The Vagaries of Tuamotus Tides Or In Praise of Art

On board references used:

1.  Polynesia, Charlie’s Charts, 7th Ed

2.  The 2016 Tidal Guestimator, S/V Visions of Johanna

3.  Sailing Directions – Pacific Islands (Enroute), US Publication No 126, 2014

4.  Ocean Passages and Landfalls – Cruising Routes of the World, Imray, 2nd Ed

5.  The Pacific Crossing Guide, RCC Pilotage Foundation with the OCC, 2nd Ed

6.  South Pacific Anchorages, Warwick Clay, Imray, 2nd Ed

7.  The Tuamotus Compendium, S/V Soggy Paws

8.  Pacific Island Pilot Vol III, British Admiralty NP 62 ,9th Ed, 1969


This started as a quick bit of analysis for my own benefit after our entrance experience to Kauehi, listening to others talking about the difficulties of estimating slack around the atolls and the use of the Tidal Guestimator. It took on a life of its own. The more I read and looked up, the more I became sure that I wasn’t going to find how I could calculate a definitive answer. What I do now have is a far better understanding of why working out when slack tide is in the Tuamotus will always be an imperfect exercise, what the factors involved are and how I can get closest to an accurate answer. I know friends who sat scratching their heads for more than seven hours after what they thought would be slack water, watching the water rush out before finally being granted those elusive few minutes of quiet water to pass safely in to an atoll. I hope this missive will help them understand the circumstances that may have caused their delay and confusion. SH Sep 16.


All those that sail the coastal regions of the oceans are used to dealing with tides. Every few hours the tide turns and the sea runs in or out, generally by the rule of twelfths. Around the UK, we have wonderfully accurate tables and a full list of primary and secondary ports with nice graphs for tidal curves to work out what the tide will be doing to us at any particular moment in time. Our RYA test papers ask us to the minute when we will be able to cross the bar to enter harbour, which, with the accuracy of the data available to us, we can smugly do. Springs and neaps do complicate matters but our data deals with those.

There are a few tidal gates around the UK coast that we know that we have to hit if we want a quick and easy passage. Missing most of these gates will but inconvenience us a little, although those with West coast of Scotland experience might hold a stronger opinion!  I have found in conversations with other international cruisers that UK sailors are far less fearful of dealing with tidal problems, perhaps simply because we are exposed to more frequent and greater extremes than most. As prudent sailors, we are in the habit of looking at the tide tables before we leave harbour. But we can still be caught out. I remember fondly a trip with my father many years ago, delivering a boat to Ardfern, Scotland and ghosting along in very little wind,  being caught by the tide boiling past us just a few miles from our destination. I was astounded to feel the boat pirouette, completely out of our control, as the tidal front, whirlpools forming around us, grabbed us and spat us out. It was a salient lesson for a teenager on the power of the tide.

Here in the Pacific it is a bit different. For most of us sailing W in our crossing, the tide is of little interest. We look for current and any change to the normal Trade wind which will effect the wave pattern and size of seas we will experience. The current that runs most of the year E to W in a huge band S of the Equator can provide you with an excellent 0.5-2kts lift at times and this passes through the Tuamotus. 

Where tide is of interest is around the atolls, specifically when you are trying to enter or leave one. Each pass is a tidal gate. The best comparison I know of in UK waters is the pass at Cuan Sound on the West coast of Scotland where the tide flushes one way then reverses at high speed offering but a few minutes of slack for a slow moving yacht to transit, the rest of the tide being a white knuckled ride trying not to hit the Cleit Rock. The difference is the passes here are far more affected by wind and wave effects. 

After visiting our third atoll and listening to the daily dramas of others on the net, we are coming to the conclusion that working out when that elusive slack is going to happen involves at least as much art as science.

In its section on estimating slack water, Charlie’s Charts talks about using the NOAA tide charts then references another publication, Sailing Directions (Planning Guide), South Pacific Ocean (Appendix Atlas) that suggests that

“ the minimum current is most likely to occur one hour after Low tide and one hour after high tide as specified in the Tide Tables” but does not give a justification for this arbitrary fixed figure.  It also gives some other guidelines for individual passes, again in relation to a fixed time from a primary port. The other publications have the same problem. They want to give you guidance but are limited by the fact that any guidance they do give is likely to be inaccurate. So they stick to broad definitions which are worse than useless in my book as they give false confidence and cause confusion.

