Ok now done! Be aware it is long and needed to be split across three posts! It might be best to tackle it in several parts at a time.
Understanding Port Phillip Bay Tides - How Weather Affects Tides, Streams, & Slacks
********* Intro *********
Actual slack water times at Port Phillip Heads can differ from even the best theoretical predictions due to the effects of the weather. Trying to correct for the weather is a very tricky area, and not well understood by most divers. Even armed with the best understanding of the factors involved, you may get only partial success at best.
Being a Bit Brutal: It is likely the majority of those who dive the Heads region still have at least one, (and possibly two), commonly held beliefs on slack water that are simply untrue and unhelpful. Despite it being said and written a zillion times, (even in government/shipping media), the fact is slack water at the Heads DOES NOT OCCUR when, or because, the Bay and Ocean levels become equal.
Until divers manage to shake off this historical and simplistic myth, there may be little point in continuing with this post right now. Instead they first need to read the last couple of screens of the top post in the packo dive-Oz "final smack-down" thread starting at the heading "*** OTHER STUFF: The Equal Levels Moment,. .".
For those who need to see it in the real world, simply observe water levels at Pt Lonsdale (or Pt Nepean) reef platforms at a mix of flood and ebb slacks. In each case the reef platforms will be either just drying, or just submerging, with the ocean level being roughly the same for any slack to within roughly +/- 10cm except on some days where very unequal tides occur.
In contrast we know the central and north bay areas have water levels at these times close to either high tide or low tide. The hi/lo level difference in the north is typically around 55cm. This shows bay levels are definitely not equal to the ocean level when slack water occurs.
In fact the very brief equal levels moment occurs roughly 40 to 60 minutes before slack water. At this time the stream is still running at around 2 knots. The water's momentum allows it to continue moving in the same direction despite the changes in ocean and bay levels now creating an uphill slope. Naturally this creates a force which begins to retard the flow as now it must climb slightly uphill.
This retarding force grows with the increasing slope as the ocean and mid-bay levels keep moving in opposite directions. During this uphill flow phase the frictional forces quickly drop away as the speed falls. The reverse water slope then becomes the dominant "stopping force" approaching the slack water diving window.
After the water stops briefly, it begins to reverse down the slope. The slope continues to grow, but at a slower rate since the mid-bay level is no longer changing. The increase in friction as the water speeds up is now opposing the down-slope forces. The net result is that the acceleration number, or rate of change of current with time, reaches a slight peak just after slack water. Nevertheless the rate of change during the 20-60 minute diving window can be considered as roughly constant.
I have chosen to measure these acceleration numbers at slack water in units of "knots per ten minutes". It is simply a pragmatic and sensible choice. If given in the more scientific units of "meters per second per second", the numbers would be hard to remember with three extra zeros needed. Any necessary mental arithmetic would then be more difficult.
While novice boatmen have been mocked for incorrectly saying "knots per hour" as a unit of speed, note that any form of "knots per unit time" is a perfectly valid way of expressing an acceleration value. Note also that here "acceleration" is used in the scientific sense with both a value and a sign. Depending on the combinations of the signs of both the velocity and the acceleration, it may be describing either a "slowing down" or a "speeding up" tidal steam.
The height of the hill that develops while the water is being brought to a halt varies between 12cm and 45cm depending on the particular tide cycle. This height is distributed over a distance of about 15km so the slope can't be seen with the naked eye. It is however clearly visible in the tide gauge data once these are corrected to a common datum level. This link shows some tide curves and the "uphill flow phase":-
Seven days of tide curves
It is the "hill height" at slack water that determines the rate at which the tide stream stops and reverses for each particular slack water event. This "acceleration number" not only determines how long the "diving window" is for that particular slack water, but also how vulnerable it is to having its timing pushed around by the effects of the weather.
