Investigate feasibility of 4-dimensional interpolation. 
appendix a 
The Cumulative Frequency Diagram Method for Determining Water Quality Attainment 
