Proper depth !

[paulburn]

Just got new chartplotter - Garmin 95sv - excellent - big screen, low price.

But got me thinking....

Why don't electronic charts display proper depth ? ie the current depth of water, underneath your boat, taking into account both the sea bed topography and tides.

So my ideal would a to have a choice of all options below....

1) real time proper depth (to use 90% of the time)
2) a time slider - to allow selection of depth at any future date/time (so you know depth at a predicted ETA)
3) chart datum depth (to have as a historic reference)

Paul

 

[Lon nan Gruagach]

Storm surge and atmospheric pressure would be inordinately difficult to reliably implement and do have a significant effect.

 

[PaulRainbow]

Not sure if i've understood "real time proper depth", but if you adjust the offset you can get the plotter to display a figure plus or minus x feet (m) below or above the transducer. So you can have, depth below the keel, depth below the transducer, actual depth of water or any other odd figure you want. See page 19 of your manual.

 

[prv]

Storm surge and atmospheric pressure would be inordinately difficult to reliably implement and do have a significant effect.

True, but then you might as well throw all the other tidal prediction stuff out the window too. The standard predictions and curves are good enough for most purposes, you can have a think about pressure etc if you're really planning to cut something fine.

At first this seems like a really silly idea, but in fact it's hard to come up with a strong counter-argument. I think it's just that those of us who've sailed for years are completely familiar with the status quo - a chart with soundings reduced to datum, and the rise of tide added mentally on top - so anything else sounds crazy to us. But if you were starting from scratch, why not show real-time depths on a chart image which is being generated on the fly anyway? It's no less accurate than doing the same thing in your head based on the same data.

The only real problem I can see is if someone thought a plotter was set to datum but it actually had the rise of tide already accounted for, and they added the tide manually again. As well as an explicit caption, one helpful indicator might be to use the traditional chart colours when in datum mode, and a new set of colours when the tide is included. This neatly allows a custom contour which you can set to your draught plus an appropriate margin, and colour everything deeper than it in a safe colour, everything shallower as dangerous. Then you can see at a glance where is currently safe to travel over, and where is not. A boon in somewhere like the Channel Islands, where lots of the chart is coloured green but is actually perfectly navigable 90% of the time. The occasional dangerous pinnacles would stand out a lot more that way, currently you need to inspect the drying height of every rock.

So, in my opinion, good idea, at least as an option. Not included mostly due to habit and inertia.

EDIT: Just noticed the OP was talking about real-time depth below the keel, whereas I was considering the same thing with depth to the surface. Basically the same idea though.

Pete

 

[prv]

Not sure if i've understood "real time proper depth"

You haven't

Pete

 

[paulburn]

Sorry - no - i mean depth displayed on the chart (as opposed to transducer depth) - ie the colours, contours and spot numbers used to indicate depth.

 

[coopec]

What about adjustment to the depth reading for when the yacht is heeled too!

 

[bitbaltic]

At first this seems like a really silly idea, but in fact it's hard to come up with a strong counter-argument.

I can think of a couple.

The model of the sea bed provided by a chart is just that- a model, ie an approximation. In some parts of the world a very close one, in others less so. The model of height of tide provided by any set of tables or tide gribs is also just that- a model, ie an approximation. Under some conditions a very close one, in others less so. Combining the two models together will compound the uncertainty in the models in a manner where they can't be seperately assessed by the skipper. That's something I'd be uncomfortable with, because assessing the uncertainty of each is part of my safe navigation.

It's no less accurate than doing the same thing in your head based on the same data.

This is the issue. Whilst there are many places where the resultant errors will be low and this end product would work really well for the user, it's going to present you with something much more precise on the screen than a calculation in your head, leading to an apparent certainty which- in some places- may be false. Electronic charts tend to be amenable to a false sense of certainty, especially for beginners, and a product that increases that perception might also increase the number of plotter-assisted pile ups. I'm especially thinking of the number of drying banks around the Bristol Channel which have charted depths much deeper than they really are; and I don't even want to think about how a 'dynamic depth chart' would be integrated with the generalization of depth contours at low levels of vector zoom.

