THEN: Local Knowledge ("LK") developed its first current model for San Francisco Bay in 1996-1997, with the intention to provide racers with a continuous, all points, PC program that could support tactical computations and compute current maps of predicted currents. At that time, the only alternative was NOAA documentation of current at discrete points, too widely and randomly scattered to provide much guidance at the actual location of your boat. The LK model took the available NOAA data, supplemented by data taken while racing on large boats such as Sayonara and Zephyrus, and used it as input to develop a hydrodynamic model of the entire Bay. At that time, much of the NOAA data was decades old and untested. Fortunately, the PORTS system was underway and began to provide continuous monitoring of current strength and direction at several new locations in the Central Bay, supplementing the historic NOAA data and enabling a test of the LK model by comparing its predictions to the PORTS measurements. These comparisons validated both the general accuracy of the underlying data and the methodology of the LK model. That said, some parts of the Bay are better determined by the historic data than others, and the model was supplemented in subsequent years with new data supplied by racing clients. The associated software, marketed in 1996 as "Force 2," also became more powerful, adding new tactical functions and evolving to "Force 3" and "Force 4" by 2002.
The new LK model is implemented by Advantage Racing Software, a vast improvement over the original "Force" programs. Advantage has been optimized for the newer PC operating systems, and is much faster, more stable and more powerful than the earlier programs, including many more (and better quality) charts, more sophisticated chart handing and a broader range of display options and tactical functions. Advantage is installed as a single program that can be licensed to a user at one of three "Levels," depending on intended use (and budget). Level 1 is sufficient for users who don't intend to use a PC "live" on the boat, relying instead on pre-race studies of predicted currents and printouts; Level 2 adds primary buoy racing functions and offshore routing capability; and Level 3 an additional suite of unique starting line and buoy racing functions that are available nowhere else in the industry. [For more details about Advantage, use the link at the bottom of the page]

NOW: In 2004, NOAA began to completely revise its current predictions for San Francisco Bay, adding new Primary Reference Stations inside the Golden Gate Bridge, south of Treasure Island, and south of Red Rock, modifying the timing predictions of most Secondary Stations to reference the new Primary Stations, and adding a number of new Secondary Stations. And, in the past year or so, new current sensors are operating in the Central Bay. In response to this wealth of new information we have completely reformulated the Local Knowledge current model for 2011. The new model is more detailed, more accurate and incorporates a much better description of the shorelines (important for routing computations). Like the original Force 2 program, it will also enable a user to "tune" the program in the spring run-off season using online data from the new current sensors.
New 2011 Current Model for San Francisco Bay

Advantage offers a variety of options for displaying current distributions. There are three "vector" displays, in which the length of the streamer, and size of the arrow or chevron, is proportional to the strength of the current, and a non-vector display in which a digital value for current speed is shown in a nearby box, and streamer shows direction. Vectors are chosen automatically by the program at a grid of points covering the chart on view (the closer you zoom in, the more detail is shown). In the non-vector display, the user defines the display points at whatever locations desired.
Current Streamers
Current Chevrons
Non-Vector Display

The new LK model covers the entire San Francisco Bay region, south to the San Mateo Bridge, north to include San Pablo Bay and east to Suisun Bay, and west through the Golden Gate beyond the SF entrance buoy. The image below shows flood current in the Central Bay. Advantage comes with model coverage throughout the calendar year, plus several prior years. Coverage in subsequent years is available for an annual renewal fee of $100, which also includes the latest version of the program and auxiliary programs. You can run Advantage in real time, at the date and time of your choice, or speed up time evolution to watch currents change in a dynamic display.

Strong non-uniform currents result in curved laylines, as shown in the image below. The solid blue lines emanating from Mark 5 show the laylines in the absence of current, and the magenta dots show the current-corrected laylines (a boat sailing at its optimal tacking angle will follow the curved path to the mark). Likewise, the path followed by the boat to the layline is curved (the orange and green lines show the predicted track on port and starboard tack, respectively). The layline function is available even in Advantage Level 1, and can be used to explore different wind and mark scenarios in advance of a race. For users with a PC onboard, more powerful functions are available that determine the optimal sequence of tacks to the mark and predict the time to reach either layline (see Advantage link below for more details and examples).

During the 2010 Big Boat Series, a company called Tidetech announced a new current model for San Francisco Bay with a lot of fanfare. We have analyzed the sample current grib file distributed by Tidetech for September 7, 2010 and found numerous errors. Most critically, compared to established NOAA data, Tidetech systematically underestimates current magnitudes in the Bay by 20-25% (and by more in some locations), and also exhibits significant differences in current direction and timing. For a detailed comparison with NOAA and LK, use the link below.
TT Comparisons: NOAA Locations
TT Comparisons: Distribution
TT Comparisons: Timing