********* Part 0 - Do I Really Want to be Here? *********
It is a difficult area to learn about, but in the long run the effort may be worth it. A diagram of tides during a large storm is included in the post. Storms greatly exaggerate weather effects and the diagram should help visualise what is being presented. Various other diagrams assist in showing how the theory is applied in practice.
Many divers may not wish to get into this topic. They can avoid the learning effort simply by always making a generous "weather allowance", and arrive at the site 30-40 minutes before the expected "drop-in" time. However long waiting periods can be frustrating and can cause distraction from the key task of monitoring the current as it decreases towards your chosen "drop-in value".
There will also always be situations when for one reason or another you can't make such a weather allowance, or might prefer not to waste that time if there are reliable ways to predict if an upcoming slack water is likely to be early, on-time, or late.
This long post is written for those who might want to better understand the factors involved. Hopefully that may result in better dive planning decisions. Due to its length, the post might best be tackled in chunks of 3 or 4 parts at a time.
********* Part 1 - The Observed Tide Curve *********
The sequence of readings from a tide gauge trace out a curve known as the "Observed Tide Curve". The readings are typically at 6 minute intervals with measures being taken so the short term variations in water level due to waves and swell are averaged out over this time period. This is not always entirely successful. In rough weather the observed tide curve from many bay tide gauges is often quite noisy with many small random height variations evident in the overall curve.
There are three main factors responsible for producing the observed tide curve:- Motion of the Moon, motion of the Sun, and the motions of weather systems down here on Earth. The first two have been determined very accurately and together with long term tide observations at a location, can be used to accurately predict the "Astronomical Tide" height at that location at any future date and time.
Those with a solid understanding of the astronomical tide may wish to skip over Part 2.
********* Part 2 - The Astronomical (or Predicted) Tide Curve *********
Despite the enormous differences in the mass of the Sun and the Moon, the difference in distance from the Earth means their ability to raise tides is roughly similar, but with the Moon's influence being 2.1 times as strong as the Sun's.
Unfortunately some enthusiastic school teachers left pupils with the notion that on Earth the Moon's gravity is stronger than the Sun's because it is so much closer. If that were true, the Earth would orbit the Moon rather than the Sun! In actual fact the Sun's gravitational pull here on Earth is 178 times stronger than the Moon's.
It turns out the tide raising forces do not depend directly on the strength of the gravitational pull of a celestial body, but rather on how much that pull differs between the near and far sides of the Earth. It is because the Moon is so much closer than the Sun that its gravitational pull, even though very much weaker, actually changes in strength by more than twice as much as the Sun's pull changes over the distance of one earth diameter.
This "Differential Gravity Field" is also the reason why two equal water bulges occur for each celestial body, one on the near side directly "under" the body, and one on the directly opposite side of the Earth.
There is a complex interplay between the two sets of tidal bulges because the Sun and Moon appear to rotate around the Earth with slightly different cycle times (24 versus 24.8 hours). This gives us roughly two Bass Strait high tides per 24hr day, with every second high tide occurring on average 50 minutes later each day. However this progressive daily delay can vary between 35 and 65 minutes later each day depending on the phase of the Moon.
Also varying with the phase of the Moon will be the amount by which the tidal forces of the Sun and Moon either reinforce, or partly cancel each other. This gives us either "Spring Tides" of a higher range at maximum reinforcement, or "Neap Tides" of a lower range at maximum cancellation. Although the net tidal forces will peak at both New and Full Moon times, there is a delay of around five days for the oceans to fully respond.
The highest tidal range and shortest slack waters therefore tend to be around 5 days after a New Moon or a Full Moon. That is at day_#5 or day_#19 of a lunar cycle. Similarly, the weakest tides, and the longest slack water times, are generally found around day_#12 or day_#26 of a lunar cycle. Divers should keep this in mind as the duration of a slack water "dive window" can vary by a factor of 3 over a complete lunar cycle.
