Guide

Reading a tide table on this coast

Checked September 13, 2026 against NOAA Tides and Currents.

The Tides tile repeats NOAA’s predictions for the station nearest the town you are reading for. This page explains what the zero on that scale is, which stations on this coast are real instruments and which are arithmetic, and why the water is not always where the prediction says.

Zero is not the sea bed

Every tide height on this coast is measured from mean lower low water — MLLW — which NOAA defines as “the average of the lower low water height of each tidal day observed over the National Tidal Datum Epoch.” In plainer terms: of the two low tides most days bring, take the lower one, and average it over about nineteen years. That average is zero.

Because it is an average, roughly half of all lower low tides fall below it. A prediction of −1.5 ft is not an error; it is a tide a foot and a half lower than the long-run average low, and those are the tides a clam digger wants and a boat crossing a bar does not.

It also means a charted depth of 6 feet is 6 feet at MLLW. On a minus tide there is less water over it than the chart says.

Some stations measure. Some only calculate.

NOAA distinguishes stations that carry tidal harmonic constants from stations that do not. A harmonic station has the constants, and predictions “are generated directly from the harmonic constants.” A subordinate station has none, so its high and low tides are produced by predicting them at another station and then applying time and height corrections.

That is a distinction about arithmetic, and it is worth saying plainly that it is not the same as whether there is an instrument in the water. The Pacific County board reads nine stations, and only two of them have a gauge in the water: Toke Point and Cape Disappointment. The Grays Harbor County board reads five stations, and one of them has a gauge in the water: Westport. Three of those five — Point Grenville, Point Brown at Ocean Shores, and Cosmopolis — are subordinate stations that NOAA predicts from Toke Point, on Willapa Bay. Everything else on either board is a prediction, and a prediction cannot tell you the water is already over the road.

Five of the seven can tell you what the water is actually doing right now. Aberdeen and South Bend can only tell you what it was expected to do — and the expectation has had no instrument checking it for twenty years in one case and forty in the other. That is not a reason to distrust the prediction; it is a reason to know which kind of number you are looking at.

The north Olympic board reads ten stations, and four of them have a gauge in the water: Neah Bay and La Push and, at the east end, Port Angeles inside Ediz Hook and Port Townsend on Admiralty Inlet, which Port Hadlock reads five miles up the bay. Sekiu is a harmonic station with no gauge, as are Crescent Bay off Joyce and the Sequim Bay entrance, and Kalaloch and Queets read Destruction Island, six miles offshore, which is a subordinate station predicted from Toke Point in Willapa Bay, seventy miles south. It is the only prediction anywhere near that stretch of Park and reservation beach: there is no station at the Hoh, the Queets or Kalaloch itself.

The two newest are on Hood Canal, and neither has a gauge either. Quilcene, in Dabob Bay, is a harmonic station — NOAA carries its constants, so its predictions are worked out from the water there rather than borrowed. Pleasant Harbor, a mile south of Brinnon, is subordinate to Seattle, forty miles away over the Kitsap Peninsula, and is the only prediction near the Dosewallips. The canal runs a bigger tide than the Strait does — a range of ten feet and more at the head of it — and it arrives later, because the water has to come the whole length of Admiralty Inlet first.

The Puget Sound board reads sixty-five stations, six in Whatcom County, eight in the islands and Skagit, six in Island County, six in Snohomish, seven in King, ten in Kitsap, nine in Pierce, five in Thurston and eight in Mason, and two of them have a gauge in the water. In Whatcom it is Cherry Point, at the refinery pier four miles south of Birch Bay, which is what Birch Bay reads. Point Roberts, Blaine, on Semiahmoo Bay, and Village Point on Lummi Island are harmonic stations with no gauge; Sandy Point on Lummi Bay, which Ferndale reads, and Bellingham itself are subordinate to Port Townsend, forty miles down Rosario Strait. The tide up here arrives an hour or more after Port Townsend’s and runs bigger, because the water has to come round the islands and pile up at the end of the Strait of Georgia.

In the islands the gauge is Friday Harbor, the only water-level instrument in either county. Eastsound reads Rosario, on East Sound itself, and Lopez Island Upright Head at the island’s north end, both harmonic stations; Roche Harbor and Orcas are subordinate to Port Townsend. In Skagit, Anacortes and Guemes Island read the Guemes Channel station and La Conner and Mount Vernon the Swinomish Channel at La Conner, both subordinate; Edison, on Samish Bay, has no station of its own and reads the Swinomish Channel entrance on Padilla Bay, the nearest prediction. Mount Vernon is ten miles up the Skagit River from La Conner’s water, and the river at the town is not the table.