There are few primary ports here and the distance between where you are trying to work out your tides and where the primary is can run to hundreds of miles. The two closest to us at the moment are Rangiroa (160miles away) and Tahanea  (about 100 miles), neither exactly close. Not a problem, I hear you say. It should, if measurements have been taken over the years, simply be a case of adding or subtracting times to suit your atoll and Bob’s your uncle, there you have your slack times. Sadly not. It is not possible to be able to look at a tide table of an atoll and know exactly when slack is going to occur. The Tide Table is just the starting point.

A very popular tool, first released in 2011 and used by many cruisers here is called “The Tidal Guestimator”. Some go as far as to call it the “Confusimator” but I rather think that they have decided that slack calculations goes beyond art and into witchcraft. The Guestimator is an Excel spreadsheet that works out the slack tide timings based on a  +/- to the data from a single primary port, Rangiroa, one of the northern atolls and then adding in further corrections. Whilst it looks good, it is not a wholly accurate predictive tool as it requires you to guess/use art/estimate for one key bit of information which has a massive effect on the result you get out of it, hence its name. More on this later.

The big problem with the Tuamotos is that there are additional factors that can significantly alter the time of slack water or even negate it altogether. Although tides are small, rarely over one metre even at springs, the base outward current velocity at passes in periods of clear, still weather runs between 4 and 6kts, dependent on the atoll. With huge amounts of water racing in and out of the often narrow passes with reef tight either side, races, boils, standing waves and overfalls are common outside the short period of slack, meaning transit during slack remains preferable. However, the key statement that you must understand is that the outgoing current speed is effected by weather and can, at any point of the tide, be significantly higher a velocity than the incoming tidal stream.

The effect of weather is always considered when consulting UK tidal tables. Wind over tide is a major consideration – how I hate the Solent chop. In the Tuamotus, one must not presume to calculate using just the current weather. It must be considered for days previous to the time you wish to use a pass. Let us look at why.

Atolls mainly run SE to NW in attitude. Most atolls are made up of motus (the smaller islands making up the atoll’s edge) on the E side, where thousands of years of the normal Easterly Trade winds have deposited sand and spoil on the inside of the reef to form them. The protected W side is far more barren and most often consists simply of reef with a few small isolated motus only. The S and N ends of atolls will normally have some motus due to the wrap around effect of the seas but they will be small and usually have large gaps of reef segments only between them.

This means that if there has been a lot of wind with a S or W component then more sea than normal will break across the open reef, filling the lagoon up.

Some atolls (Makemo for instance) have few very long motus on its E side which gives wonderful protection to the lagoon. However, the greater the number of motus making up the E side of an atoll (Raroia has lots), the greater the number of gaps in between where the water can flood in over the reef. So, this suggests that a stronger than normal E wind (known as a reinforced Trade) would spill more water into some atolls too. This is unaccounted for in the guestimator calculations and I found a passing reference to it only, mainly to do with wrap around waves rather than anything directly from the E.

Once the water is in the lagoon, it needs to escape and most atolls have but one or two main passes where the water can escape from. Limited amounts of water will flow back out through the gaps between the motus or back over the reef.

Pressure will change matters too but for our cruising period and for most times of the year, the Tuamotus are blessed with near continual highs rolling E about a 1000 miles to the S which keeps pressure near constant. We will disregard this as I have no data to account for its effect.

The direction a pass faces is a factor. One that faces E, into the Trades, will always have a wind over tide effect during an outgoing flow, holding back and slowing the outward flow down. For a pass pointing W, then the opposite effect.

Then there is the size of the pass. The Makemo pass is 500m wide and 15m in depth. A good wide, deep channel allows more water to pass out of the lagoon. A small narrow one will have a limit on the water it can pass which also causes a delay to the slack. This is the Slack Delay Timing,  the least well defined figure in the Guestimator.

And lastly the size of the atoll and so the size of the capture area for water flowing in to the lagoon. The strongest of tidal races is at Hao, a very large atoll. Over 30 miles long it has a single pass facing N. With so much water coming into the lagoon thrown over the reef, the race can go well past 10kts on an outward flow. We were discouraged from going here as our first atoll due to the write up on Hao to be found in the US Sailing Directions, an excellent publication, which frankly worried me.

“ The rate of flow entering Passe Kaki can reach 3 knots at HW, when the water level in the lagoon is low. The outflow can exceed 12 knots, 6 hours after HW, when the water level in the lagoon is high. A S swell may lead to a phenomenon of water piling up in the lagoon, with result in large and sudden changes in water levels of up to 1.8m. A tidal race and overfalls may extend up to 0.8 mile seaward of the channel entrance.

South Pacific Anchorages goes one better claiming the current at Hao “can obtain 20kts”! I think this is a typo as no other publication supports this. Either way, scary.

