Engineering design when the decision making of animals dictates success.
High–resolution engineering models can create an accurate 'virtual reality' of the hydraulic and
water quality environment associated with design or operational alternatives of a dam. A fish swim path selection algorithm, based on cognitive ecology, added to a particle–tracking model (that
forecasts passive transport of neutrally–buoyant particles) creates a “smart particle” or Numerical Fish Surrogate (NFS). The NFS is based on the Eulerian–Lagrangian–agent Method (ELAM). The NFS can forecast fish x, y, z, time,
swim speed, and swim orientation information in 2– or 3–dimensions. The underlying “traffic rule” used by fish to navigate complex flow and water quality fields can be discovered by
recursively reducing differences between forecasted paths and field observations.
ELAM simulated fish vs. passive particles (flowlines) approaching a dam.
Movement of Numerical Fish Surrogate (NFS) virtual fish [top] and passive particles [bottom] approaching Lower Granite Dam (Snake River, Washington USA) in virtual reality. Note the distribution of behavior within the fish population changes as the fish approach the dam. Upstream of the dam the default behavior B0 dominates, but the other behaviors become more important as the hydrodynamic environment becomes more complex near the dam.
Cognitive analysis of complex 3–D fish movement behavior.
Project features at Lower Granite Dam on the Snake River, WA USA for 2000 studies include a Behavioral Guidance Structure (BGS) intended to guide fish to the Surface Bypass Collector (SBC) and occlude them from the 3 turbine intakes nearest the shore. The movement of an acoustically–tagged fish (real fish track) is colored green in the upper–left plot, provided by the U.S. Geological Survey.
Population passage is the sum of the behaviors of individuals and cannot be accurately forecast unless the behavior of individuals is captured. In the first example [C], the NFS duplicates the pattern of a real fish following the trash boom without visual cues by integrating information between the total hydraulic strain and velocity fields (see below). At time t = 2120 sec the virtual fish approaching the trash boom perceives total hydraulic strain exceeding threshold k1 identifying the environmental agent as friction resistance (environmental agent A1). Agent identification triggers a sequence of events increasing the expected utility U1 (the motivational value of response B1 to stimulus agent A1). A short time later U1 exceeds utility U0 (the motivational value of the default behavior B0). Only when U1 exceeds U0 in [C] at t = 2128 sec does the behavior switch from B0 to B1. The delay between stimulus identification and observed response is the ‘response latency’, an important feature of fish behavior.
In the second example [D], the NFS duplicates the pattern of real fish milling between the Surface Bypass Collector (SBC) and trash boom using only hydrodynamic cues. In the high–energy
hydrodynamics near the SBC (see below) the virtual fish perceives total hydraulic strain exceeding k2 at t = 2928 sec in [D] (strain plot) signaling environmental agent A2. After a 50 sec latency, utility U2 exceeds U1 and the virtual fish switches to behavior B2 leading
the fish upstream. Even though total hydraulic strain diminishes as it moves away from the SBC [D] (strain plot), the fish continues swimming upstream because of response latency. Perceived total hydraulic strain begins to drop significantly after the fish moves upstream of the boom just prior to t = 3200 sec and utility U2 of behavior B2 drops correspondingly. While the fish may repeatedly journey between the SBC and trash boom, the
intensity of total hydraulic strain to which the fish is acclimated, Ia(t),
increases with each cycle and eventually the perceived total hydraulic strain at the SBC entrance does not exceed k2. At that time, the fish would either swim with the flow (B0) or into increasing water velocity (B1), both of which
would cause the fish to enter the bypass channel. The boom’s impact on the flow field dissipates with depth and, correspondingly, neither deeper swimming virtual fish nor deeper swimming tagged fish show a response (Goodwin et al., 2006).
3–D hydraulic pattern at Lower Granite Dam.
CFD model depiction of the patterns in velocity magnitude (m/s) and total hydraulic strain (s–1) for a Lower Granite Dam SBC forebay configuration in 2000. Dam operations and other information described in detail in Goodwin et al. (2006).
3–D structured mesh CFD model.
CFD model accuracy depends intimately on the resolution provided by the
underlying mesh (upon which hydraulic information is calculated by the
CFD model).
Example to the right illustrates tessellation of a hydropower dam
forebay into a structured hexahedral (8–node) element mesh. The
mesh is composed of many blocks, represented here with different
adjoining colors. Mesh elements are essentially distorted bricks
and “near–orthogonal” meaning that each corner of the element is as
close to 90° as possible.
3–D unstructured mesh CFD model.
Tessellation of a hydropower dam forebay into an unstructured tetrahedral (4–node) element mesh.
Refereed Conference Proceedings Goodwin, R. A., Nestler, J. M., Anderson, J. J., and Cheng, J.–R. (2007). “Understanding hydrodynamics from the fish’s point of view, Part I: Integrating CFD modeling, individual movement, and spatial/cognitive ecology.” Proceedings of the 6th International Symposium on Ecohydraulics, 18–23 February 2007, Christchurch, New Zealand.
Nestler, J. M., Goodwin, R. A., Anderson, J. J., and Smith, D. L. (2007). “Understanding hydrodynamics from the fish’s point of view, Part II: Integrating flow field distortion, sensory biology, and geomorphology.” Proceedings of the 6th International Symposium on Ecohydraulics, 18–23 February 2007, Christchurch, New Zealand.
Goodwin, R. A., Nestler, J. M., Anderson, J. J., and Weber, L. J. (2004). “Virtual fish to evaluate bypass structures for endangered species.” Proceedings of the 5th International Symposium on Ecohydraulics, 12–17 September 2004, Madrid, Spain.
Goodwin, R. A., Nestler, J. M., Anderson, J. J., and Weber, L. J. (2004). “Forecast simulations of 3–D fish response to hydraulic structures.” Proceedings of the World Water & Environmental Resources Congress, American Society of Civil Engineers, 27 June – 1 July 2004, Salt Lake City, Utah.
Goodwin, R. A., Anderson, J. J., and Nestler, J. M. (2004). “Decoding 3–D movement patterns of fish in response to hydrodynamics and water quality for forecast simulation.” Proceedings of the 6th International Conference on Hydroinformatics 2004, Liong, Phoon, and Babovic, eds., World Scientific Publishing Company, 21–24 June 2004, Singapore.
Goodwin, R. A., Nestler, J. M., Weber, L., Lai, Y. G., and Loucks, D. P. (2001). “Ecologically sensitive hydraulic design for rivers: lessons learned in coupled modeling for improved fish passage.” Proceedings of ASCE Specialty Conference on Wetlands Engineering & River Restoration 2001, 25–31 August 2001, Reno, Nevada.
Goodwin, R. A. and Nestler, J. M. (2000). “Coupled Eulerian–Lagrangian hybrid (CEL hybrid) ecological modeling and hydroinformatics.” Proceedings of 4th International Conference on Hydroinformatics, 23–27 July 2000, Cedar Rapids, Iowa.
