My interest in computational fluid dynamics (CFD) models has been motivated by three factors:

• To better understand the influence of channel morphology on hydraulic processes
• To infer geomorphic processes from hydraulic processes (e.g. pool-riffle maintenance mechanisms research)
• To assess physical habitat quality for fish (e.g. ecohydraulic modelling, spawning habitat rehabilitation, habitat heterogeneity research)

To date, I have primarily used a two-dimensional (2D) code called FESWMS and marketed by BOSS as SMS (Surface Water Modelling System). In general, with good topographic boundary conditions we can produce very reasonable simulations over short temporal scales and over small reach scales (e.g. figure at right). CFD modelling is very time consuming and computationally demanding, but is very helpful for gaining insight into research questions at sub-reach scales. However, 2D and 3D CFD models of meaningful spatial resolutions are impractical at catchment scales or over longer time-scales. Furthermore, with a few notable exceptions (CFD codes with sediment transport capability for sand-bedded rivers), most CFD codes are impractical to run with a dynamic bed boundary condition. These limitations have led me to explore a reduced-complexity class of models (cellular automaton).

---

Example CFD Simulation. A) Portion of a 2D CFD simulation of flow around and over boulder complexes in a spawning habitat enhancement site on the lower Mokelumne River. The intensity of blue is scaled to the water depth solution (dark blue is deep); the arrows represent depth-averaged flow directions with the color scaling to the velocity magnitude (red is slow, green and blue are fast). Notice the boulder complex in the upper right where the model effectively captures flow around the boulder, and the hydraulic jump over the submerged boulder complex. B) A photograph of the boulder complex in the upper right of figure A at the same flow as the simulation (approx 6 cumecs) ©2002 Wheaton (See Photo Copyright Disclaimer before downloading).

---

CASE STUDIES

• Mokelumne River, California USA
• Dry Creek, Winters, California USA

Publications using CFD Tools

• MacWilliams, M. L. Jr., Wheaton, J.M., Pasternack, G.B., Street, R. L. and Kitanidis, P. K. , 2006. Flow convergence and routing hypothesis for pool-riffle maintenance in alluvial rivers, Water Resources Research, 42, W10427.
   
• Pasternack, G.B., Gilbert, A.T., Wheaton, J.M. and Buckland, E.M., (In Press). Error Propagation for Velocity and Shear Stress Prediction: Using 2D Models For Environmental Management. Journal of Hydrology.
   
• Wheaton, J. M., Pasternack, G. B., & Merz, J. E. 2004. Spawning Habitat Rehabilitation - II. Using Hypothesis Testing and Development in Design, Mokelumne River, California, U.S.A., International Journal of River Basin Management, 2(1): 21-37.