So, the amount of water in the lagoon trying to escape via the pass at any point of tide is the sum of what the weather has dumped across the reef and the normal tide.  The more water there is in the lagoon, the greater the current strength and speed of the outgoing stream. This then effects a difference between calculated and actual slacks times. And if the lagoon is very high after a period of bad weather, you may find that the outward flow negates the incoming tide, meaning no slack period at all.

The Tidal Guestimator gives you a starting point to work from to estimate what the current flow will be. You have a base tidal stream velocity, of which each atoll is given its own value. This is between 4 and 6kts. The correction you need to make to this is called the Wind/Wave Current Factor.

The Guestimator provides the following additions to current outward stream velocities:

1. Add 1 kt for every day the wind has been blowing over 20 kts from a S or W component

2. Add 0.5 kt for every day the wind has been blowing over 15 kts from a S or W component

3. Add 0.5 kt for every 1/2 meter increment of southerly swell over 1.5 meters (ie 3 meter swell = +1.5 kt) 

4. Cap the Wind Wave factor at 1.5 times the Normal Max Current

5. Subtract 0.5 kt for wide/deep passes and for each extra pass that an atoll has

As the Guestimator has been published for several years, these numbers must have been worked upon and do seem roughly accurate. They have worked for us so far. Note – there is no addition here for a strong E wind, one of the identified factors above. I think, for a few atolls, it may be a necessary addition.

Whilst we found the slack delayed at Raroia and Makemo, neither by a significant amount to our calculations, our own experience in entering Kauehi proved to us that you do need to be aware of the vagaries that occur here and take advantage of what you see in front of you.

After delaying our departure at Makemo and getting a bit bumped around mid tide, I decided I would be a bit more careful entering Kauehi, where the normal tide is said to run to 4kts and to enter at the end of the incoming tide. The Tidal Guestimator, based on a simple time variation from the Rangiroa measuring station, gave me an initial high tide slack time of 0920hrs for 7 Sep 16.


I corrected it by adding in the Wind/Wave Factor, which we had guestimated to be 2kts and plugged the numbers into the spreadsheet. We had sailed here on the first day after a long period of reinforced Trades, blowing at 20-25kts from the SE. I decided to go back three days. That gave me two days of 20+ and one of 15+ and then I took 0.5kt off for the wide pass. Out spat the graph below. You can see two slacks in the morning, quite close together. The slack I needed with enough light to see to be safe is at 0815hrs.


Being prudent I had also looked at Charlie’s Charts for pass information, which then gave another direction for working out slack timings. 

“The current is slack 1.5hrs after low water in Tahanea, the closest tidal reference point”

Using the web site (the official French Maritime web site – an excellent resource as long as you have internet), this offset gave me an initial time of slack for Kauehi of around 1045hrs.  A difference of 1hr 25mins from the Guestimator.  A huge variation! And using a standard sinusoidal graph, then I should expect the same current going in as out? Some mistake, surely?

Unsure what time to plan to and of my estimate of the wind/wave factor, we decided to arrive early. By 0745hrs, we were sitting off the entrance to the atoll, expecting to see the water overfalls the pass is known for during an incoming tide and to have to wait a while before it became calm enough for an entrance. Lou and I both checked with the binos. Nothing. No standing waves on the outside either. Recheck – still nothing.

At 0800hrs, we decided to take a turn in and ran into the pass expecting the tide to make itself known. Absolutely zip.

We pushed through quickly with less than a knot of tide against us and flat water. So, according to what we experienced, the tide was outgoing but near slack, which we must have missed by less than 30mins. Our estimate of the wind/wave factor had been about right, (maybe a 0.5kt light) and we got the conditions we needed to get in safely. Just look at how little time the incoming tide was able to negate the outgoing flow. We probably should have gone straight  in to the pass rather than standing off and debating what we were seeing. 

Perhaps the Guestimator’s accuracy could be improved. Some thoughts:

1. The addition of properly researched Rangiroa offset times for each atoll,  to me a screaming omission.

2. The Tidal Slack Delay times, if it is said to be a constant (and for the islands with data against them, it is), should be a completed column too but this requires some hard data to be collected from cruisers or extracted from the SHOM web page.  It may (as suggested in the Guestimator instructions) be easier to leave this column blank and discount it. If left set to 00.00, it will mean you won’t miss the slack because of it as you will always guarantee yourself that you will be early. However, this addition, given in increments of 10 mins up to 1hr 30mins, suggest that this figure should be nailed down further.

