Tidal Water Volume Slop

Tidal Water Volume Slop

Post by Chris » Wed, 10 Apr 2013 06:26:15


Quote:


> >> The average is calculated for 130 years' worth of records for Kingston from the National River Flow Archive. This is about 101 cumecs for March, 75 for April. This year on 31st March there was about 180 cumecs on the day - it is difficult to get reliable figures until they're archived, but this was countered by a reasonably good tide and a following wind, hence a reasonable quick time, 12th fastest if I've counted correctly. I'll convert my spreadsheet for calculating boat race winning times and stick it on Google docs.

> Ok, so we know how much water is coming down on a day-by-day basis

> thanks to the NRF archives, but for the flow on the tidal part of the

> river you speak of a "tidal stream calculator"....? So is this an

> estimate or an actual device?

It's a numerical integration of the Thames based on the work of Rossiter and Lennon in 1965. Divides the Thames from Sheerness to Richmond into 12 compartments and calculates the tide level and stream speed in each compartment every 10 minutes. Initial values are the tidal levels at based on 19 years of hourly tidal observations from the Liverpool Tidal Institute, and the land water at the other end from the NRFA.
 
 
 

Tidal Water Volume Slop

Post by Chris » Wed, 10 Apr 2013 06:43:54

Quote:

> Here are the past twenty years of Boat Race times and water flow:

> 27/03/1993 29.5    17

> 26/03/1994 71.4    18.09

> 01/04/1995 70.5    18.04

> 06/04/1996 34.5    16.58

> 29/03/1997 7.09    17.38

> 28/03/1998 48.4    16.19

> 03/04/1999 45.9    16.41

> 25/03/2000 48      18.04

> 24/03/2001 346     19.36   (Goldie)

> 30/03/2002 68.2    16.54

> 06/04/2003 52.4    18.06

> 28/03/2004 49.9    18.47

> 27/03/2005 21.9    16.42

> 02/04/2006 114     18.26

> 07/04/2007 62.1    17.49

> 29/03/2008 153     20.53

> 29/03/2009 49.9    17

> 03/04/2010 159     17.35

> 26/03/2011 40.3    17.32

> 07/04/2012 15.8    16.41   (Isis)

> I looked at this and thought: that looks like bollocks, but popped it

> into Excel anyway and did a scatter graph with the flow represented

> logarithmically and I was surprised to see a definite trend there.

> I am forced to admit that scatter graphs can reveal non-obvious stuff.

> CAVEAT - relying on flow volume to PREDICT a (record) time is still idiotic.

Yes it is if you ignore the wind and the strength of the tide.

Here are my predicted times here

https://docs.google.com/spreadsheet/ccc?key=0AuSZSCdZwP09dDRqZGdIOGdT...

It makes several assumptions

1. Crews get faster over time - approximately 0.11 cm/s/yr (rate of improvement in world best times since 1964)
2. Flow at Kingston is approximately the same as at Teddington
3. Stream speed is as the model predicts
4. The wind speed and direction at LHR approximates to that over the course. The wind model is based on that proposed by Burnell in his 1954 book celebrating the 100th race. Burnell's model however, has serious deficiencies because he gives very different values when there is hardly any wind. I have adapted this on the same principles, i.e. that a head wind is always unhelpful and a tail wind is initially helpful until it is so strong that it is a hindrance. This is then adjusted to affect the crew speed to get best fit with the data.

Several questions spring to mind

1. Why was the 2008 crew so slow compared with predicted - poor recruitment in an Olympic year perhaps?
2. Why was the 1999 crew so fast?

When I get the time I'll do the calculations for the Isis/Goldie races as well.

 
 
 

Tidal Water Volume Slop

Post by Chris » Wed, 10 Apr 2013 06:45:46

I should add that this model has been a nice little earner on the boat race winning time market over the last few years.

 
 
 

Tidal Water Volume Slop

Post by Alistai » Wed, 10 Apr 2013 06:45:48

Quote:

> It's a numerical integration of the Thames based on the work of Rossiter and Lennon in 1965. Divides the Thames from Sheerness to Richmond into 12 compartments and calculates the tide level and stream speed in each compartment every 10 minutes. Initial values are the tidal levels at based on 19 years of hourly tidal observations from the Liverpool Tidal Institute, and the land water at the other end from the NRFA.