In addition to the differences in east-west motions of the Sun and Moon causing variations in the tides, movements north and south of the equator also have a significant effect. For example as the Sun moves northward for our winter, it drags its day-side bulge further north. Its night-side bulge will move southward. As the Earth rotates, our southern coastlines are dragged through the smaller edge of the day-side bulge but closer to the middle of the night-side bulge.
This effect gives unequal sizes to consecutive high, or consecutive low tides. This is called the diurnal inequality. It also means the height range between successive highs and lows can change significantly over a tidal day. This means the duration of the diving window at each of the four slack water times will vary over a 24hr day.
The Moon's north-south wanderings further complicate the picture because its north-south cycle is rather rapid at 28.4 days, compared to the Sun's 365 days. In addition, the range of its north-south journey also changes between a latitude band of +/- 18 degrees and a latitude band of +/- 28 degrees over a very long 18.6 year cycle. The Sun's yearly latitude band limit is essentially fixed at +/- 23.5 degrees.
The interplay of the motions of the Sun and Moon give a complex astronomical tide curve with daily, monthly, yearly, and longer term variations. Further minor complicating factors are the slightly non-circular orbits of the Moon and Sun as well as a few other orbital quirks.
Despite all these complexities, future predictions of the astronomical tide can be made for certain locations many months or years in advance. Their height accuracy is around just a few millimetres.
********* Part 3 - The Residual Tide Curve *********
When the highly predictable "Astronomical Tide Curve" is subtracted from the "Observed Tide Curve", the difference produces a third curve called the "Residual Tide Curve". The residual tide curve is a grab bag of left-over positive and negative height differences that are attributable to terrestrial causes. The main contributor here is the weather.
Those weather effects range from the short term passage of squalls and weather fronts, to the usual 3 - 7 day cycle of high and low pressure weather cells, through to seasonal changes, and then to longer term trends such as the El Nino or La Nina climate cycles.
********* Part 4 - A Monster Storm *********
In late June 2014 the biggest storm in perhaps 20 years hit the Victorian coastline. The observed, astronomical and residual tide curves for Williamstown are helpful in understanding what happened.
Note that to a rough approximation, actual slack water at the Rip occurs just before the (red) observed Williamstown tide curve reaches a maximum or minimum, while the predicted slack water time is just before the (green) astronomical tide curve reaches a maximum or minimum.

Significant points during the progress of this storm are numbered in the diagram above and are discussed below. Note the dates and times in the diagram are UTC and 10 hours needs to be added to give the local times as mentioned in the text below.
#0 - The storm began on June 23rd and the residual tide at Lorne rose to +50cm over about 8 hours. At the Rip a flood slack was due shortly after the residual tide began to rise.
#1 - This flood slack was delayed by 80 minutes as the rising residual tide kept pouring more and more water into the bay.
#2 - When the ebb stream did finally begin, it was weak and lasted only 3 hours instead of the usual 6.
#3 - The following ebb slack arrived 90 minutes earlier than predicted.
#4 - A long and enhanced flood stream then followed. The residual tide stabilised at around +45cm for the next 16 hours of the storm. The streams and slack water times quickly reverted back to the predicted times, except of course that all measured heights were 45cm above the predicted levels.
#5 - At 10am on the 24th, another massive surge quickly took the residual tide to a new peak of +85cm. This gave an all time record tide height at Williamstown and flooded parts of the Southbank precinct.
#6 - Luckily the peak in the residual tide occurred somewhat after the peak in the astronomical tide and so spared the area of an extra 20cm of flooding.
#7 - The residual tide fell quickly back to +50cm not long after this and the ebb slack at the end of the outgoing stream was delayed by 140 minutes due to the extra water draining out of the bay.
#8 - At around 8pm on the 26th, the residual tide dropped sharply from +50cm to +20cm, delaying the ebb slack by 2 hours.
#9 - This extra outpouring of bay water also shortened the following flood stream from roughly 5 hours to just 3 hours.
