Painting a Better Picture
Painting a Better Picture
A new service aims to deliver highly accurate tidal data in popular racing venues in North America. Electronics Tech Review from our July/August 2011 issue.
Wouldn’t it be great if figuring out the current on the racecourse was as simple as knowing whether it’s ebbing or flooding, and how strong it’s running? That’s wishful thinking, especially considering everything that can influence tidal flow on your particular racecourse: underwater contours, atmospheric pressure, rainfall, and lunar cycles, just to name a few. Until recently, the only way most of us have been able to get our heads around the current was by using vector charts and local knowledge. As it is with wind forecasts, however, what we really want is the devil found only in the details—what’s really going on underneath our keels—in near real time. Fortunately, a new company providing subscription-based current-flow models is bringing us closer to that reality.
TideTech, an online subscription-based service (www.tidetech.org), taps into available oceanographic data (some publicly available, some under license) and runs it through its proprietary model to deliver impressively detailed current predictions—they presently have accuracy down to 100 meters, and are pushing for even better resolution where possible.
How they come up with the end product (show in the diagram above as a downloadable PDF) requires explanation. Fifty percent of the TideTech model is gathering the existing data, says co-founder Penny Haire (who is partners in the venture with Dr. Roger Proctor and Dr. Roger Flather, both of whom specialize in hydrodynamic modeling): “What we do is undergo a process where, for example, to develop a model for a popular racing venue, we’d get depth information from surveys and charts—as detailed as possible. What we then need is something called, ‘boundary conditions’—that’s the tidal regime in the area. Let’s say [Rhode Island’s] Narragansett Bay is the region; we then divide it into lots of small boxes, to form a grid. The more boxes we create, the better the resolution, and the longer it takes the model to run, so we determine what level of resolution the customer really needs—say 100 meters. The algorithms we’ve developed then allow us to simulate what the tidal streams and tides do in each of those little boxes.
“Imagine, in a simplistic way, that we know the tide will be higher in one side of the model area than the other. We can simulate what the flow will do between one side of the model and the other. So each of those little boxes is calculated as to how the flow moves between each box, and that’s determined by the tidal height, the time, and the depth as well. It’s an equation of motion, essentially.
“Once we have the basic model, we calibrate it by picking an individual point in the model—where there’s an actual tide gauge, for example—and making sure the model corresponds with the actual tidal height rise and fall at that point. If we’ve got that bit correct, the whole thing is correct. We can take measurements to further refine it.”




