ShellSanibel
Read the beach.
How the beach read is made
Anyone can print a rating. What makes one worth anything is being able to see what went into it and where it stops working. This page is both.
The honest framing, first. Conditions can improve or reduce the likelihood of productive shelling, but the Gulf decides what reaches the beach. Nothing on this site can promise you a shell.
Where the numbers come from
-
Tide predictions
NOAA CO-OPS, St. Petersburg harmonic station
Shifted to each beach using NOAA’s published subordinate-station time and height offsets. Every beach page shows which offsets it used.
What it cannot tell youA harmonic prediction is astronomy. It knows where the moon is; it does not know what the weather has been doing, which is what the correction below is for.
-
Wind correction on the tide
NOAA CO-OPS, Fort Myers gauge — the same gauge the water level is measured at
Wind blowing along the coast piles water against it or draws it off, and a harmonic prediction cannot know that. We take the average alongshore wind over the twelve hours before each tide, subtract what that month normally does, and shift the predicted height by roughly 0.06 to 0.09 ft for every knot of the difference.
What it cannot tell youThe relationship is measured at a gauge up the estuary, because there is no water-level gauge on the Gulf beach itself — so the size of the correction is a range and not a single number. Wind we have not observed yet gets no correction at all rather than a guess, which is why tomorrow morning is not adjusted until tomorrow morning.
-
Red tide
Florida Fish and Wildlife Conservation Commission
FWC’s own abundance band for recent samples near the islands — we publish their category, never a number we derived ourselves.
What it cannot tell youSampling is periodic and patchy. "No recent samples" is not the same as clear water, and the page says so rather than implying safety.
-
Wind
The National Weather Service, at the island coordinates
Speed in mph and direction, used to judge whether anything is being driven ashore.
What it cannot tell youA point forecast. The beach you are standing on can differ, especially near the passes.
-
Sun and twilight
Computed locally with suncalc
Civil dawn and civil dusk bound the walkable day — shellers start at first light, not at sunrise.
What it cannot tell youAstronomy is the one input here that is genuinely exact.
What actually moves the read
These are the only things that do. There is no other input, no editorial thumb on the scale, and no number we invented.
- How low the water gets in daylight The largest single influence.
- A daylight low at or below 0.3 ft counts as excellent and moves the read two steps; at or below 0.8 ft, one step. Higher than that and comparatively little bar is uncovered. Those cuts are where the water actually sits, which is not the same as what a tide table says — see the row below.
- Why our grade can disagree with the height we print Shifts the read, never the number.
- NOAA's tide predictions are referenced to a tidal datum averaged over 1983–2001, and mean sea level has risen since. At the St. Petersburg gauge these predictions come from, the water has run about 0.38 ft above its own prediction over the last nineteen months, and the Fort Myers gauge agrees within three hundredths of a foot. Every tide table and phone app inherits this; it is not peculiar to us. So we print NOAA's number, because that is the one on your phone and in the printed tables, and we grade against where the water is actually going to be. In tide-table terms the cuts land at -0.08 ft and 0.42 ft. Graded on the raw prediction instead, more than a third of daylight lows here came out excellent and nearly two thirds good or better, which is not a rating so much as a yes.
- Whether that low lands near first light Improves the read.
- A low within 90 minutes of civil dawn is the premium combination: lowest water, low sun, and a beach nobody has walked yet.
- Wind direction Improves or worsens it.
- Onshore wind can drive material in, and how onshore a wind is depends on which beach you are standing on: these seven Gulf beaches face across about 135° of compass, from the island’s southeast tip round to west-northwest at Captiva. We measure the part of the wind aimed straight at that beach, so the same breeze can be onshore at Captiva and off the land at Point Ybel. Each beach page shows the direction it faces. Wind off a beach moves its read down, because nothing new is arriving there.
- Wind strength Improves it, then ruins it.
- Onshore wind counts from 9 mph and counts strongly from 17 mph. At 29 mph and above the water is churned and the wash is hard to read whatever the tide is doing, so the read drops.
- Red tide Caps the read outright.
- Medium or high Karenia brevis near the islands caps the read at Thin regardless of a perfect tide, and leads the reasoning. Respiratory irritation and dead fish decide a morning in a way that a good low water does not offset. Low abundance moves it down one step.
- Darkness Overrides everything.
- A low that falls outside civil twilight is not walkable. If both of a day’s lows are in the dark, the read is Thin and no window is offered, however deep the water drops.
Why there is no score out of ten
The read is one of five words — Promising, Worth a walk, Mixed, Thin, Too uncertain to call — and never a number. A 7.6 out of 10 would imply a precision that wind, tide and a Gulf doing as it likes cannot support. The reasoning underneath is the part worth reading; the label is only an index into it.
What happens when a source fails
Each block is checked for freshness on its own and shown as Unavailable rather than carried forward. Tides count as stale after 24 hours, red tide after nine days — FWC samples periodically rather than daily — and wind after 3 hours. A missing input never silently improves the read — it lowers the stated confidence, and the page says which input was missing. If the page itself is more than 18 hours old, a banner appears at the top saying so.
We would rather show you nothing than a stale red tide reading. A wrong number on that block is a health question, not a cosmetic one.
Update timing and what is automated
These are static pages. Nothing is fetched while you are reading, so the page you have is complete and unchanging until the next build — which is why the timestamp matters. The schedule runs three times a day: before first light, at midday, and in the evening after FWC posts its daily sample map.
Believe the timestamp in the footer over that last sentence. It records when this page was actually built. A schedule is an intention and can fail; the timestamp is a fact, and if it is more than 18 hours old a banner at the top of the page will say so before you read anything else.
The interpretation is code, not a person: the same inputs always produce the same read. The beach descriptions and the shell guide are written by hand. Scientific names are checked automatically against the World Register of Marine Species.
Limits, stated plainly
- Conditions can improve or reduce the likelihood of productive shelling, but the Gulf decides what actually reaches the beach. Finds are never guaranteed, and a Promising morning can produce nothing.
- The read describes the conditions, not the sand. It cannot know that a hundred people walked the same stretch an hour before you, and on these islands that is often the deciding factor.
- Tide heights are predictions. Sustained wind and pressure move real water away from them, sometimes by a foot.
- Red tide is reported from periodic samples at fixed points. Conditions can change between samples and can differ sharply within a few miles of coast.
- The read is built from four inputs. If more than 1 is missing, or if tides or sun times are missing at all, it refuses to call it and says "Too uncertain to call" instead of guessing.
- This is not a safety service. It is not for navigation, and it does not tell you whether the water is safe to enter. Check official sources for that.