Additionally, in areas with large tides like the Bristol Channel, this sort of visualisation would make it difficult for a navigator to get an overview of a the seabed topography of a large area as, at any given time, some of the drying areas will have water over them and some not as the tide moves. It will obscure where CD is and what dries and what doesn't, leading to constant re-checking and a background feeling of uncertainty for someone unfamiliar with the area and navigating a long passage.

As I say I can see situations where the idea works- largley for limited or local use. But I can imagine most chart suppliers are reluctant to become responsible for a product of this kind. And I'm not teccy enough to answer this last question but what about processor power required to constantly recalculate everywhere on the displayed chart- will it not also mean a rather expensive plotter is required?

 

[PaulRainbow]

You haven't

Pete

Didn't think so.

 

[shaunksb]

At first this seems like a really silly idea, but in fact it's hard to come up with a strong counter-argument.

Pete

It would be absolutely useless (and dangerous) to use it to navigate over Caernarfon Bar or up an East Coast creek or anywhere the topography changes after a good storm and you need to "feel" your way in.

________________________

 

[[70521]]

Why don't electronic charts display proper depth?

Unless you are trying to get up a river with 0.05m under your keel you don't need that level of accuracy.

 

[SiteSurfer]

Navionics on the iPad/iPhone/Android (other platforms are available) does a similar thing to one of your points.

You can select the day, slide across the tidal times and see the chart depths change - most useful as an aide memoir and emergency check. (That's options 2/3 in the OP)
I think these days that package counts as a chart plotter of sorts.

 

[paulburn]

It would be absolutely useless (and dangerous) to use it to navigate over Caernarfon Bar or up an East Coast creek or anywhere the topography changes after a good storm and you need to "feel" your way in.

________________________

I think your argument is an a prior one against using charts of any kind, paper or electronic, which I'm afraid I disagree with !

 

[paulburn]

So, to compare the steps needed to know the depth at any place on the chart

Using my wacky idea of chartplotters displaying proper real time depth date (combining seabed depth and tide depth)
1) look at the chart depth (instantaneous and pretty accurate)


Using current method of using chart datum (takes ? 30s if done properly each time and involves guess work)
1) look at the chart depth
2) look up current time
3) look up tide timetable and find correct entry for place and time
4) note the HW and LW times that straddle current time
5) mentally note where current time is relative to HW and LW times
6) note HW and LW tide heights
7) use (5) and (6) and mental visualization of tide curves to determine current tide height
8) add (1) and (7) together

 

[SiteSurfer]

But if you were starting from scratch, why not show real-time depths on a chart image which is being generated on the fly anyway? It's no less accurate than doing the same thing in your head based on the same data.

As a bigger boat sailor who literally did start about 4 years ago and only in the last year actually get and take responsibility for a larger boat (bigger than 12ft) then I have to admit that from the get-go I've worked with electronic charts more than paper charts - not to say that I don't understand or use paper charts - but I don't know any different and see little merit in going backwards (other than the remote possibility that all 4 devices will fail in a cascade fashion) - so for me (and I stress personal opinion) the idea of using the dynamic chart soundings seems totally obvious.

I suspect that there are several hundreds if not thousands of people just like me, I don't discount the paper charts and traditional methods of checking depth as they have worked for millennium - but I find myself literally never getting the paper charts and tide tables out during my boat usage - I have found that the dynamic data is always near as dammit ok for me.

Small caveat that I only draw 800mm so almost always CD works for me

Note: For the benefit of doubt, I'm not arguing that paper charts are a thing of the past. I've done all the chart work theory and exams - so i do know how do use them, it just seems like a royal pain in the arse when the same info is available instantly. Does that make me a 'bad' sailor?

 

[paulburn]

Navionics on the iPad/iPhone/Android (other platforms are available) does a similar thing to one of your points.

You can select the day, slide across the tidal times and see the chart depths change - most useful as an aide memoir and emergency check. (That's options 2/3 in the OP)
I think these days that package counts as a chart plotter of sorts.

So they do - I never noticed that !

 

[SiteSurfer]

I've taken to having the tidal screen open under the chart on Navionics for the tidal area I'm currently passing through which then not only gives me the (approximate) full height of the water in real time - but allows me to look ahead (in time) and double check I'm ok - its also brilliant for the tidal current/flows direction as they also show in real time on the charts too.