In all the storm event lasted about 4 days after which the residual tide fell to near zero and the observed tide curve quickly returned to follow the predicted tide curve to give normal slack water timings.
In storm events like this we have a large residual tide anomaly, usually called a "storm surge", being superimposed on top of the normal astronomical tide to give a "storm tide". Note that at Williamstown, the best efforts of Moon and Sun can raise the water level by about 35cm above mean sea level, whereas large storms like this one can raise bay water levels by 85cm above normal.
The timing relationship between a storm surge and the astronomical tide is completely random, and so the severity of the resulting storm tide depends on chance coincidences between the timing of peaks in the surge, and peaks in the astronomical tide.
The same is true for the effects on streams and slacks. Had the timing of that storm been slightly different, any of the following could have occurred:-
- The maximum Rip current could have exceeded 10 knots.
- The flood slack could have been delayed by up to 5 hours
- The ebb stream could have been entirely obliterated, with water flowing in continuously for nearly 18 hours.
While the drastic effects of storm tides on beach erosion are related to the height of the residual tide, the effects of changing the strength of a tidal stream or of changing slack water timing, depends instead on the rate at which the residual tide changes with time.
In this storm the bay-wide residual tide height was changing at sustained rates of up to 9 cm/hr. The relationship between Rip current and average Bay height change rates is roughly:-
1 knot of Rip current <==> 3 cm/hr bay level change
Note this relationship is fixed for our particular combination of bay surface area and Rip cross-sectional area. It holds regardless of whether the bay level change is caused by the residual tide, or the astronomical tide, or any combination of the two.
Bay water height changes as seen in this storm correspond to a sustained "residual current" (or "weather current") component at the Rip of up to 3 knots. It is not surprising that streams and slacks are significantly altered during big storms.
The story of this storm is to reinforce three important weather effects:-
a) When the residual tide is rising, flood slacks are delayed and ebb slacks arrive early.
b) When the residual tide is falling, ebb slacks are delayed and flood slacks arrive early.
c) If the residual tide is not changing, even if it is at a very high or very low level, slack water timings are not affected.
********* Part 5 - Everyday Weather Effects *********
Although this storm was an extreme example, it turns out that even in ordinary weather the residual tide level can move around by up to +/- 4cm/hr on some diving days. This is equivalent to a "weather current" of +/- 1.3 knots.
Note that since the normal astronomical tide current, and any weather current exist totally independently of one another, the total Rip current is simply the sum of the two components, taking into account their individual signs.
So the new time of slack water in the presence of a "weather current" is simply that new time when the total Rip current is zero.
If the weather current is roughly constant, ie. slope of the residual tide curve is roughly uniform over the time of interest, then estimating the time correction is fairly straight forward. We first need to know the rate at which the astronomical Rip current changes around slack water. From this we can estimate how long before or after the predicted slack time will the astronomical current component be sufficient to exactly cancel out the weather current. Adding this time correction to the predicted slack time will give a new "weather corrected" slack water prediction.
One of the reasons for producing the "packo predictions" material I have been releasing is to provide exactly such acceleration or rate of change numbers for each slack water event. Even if you don't want to get into the details, it is a useful thing to know that slacks with the higher acceleration numbers are less vulnerable to being moved about by the weather.
Conversely the timing of slacks with low acceleration numbers is far more readily shifted about by the weather. You can save a lot of time by simply adjusting the size of your "weather allowance" to be inversely proportional to the acceleration number of the particular slack you intend diving.
The magnitude range of slack water acceleration numbers is roughly from 0.20 knots/10 mins up to 0.60 knots/10 mins. If the number has a negative sign it means the acceleration force is directed outwards through the Rip. Flood slacks have negative values as the bay level is higher than the ocean level at the time of slack water. The outward directed force is responsible for slowing the last of the flood stream before slack water and also for speeding up the new ebb stream after slack water.