3. Tighter instructions on the use and calculation of the crucial wind/wave factors, perhaps with advice on a cut off on the number of days you need to look back on. Three days seems to work for me but we have been here with reinforced Trades only rather than anything truly foul.

However, if we accept calculating the slack really is at least as much art as science, then the current formula, used properly, is probably already as good as you are likely to get.

In the meantime, Gram Schweikerk who took the time to make the Guestimator up, still has it updated annually with new tidal information and as the only free resource out there for this problem, he and his helpers must be thanked and praised for their efforts. I will continue to use it and now have a far greater understanding of how to get the best out of it. I will also be donating to his beer fund which thoughtfully comes as a link on the document.

We have come to the conclusion that you need to take all published information with a huge pinch of salt. If not wrong, it will certainly be inaccurate as it is most likely written for a set of generic weather conditions you are not in.

The Tidal Guestimator is of great use but you need to think hard about the data it asks for if you are going to use it properly and get any semblance of an accurate answer from it. It is not difficult but it does require you to engage the brain. For ourselves, we will continue to aim to be early outside the pass, preferably waiting for the end of an ingoing tide,  watching for the period where the sea calms and slack approaches. We will continue to base it on tidal data and our estimate at the wind/wave factor which I hope will only get more accurate as we gain more experience in the atolls. Sometimes we might even have to push the tide but we will understand the why and when it would be best to do so.

I could say that we will try to get better with our Scientific Wild Ass Guessing. I prefer to say we will practise art.


It is interesting to note that the US Sailing Directions, Ocean Passages and Landfalls, The Pacific Crossing Guide and South Pacific Anchorages all discuss using the times of moonrise and set as the basis to work out slack water. I have to admit I have not used this at all as there seems there is considerable confusion between publications on just how to work out your start points.

From the US Sailing Directions for Hao:

“To avoid a difficult passage through the reef, vessels should wait for the two periods of slack water associated with the flood current, which are short. Slacks usually occur about 4.5 hours and 2 hours before moon rise; and again 5 hours and 3 hours before moon set. When the tidal race slows or stops, the channel may be entered. Caution should be observed, as the information given above is for average conditions only. Current rates and the times or presence of slack waters may differ from those the vessel may experience.”

With considerable experience as a UK Army staff officer (don’t believe what they say about us – we even understand what Navy people write), I translated this to “ hold on to your pants ‘cause we don’t have a clue what you’ll find but it will be hairy!”

Ocean Passages and Landfalls states that for the Tuamotus:

Slack water will be 12 hours from moonrise or moon set.”

Then it states

Using the 12 hour moonrise/set system you may find that the time for slack water does not exactly coincide with the tide tables but it will be close”

Lastly The Pacific Crossing Guide states:

“…the tidal flow in the passes can be predicted from the time of the moons rising and setting. The full details are given in the BA Pacific Islands Pilot Vol III. Briefly, there is a slack water 5 hours after moonrise, followed by the inflow, and slack water again 4-4.5 hours before moonset when the flow reverses; 5 hours after moonset and 3 hours before moonrise, the pattern is repeated. However, tidal flow can be dependent on swell and wind, and it would be wise to back up calculation for any specific atoll with local knowledge”

Compare the three and there seems to be a significant difference in opinion. Ocean Passages and Landfalls I am discounting as it looks like it is using a overly simplistic model. The Pacific Crossing Guide and US Sailing Directions are close although the US Sailing Directions is detailed only for Hao rather than for all atolls and seems to me to already have a calculation incorporated to account for a Wave/Wind factor.

The Pacific Crossing Guide description appears to be the best and applicable to use throughout the atolls. Knowing that the RCC and OCC both have a strong tendency towards well researched competency, I’ll trust my initial tests against tide timing data to their timings. Bless the British Admiralty and the days Britannia Ruled the Waves! The Pilot referred to is a wonderful read on the history and discovery of the Pacific Islands and contains fascinating information on weather with 30 years of measurements for the stations it uses. Sadly, in terms of tidal information, it expands only slightly on the information contained in the RCC Guide. My copy is the 9th Ed from 1969. I’d be very interested to compare it to a newer digital age edition.

In the meantime, I will need to look at this technique a little closer before I use it and will make up a sheet to write down the simple calculations although it would be good to be able to talk about time in increments of less than 30mins. Of course, you need accurate moon data for your lat/long but I have this from the excellent app, Star Walk. The answer will still need to be adjusted for Wind and Wave, so just as it does using tide tables, I think art may come in to the completed solution too.

 The Vagories of Tuamotus TidesThe Vagories of Tides

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s