How the fishability model works

Green means the model sees a credible opportunity and the spot should be fishable enough to act on.

Yellow means there is still a path, but the spot is conditional or requires accepting tradeoffs.

Red means either the setup is not turning on, the execution cost is too high, or both.

• NOAA current stations are still the main anchor for tidal movement where a nearby station exists
• SFBOFS now adds modeled bay-flow guidance for current speed, eddy behavior, and spots with weaker station coverage
• NWS hourly wind forecast now anchors the broader bay wind read by zone, with gusts carried through when available
• wind direction is interpreted against each spot's modeled exposure, so leeward/protected spots are treated differently from windward or open edges
• NWS sky-cover forecast now anchors the cloud read and 7-day outlook context
• NOAA and USGS station data now anchor water temperature where possible
• USGS turbidity now anchors the clarity heuristic by zone when a nearby station exists

• clarity is inferred from rain, wind, exposure, zone, and resilience of the spot unless a live turbidity anchor is available
• bay spot fit is inferred from local structure type, preferred tide direction, current dependency, shoreline-pinning behavior, and light preference
• beach spot fit is weighted primarily around wind, tide push, and swell window, with swell direction used as a beach-orientation check for how that energy enters the surf zone
• the final score is an opinionated fishing recommendation, not a direct measurement
• water temperature is treated mainly as seasonal context; normal Bay-range differences have only a small score effect, while unusual cold or warm water receives a modest caution
• barometric pressure is included as a bay-wide weekly context read, not a spot-ranking driver
• moon phase is included as a daily context callout for spring and neap tide windows, not as a dominant bite predictor

• Bay current direction and speed are primary only where the mapped structure depends on a seam, point, shoreline pin, or exchange edge
• current effects are capped by spot dependency so overlapping current rules cannot repeatedly penalize the same weak or slack condition
• beaches use tide level, wind, and swell rather than inheriting Bay-current scoring; river mouths retain a modest exchange-current effect
• wind is a major practical limiter when exposure makes casting, presentation, wading, or boat control unreliable
• direct turbidity can materially affect clarity scoring, while rain-and-wind-only clarity estimates stay deliberately muted
• light and water temperature are supporting context, not dominant hour-to-hour ranking variables
• pressure and moon phase remain planning context and do not directly alter spot scores

• spot data and scoring rules are now split out of the homepage into reusable model files
• wind is now driven by NWS hourly zone forecasts instead of a paid third-party map dependency, and gusts now influence casting difficulty
• SFBOFS modeled flow guidance now supports current speed reads when station coverage is weak or local flow gets weird
• the homepage now clearly distinguishes trusted source data from inferred values
• directional shoreline spots now get extra credit when the tide pins bait into the modeled edge, and a stronger penalty when the tide pulls away from that setup
• coastal spots now distinguish sandbar/trough beaches, river mouths, pocket beaches, rocky edges, and pier-adjacent beach water when scoring swell windows
• swell direction now has a real score effect for beach spots because wave angle changes exposure, longshore current, and rip/edge formation
• beach scores now rebalance the model so wind, tide, and swell are the dominant drivers, while clarity, light, and temperature stay as secondary context
• water-temperature scoring is now deliberately low-weight because research supports it more strongly as a migration and seasonal-habitat signal than as an hour-to-hour local bite predictor
• Bay current contributions are now bounded by spot dependency, preventing direction, seam, current-window, and slack rules from stacking without limit
• standard beaches now score normalized tide level instead of Bay current direction; this better reflects how water depth changes bar, trough, rip, and access geometry
• clarity tolerance now controls how strongly dirty water affects each spot, while heuristic clarity remains lower-confidence than direct turbidity
• a 7-day planning window now shows cloud, rain, wind, pressure, moon, and temperature context for the selected spot
• the map now has a restrained zone-level wind overlay so wind supports the fishing read without turning the app into a weather site
• the detail panel still keeps the condition breakdown and source story visible for the selected spot

• 7-day calendar views and future-window planning
• fly recommendations by spot, condition, and likely water color
• per-location or zone-based temperature differentiation
• better tide translation by zone
• deeper clarity logic by zone and exchange
• real-time beach-bar, trough, rip, and water-color observations for coastal spots