Assume for example the residual tide is changing at +3cm/hr, giving a weather current of +1.0 knot. Now assume that a flood slack with an acceleration number of -0.5 knots/10 mins is approaching. That slack will be delayed by (1.0/0.5) x 10 = 20 minutes beyond the original slack water prediction.
At that new time, the outward astronomical current component will be -0.5/10 x 20 = -1.0 knot which will exactly cancel the +1.0 knot weather current component to give zero net current.
If the flood slack event had a lower acceleration number like -0.2 knots/10 min, then the time delay would be longer at (1.0/0.2) x 10 = 50 minutes. If instead the slack was an ebb slack at +0.2 knots/10 mins, then both the inward force opposing the ebb stream and the inward weather current would combine to cause the slack to arrive 50 mins earlier than the original prediction.
An important point in making an adequate "weather allowance", even if you have no idea what the residual tide is doing, is that for low acceleration slacks it should be very generous. This is because firstly the best "drop-in" time might be 20 mins before dead slack and secondly a weather shift may be say an extra 30 mins earlier (or not! ).
The rest of this post looks at how a diver can get some idea of what the residual tide curve might be doing to try to prevent some unnecessary time wasting or improve some other dive planning decisions. For example if you think you might be running late to get to the site, it may help you better decide if you should press on, or give up and divert to a closer alternative dive site that you may reach in time.
********* Part 6 - Components of the Residual Tide *********
There are several ways that the weather affects the residual tide value. However untangling the exact relative strengths of each of the components is difficult and involves some guesswork.
1) Inverse Barometer Effect: As the high and low pressure cells of our weather systems move across the oceans, the high cells press down harder on the ocean surface and the water level there drops. Under the low pressure cells the ocean surface rises. As these weather cells move through Bass Strait the sea level will rise or fall oppositely to the movements in barometric pressure. In turn this will raise or lower the bay levels relative to the normal astronomical tide curve.
An intense low pressure cell may lead to a rise in the residual tide of up to +40cm. Strong high pressure cells, which are geographically more spread out, may produce a residual tide component of around -25cm.
2) Ocean Wind Set-Up: Winds blowing across an ocean will move some of the water in the direction of the wind. If the wind is onshore, the extra water piles up against the shore and raises the water level there. If the wind is offshore, some water is blown out to sea and the inshore level falls.
These effects on the residual tide is guessed at being -15cm for strong offshore winds and up to +40cm for strong onshore winds.
3) Bay Wind Set-Up: Winds blowing across the bay also affect the bay level by reducing the level on the upwind side and increasing it on the downwind side. If the bay were a closed basin or lake, the level in the middle would remain roughly the same - as would the total amount of water.
However if the upwind "low spot" side of the bay is near the Heads, extra ocean water will enter the bay and try to fill up this low area. The wind redistributes this extra water and the average level rises everywhere. This happens for winds between west and south.
Winds between northeast and east do the opposite. They allow "built up" water near the Rip from the downwind "high spot" to leave the bay and so the average level falls.
The shape of the coastline just inside the Heads seems to suggest that northwesterly winds create a low spot near the Heads and so extra water flows in to raise the average level. Southeasterly winds do the opposite.
Guesstimates of "bay wind set-up" residual tide components are +10cm (NW to SW) down to -8cm (NE to SE). The effect of winds directly from the north or south is a little unclear.
4) Wind Helping/Hindering Rip Tidal Streams: Winds between west and south will help the flood stream and hinder the ebb stream. Both scenarios raise the bay's level above the astronomical tide predictions. Winds from north to east help an ebb stream and hinder a flood stream, leading to levels falling below the astronomical tide predictions.
Guesstimates are from -5 cm to +8 cm with the potential for higher positive values for flood streams if swells in the Rip start to break in strong SW winds.
5) El Nino / La Nina: These longer term weather effects give residual tide components of around -15cm and +15cm. Such components are roughly steady for some months and would not normally influence slack water timings.
. . continued in next post at: Part 7 - Trying to Put it all Together