So the input is the volume of water at Richmond, and then you do some
maths and reckon you can work out the tidal flow down the rest of the
Thames?

What's the margin of error? I would imagine it would have to be quite
high. For instance wind had a big effect on tide: if it's blowing
strongly in the estuary / sea it can significantly slow the water coming
flowing up river; and barometric pressure too will effect the amount of
water slooshing about. None of which can be taken into account by just
looking at land volumes.

 
 
 

Tidal Water Volume Slop

Post by Alistai » Wed, 10 Apr 2013 06:53:02

Here's an anecdote for you: I remember the 1998 race well and on the
start looking at the flags down the London boathouses, and they were all
completely limp. Not even a hint of a flutter.

Which doesn't really fit in with the 3.1 m/s SE you have.

In my opinion, a prerequisite for the fastest conditions to row from
Putney to Mortlake (or vice versa) is the stillest wind possible.

 
 
 

Tidal Water Volume Slop

Post by Chris » Wed, 10 Apr 2013 07:04:35

Quote:


> > It's a numerical integration of the Thames based on the work of Rossiter and Lennon in 1965. Divides the Thames from Sheerness to Richmond into 12 compartments and calculates the tide level and stream speed in each compartment every 10 minutes. Initial values are the tidal levels at based on 19 years of hourly tidal observations from the Liverpool Tidal Institute, and the land water at the other end from the NRFA.

> So the input is the volume of water at Richmond, and then you do some

> maths and reckon you can work out the tidal flow down the rest of the

> Thames?

> What's the margin of error? I would imagine it would have to be quite

> high. For instance wind had a big effect on tide: if it's blowing

> strongly in the estuary / sea it can significantly slow the water coming

> flowing up river; and barometric pressure too will effect the amount of

> water slooshing about. None of which can be taken into account by just

> looking at land volumes.

Yes, you're right it has no meteorological factors - I haven't the computing power, nor the maths - but it does model the high and low tide times and heights  reasonably well along the Thames. Height data is available from the PLA's live tides but tidal stream data is very difficult to get hold of.
 
 
 

Tidal Water Volume Slop

Post by Chris » Wed, 10 Apr 2013 07:11:30

Quote:

> Here's an anecdote for you: I remember the 1998 race well and on the

> start looking at the flags down the London boathouses, and they were all

> completely limp. Not even a hint of a flutter.

> Which doesn't really fit in with the 3.1 m/s SE you have.

> In my opinion, a prerequisite for the fastest conditions to row from

> Putney to Mortlake (or vice versa) is the stillest wind possible.

As I said it is the wind speed recorded at LHR (a good 10 miles away). If anyone has reliable data on wind recorded at Putney then please share. But even then the wind can vary significantly during the 20 minutes it takes the race to run. I'd still reckon the greatest unaccounted for difference is the variability of the standard of the crews from year to year.
 
 
 

Tidal Water Volume Slop

Post by johnflo.. » Wed, 10 Apr 2013 20:08:23

Are the flow speed numbers being quoted an average over the entire cross-sectional area of the river?  Seems likely that the flow speeds vary rather widely at different lateral positions across the river and at different depths. And that how this distribution of speeds would be affected by higher flow volumes may be complex.  What we want is the flow rate for the top say 20cm layer of water in the path the boats are taking - flows elsewhere are irrelevant.  Not that this would be easy to get....
 
 
 

Tidal Water Volume Slop

Post by Chris » Thu, 11 Apr 2013 02:40:08

Quote:

> Are the flow speed numbers being quoted an average over the entire cross-sectional area of the river?  Seems likely that the flow speeds vary rather widely at different lateral positions across the river and at different depths. And that how this distribution of speeds would be affected by higher flow volumes may be complex.  What we want is the flow rate for the top say 20cm layer of water in the path the boats are taking - flows elsewhere are irrelevant.  Not that this would be easy to get....

Yes it the velocity at mean depth for that section of the river, and the data you suggest which would be the most interesting do not exist as far as I can ascertain.