Closure Section in Final Report of the Peer Review Panel Contribution by Victor M. Ponce, Panelist August 29, 2006 Numbers refer to Issue ID in SFWMD Response. This is my contribution to these Issue ID numbers. Other issue ID numbers not shown are intentionally left for comment by other Panel members. 10. The Panel commends the District for making the effort to address individually, in a timely fashion, the concerns addressed in the Panel's Draft Report. 11. The District has responded to the Panel's concern that inadequate material was provided. The RSM documentation set proposed in the District's Response should be adequate to meet District needs. 12. The Panel is assured that typos and other minor errors will be corrected. 16. The District's response is satisfactory. The term "effective roughness parameter" is a better descriptor of the physical process than the conventional "Manning's n." 19. The Panel is satisfied by the District's argument that physical constraints will generally prevent misuse of the diffusion wave approach. The Panel recommends that the District engage in theoretical work to establish 2-D limits for the diffusion wave approach. 21. Regarding the proper value of the weighting factor alpha, the Panel reasserts that alpha = 0.5 is theoretically second-order accurate, while alpha = 1 is only first-order accurate. While the tacit use of alpha = 1 is based on issues of practicality, a good model should always give the user some guidance as to the trade-offs involved (stability vs accuracy). The Panel agrees with the District staff that this problem is not unique to the RSM, and it is assured that the District is making significant strides to achieve better modeling by candidly addressing the alpha issue in its publications. 22. The Panel retreats from a previous statement characterizing n = 1 as "unrealistically high." If data justifies it, so be it. At any rate, at this high range of n, it is virtually impossible to assure the applicability of the extension to laminar flow. The Panel reiterates that the decision to use alpha = 1 may be a practical one, but that it needs to be justified or explained in a more scientific way. For instance, as alpha increases, show the comparison between the attenuation of high-frequency perturbations (noise) with the degradation of the low-frequency perturbation (signal). 24. The Panel is satisfied that the runoff curve number method is being used in a continuous-simulation mode by adjusting the value of maximum potential retention (S) based on the available soil-moisture storage. While the method's developers did not intend it to be used outside of event modeling (see http://mockus.sdsu.edu), it is correct to state that the method has been extended, by default and by way of practice, to the continuous simulation arena. The key issue in here is to do it carefully and transparently. 25. The Panel is satisfied with the District's decision to calculate PET outside of the RSM, in order to provide consistent inputs for comparison with older models. In the future, the calculation of PET inside the model would be an enhancement. 26. The Panel is satisfied that wind stresses are of secondary importance in most cases, and adheres to the District's argument that model experience will eventually dictate whether this refinement is necessary. 29. The Panel is satisfied with the District's decision to characterize the functioning of the RSM as a "static physical system" in the "Scope and Purpose" section of the Theory Manual. Later model releases may opt to consider the succession character of managed or natural systems. 30. The Panel is satisfied that the District is fully aware of the tradeoffs involved in diffusion wave modeling, and that, if necessary, a dynamic wave approach will be developed. 32. The Panel is satisfied that rain and ET will eventually be included in the canal water balance, leading to a slight improvement in modeling accuracy. 39. The Panel is satisfied that the District will make every effort to improve the structure of the Theory Manual. The proposed modifications are significant, and should improve the readability and usefulness of the Manual. 49. The Panel welcomes the District's decision to add a volume entitled "Guidelines for Managing Numerical Error" to the documentation set. 54. The Panel is reassured that the District is committed to hiring a competent technical editor to resolve problems with language, grammar, and consistency of scientific terminology. 55. The Panel is reassured that the District will make every effort to distinguish between the three types of errors which arise in mathematical modeling. First, numerical errors should be minimized; second, physical errors should be investigated, identified, and corrected; and third, data-quality errors should be acknowledged and, to the extent possible, resolved. As the District has adroitly recognized, disregarding this triad results in bad modeling practice. 56. The Panel is satisfied that the District will constrain model parameters to physically realistic ranges during automatic calibration. 57. The Panel is reassured that the District will implement a three-stage parameter calibration to better simulate the nonlinearities inherent in the physical process. Appendix C: a. I agree with the District that the wording "steeper learning curve" does have a negative connotation. b. I believe the Panel's intent was to emphasize the difference between "reference" and "user" manuals. The existing "Theory Manual" would come under the category of "reference." The proposed RSM Documentation Set (Table 3 of District's Response) is now appropriate.

050828