Island County has no gauge at all, and every one of its six is a prediction: Crescent Harbor for Oak Harbor, Coupeville on Penn Cove, Holly Farms Harbor in Holmes Harbor for Freeland, Sandy Point for Langley, Glendale for Clinton, and Greenbank across Saratoga Passage for Camano Island. Five of the six are subordinate stations of Seattle or Port Townsend; Greenbank is harmonic.

Snohomish County reads six, four of them harmonic: Everett at Port Gardner (which Mukilteo reads too, being the nearest to it), Edmonds, Tulalip Bay, Priest Point for Marysville, and Kayak Point on Port Susan for Stanwood — the Stillaguamish gauge at Stanwood is a river station, not a tide one.

King County reads seven. Seattle and West Seattle both read Seattle, at the foot of Madison Street on Elliott Bay, the harmonic station the whole of Puget Sound’s subordinates are figured from; NOAA meters it too, but the board reads its predictions. Shoreline has no station of its own and reads Edmonds, three miles up the same shore. Des Moines reads its own Des Moines, East Passage, a subordinate two hundred metres off the marina, and Burien and Federal Way read it too, being the nearest to both. Vashon reads Burton, in Quartermaster Harbor, the island’s own harbour.

Kitsap is the most closely gauged county on the board: every one of its ten towns has a station within five kilometres, and Bremerton, at the head of Sinclair Inlet, is harmonic — Port Orchard reads it too, two and a half kilometres along the same inlet. The rest read their own water: Eagle Harbor on Bainbridge, Liberty Bay at Poulsbo, Tracyton in Dyes Inlet for Silverdale, Appletree Cove at Kingston, Harper in Yukon Harbor for Southworth, Port Madison at Suquamish, Seabeck Bay on Hood Canal, and Hansville at the north tip.

Three stations — La Push, Westport and Toke Point — report wind. That is often the fastest honest answer to “what is it doing out there” on a coast where weather stations are thin.

On the Park's beaches, the tide is the trail

From Kalaloch north, the beach is Olympic National Park and the walking route goes round headlands that the sea covers. The Park does not soften it: “The tide can trap you! Do not attempt to round headlands without knowledge of the tide heights and times,” and “People have died along the Olympic wilderness coast trying to beat the tide.” Its advice is to round a headland “within one to two hours before low tide,” carrying a tide table and a map, and its route pages give a maximum tide height for each crossing — some as low as 2 feet, and Teahwhit Head and Taylor Point cannot be rounded at shore level at any tide. Overnight trips need a wilderness permit.

The Park’s own south-coast route page sends walkers to the La Push predictions, station 9442396, which is the station this board reads for La Push and Forks.

Why the water is not where the prediction said

NOAA is direct about the limits of what a tide table is:

On this coast all four of those are routine. A hard southwesterly piles water against the shore and holds a low tide higher than predicted. Low barometric pressure lets the sea stand higher. Winter rain down the Chehalis, the Willapa or the Columbia raises the level in the estuaries and can hold the ebb back. The published time of low water can arrive late, or the low can simply never get as low as the table says.

This matters most to the two activities this coast does on a minus tide: crossing a bar, and digging clams in the dark. If the tide is running an hour behind and a foot high, both of those plans were made on a number that is no longer true. Where a station has a live gauge, the observed level is the answer; where it does not, assume the prediction is the beginning of the question.

What NOAA says about relying on it

NOAA’s own disclaimer is worth reading once: real-time data “have not been subjected to the National Ocean Service’s quality control or quality assurance procedures,” and “the user assumes the entire risk related to its use of these data.” Its model-based water level forecasts are “accurate to within 6 inches 90% of the time” — which is to say that one time in ten they are not.

None of that makes the tide table unreliable for planning. It makes it a prediction, which is a different thing from a measurement, and the board labels which one it is showing you and when it was taken.

Sources

• Olympic National Park, hiking the wilderness coast and tides and your safety — the headland rule and the tide heights
• NOAA Tides and Currents, tidal datums — the definition of MLLW
• NOAA Tides and Currents, help for tide predictions — harmonic and subordinate stations in NOAA’s own words
• NOAA Tides and Currents, frequently asked questions — what a prediction can and cannot include
• NOAA Tides and Currents, disclaimers
• Station pages: Neah Bay, La Push, Westport, Toke Point, Cape Disappointment, Aberdeen, South Bend