 

[prv]

It would be absolutely useless (and dangerous) to use it to navigate over Caernarfon Bar or up an East Coast creek or anywhere the topography changes after a good storm and you need to "feel" your way in.

But the same applies to a chart and a tide table, so Paul's idea doesn't lose anything.

Pete

 

[stevebrassett]

So, to compare the steps needed to know the depth at any place on the chart

Using my wacky idea of chartplotters displaying proper real time depth date (combining seabed depth and tide depth)
1) look at the chart depth (instantaneous and pretty accurate)


Using current method of using chart datum (takes ? 30s if done properly each time and involves guess work)
1) look at the chart depth
2) look up current time
3) look up tide timetable and find correct entry for place and time
4) note the HW and LW times that straddle current time
5) mentally note where current time is relative to HW and LW times
6) note HW and LW tide heights
7) use (5) and (6) and mental visualization of tide curves to determine current tide height
8) add (1) and (7) together

Or:

1. Look at plotter to see comparison of where you are and where you want to be.
2. Look at depth sounder to see actual depth.
3. Quick bit of mental maths to determine actual depth of where you want to be.
4. Further bit of mental maths to take time into account.

 

[prv]

Combining the two models together will compound the uncertainty in the models in a manner where they can't be seperately assessed by the skipper. That's something I'd be uncomfortable with, because assessing the uncertainty of each is part of my safe navigation.

I don't really understand what you mean about not assessing the errors separately. Can you describe the thoughts you might have while navigating in a more concrete way?

I might look at a river entrance and think "this seems like the sort of place that will shift around after a storm, and I bet it doesn't get surveyed very often. I'll add a bigger than usual safety margin, come in slower and keep a close eye on the sounder, and if I can arrange it to be on a rising tide then that would be good". Or I might think about the hot, still, settled weather and how the high pressure might be holding the water down a bit. But I can't quantify either of those things, so I just add "a bit" for safety and go cautiously. I can't see how seeing a separate figure for charted soundings and predicted rise of tide might alter my behaviour.

it's going to present you with something much more precise on the screen than a calculation in your head, leading to an apparent certainty which- in some places- may be false. Electronic charts tend to be amenable to a false sense of certainty, especially for beginners, and a product that increases that perception might also increase the number of plotter-assisted pile ups.

That's a fair point. I think most of the problem already exists though, with old and sparse surveys being converted into crisp certainty with a little animated boat driving across it. I believe inaccuracies in plotter charting (and charts generally) are much greater than errors in the tidal predictions. But I suppose a predicted current depth might induce some people to venture into complicated areas they'd normally avoid due to the effort of working out the rise of tide, and then they hit something due to chart uncertainty, not the addition of tide per se.

Additionally, in areas with large tides like the Bristol Channel, this sort of visualisation would make it difficult for a navigator to get an overview of a the seabed topography

I don't really understand why this would be the case. All the contours would still be there. If anything, the artificial distinction between dark blue and green is what distorts the overview in a place with large tides like the Channel Islands, because most of the time most of it is navigable, and some of the green is actually never exposed because the tide doesn't get very close to datum even at low springs.

It will obscure where CD is

That's the idea. Chart datum is a necessary abstraction for working with static charts, but it doesn't actually exist in the real world. You and I find it a completely natural concept because we've worked with it for decades, but if you start with a clean sheet of paper and design for someone with no prior knowledge, why would they care about it? They care what their surroundings look like now, or at some arbitrary time in the future when they're planning to be there. We have everything we need in order to show them that, and we should do so. Chart datum is an implementation detail that would be hidden behind the scenes - do you care about the Unix epoch value inside your phone?

I can imagine most chart suppliers are reluctant to become responsible for a product of this kind.

It's the plotter manufacturers who would have to implement it, not the chart suppliers. And I'm mildly surprised if they don't already offer it as an option. It's certainly more useful (and probably less dangerous due to false certainty) than the 3-d visualisations that some of the high-end plotters generate from chart soundings.

I'm not teccy enough to answer this last question but what about processor power required to constantly recalculate everywhere on the displayed chart- will it not also mean a rather expensive plotter is required?

Moore's Law has already solved that one . Modern plotters are essentially Linux PCs in a waterproof box, the performance increase over, say, a Raymarine C80 from 10+ years ago must be enormous.

Pete