Wrap-up Comments on the SFRSM Theory Manual

A.   ON THE ORGANIZATION OF THE MANUAL

  1. The main body of the manual consists of 56 pages. The remainder consists of Appendices A, B, and C. In particular, Appendix C consists of six (6) documents, the first four of which are published (or to be published) papers. I believe Appendices C.5 and C.6 contain information which should be part of the main body of the manual. It is okay to place published work in the appendix, but unpublished work, particularly if it relates directly to the subject matter, should be placed within the main body. This may require a major restructuring of the manual chapters.

  2. Published papers to be placed in an appendix (in this case, C.1 to C.4) should be in the original, published form. The proper permissions should be secured from the publishers.
B.   ON THE USAGE OF THE LANGUAGE

  1. The manual has extensive problems with hyphenation and several spelling and grammatical errors. I recommend having the manual edited by a technical writer or someone who has a high level of knowledge in the formal use of the English language.
C.   ON THE TECHNICAL CONTENT

  1. Avoid jumping over details of equations. If the manual is to be used by practitioners (consultants and others), they need to be able to see the various steps leading to the solution, within reason, of course.

  2. Need to establish a better link between the traditional equations (the differential equations of Appendix B) and the equations used in the OO model (look-up tables, regression). Are the latter based on the former? If not, how is the relevancy of the traditional equations justified?

  3. Need to be consistent on the system of units. Appendix C.5 contains SI units, while the Fact Sheet states that "all input and output data will be created in English units"

  4. The so-called "diffusion equations" calculate hydrograph diffusion, in either 1-D or 2-D. True (physical) hydrograph diffusion can only be produced by an unsteady loop in the rating curve. Disregarding the loop by using a static look-up table renders the simulation kinematic, i.e., not subject to physical diffusion. Then, any hydrograph diffusion represented in the simulation would necessarily be a function of the grid size. Please explain how extensive is the use of look-up tables in the model, and what is the effect, if any, on the calculated hydrograph diffusion.

  5. How was the threshold value δ in Eq. B.16 determined? How often is it reached? What does the model do when the threshold value is reached?

  6. The model uses the NRCS curve number method as the infiltration model. However, the latter is strictly applicable only to event (short-term) modeling. In practice, the AMC feature of the curve number method helps it account for the natural variability of infiltration response. There is no such thing as a fixed "curve number," or a constant "maximum potential retention (S)." Thus, a curve number obtained through calibration may not be applicable in the validation phase, unless the two events being used (for calibration and validation) happen to have similar AMC characteristics. This is a tough problem, and one which not many people are fully aware of.

  7. Need to better explain the determination of the Manning friction coefficient under various vegetative and other terrain (land use) conditions. If the Manning value is going to be large (greater than 0.3), it is probably out of the fully-developed, turbulent-flow regime already, and may be in the mixed laminar-turbulent regime. In this case, it is more appropriate to refer to the friction coefficient as the "equivalent Manning roughness." The latter is sometimes denoted as N to indicate that it is not the fully-developed, turbulent-flow value. What is the model's sensitivity to the chosen value of Manning friction?