ERRORS ON MARINE NAVIGATIONAL CHARTS
Copyright 2009, 2011
A discussion about chart errors. Paper vs. electronic. Raster vs. vector.
STANDARD DISCLAIMER - This is not a comprehensive guide.
They may be errors or omissions.
It is only my personal take on what some of the dangers are when using charts.
Use your own common sense when making decisions.
Paper charts can contain three main types of errors:
1. Positional errors
2. Relational errors
3. Time errors
1. Positional errors
A chart may well have been made before GPS.
Some even over 100 years ago.
The position it says you are at may be off even a number of miles.
2. Relational errors
The objects on the chart may not be correctly positioned in proper relation to each other.
3. Time errors
This is the real gotcha.
Things have changed since that chart was made.
Shoals have built up, rivers moved, etc.
Two types of charts:
1. Paper Charts
With paper charts you will not automatically know where you are.
You have to use various means to determine that.
Like taking compass bearings from shore objects.
This may seem like a disadvantage, and to some extent it is.
But if the chart is off, then let's say you use a GPS to get your exact position.
You plot that on the chart.
Since the chart is off, you have now plotted your self in the wrong position in relation to the chart.
That's the trick.
You don't care exactly where you are, you only care where you are in relation to the shoals and rocks.
That's how you use paper charts.
If there are any relational errors, they probably are not a problem if you are sticking to main channels.
These charts were most likely done by mariners who made sure the important stuff was right.
So, relational errors are usually only a factor if you are approaching a shore that a ship would not normally approach.
Then you use a little more caution.
But you always have to keep in mind that the bottom may have changed.
Someone may have put in a new jetty.
A new light may have been installed.
This is why it is critical to keep up-to-date charts.
The positional accuracy may not be improved in the latest chart, but the new light should be there.
2. Electronic Charts
Electronic charts are mostly made from the paper charts.
Even the pretty vector charts.
So they have all the same possible errors as the paper ones!
But they have more.
As stated above, if the chart is off positionally, the GPS will faithfully plot you within 10 meters of the exact place on Earth where you are.
But since the chart is off, you know exactly where you are by the numbers, but you don't know where the underwater rock is by the numbers.
You only know where it is in relation to the point and the island.
See more errors in the raster/vector discussion below.
If you are going to use electronic charts, you MUST know what the possible errors are and how to deal with them.
The only time you can totally trust a GPS is:
You have been that way before and saved the track.
While saving the track, you kept watch on what the GPS is reporting as its accuracy.
You also have a 3d type depth finder so you know that not only did you not run into anything, but it was clear for a reasonable distance on either side (or you know the waters well enough).
While you follow the track back out, you keep an eye on the GPS reported accuracy to make sure it is capable of keeping you close enough to your track to be safe.
In other words, the only time you can totally trust a GPS is when you are using it independantly from the charts.
You can trust a GPS when comparing it to data from a GPS (or to itself), but not in relation to a chart.
Note for the following section: There ARE some charts made by modern methods using GPS positions that are very accurate. These are mainly for the main shipping channels.
Electronic Chart Types:
These are made by scanning a paper chart.
It is actually just a picture of a paper chart.
It has all the errors that were on that paper chart.
Also the position encoding can be in error or a programming bug can cause the plotter to display it in the wrong position.
This is not as uncommon as you might think.
I have seen charts plotted at 45 degrees to what it should have been.
What's dangerous is when it's plotted just far enough off to be a danger, but not far off enough to be obvious.
These are also made from paper charts.
They have all the errors of the paper charts plus all the errors induced by electronics.
They are basically had traced on a computer.
That gives it the added possibility of human error.
The error that occurs the most is that an object is simply missed and doesn't appear on the vector chart.
Another problem with vector charts is that when you zoom in, the lins still look sharp and clear.
On a raster chart, it gets grainy and blury so you know you are zoomed in and detail is questionable at best.
On a vector chart you must be very aware of your zoom level.
This can be further masked when you zoom back and it jumps to a lower resolution chart and you may not notice.
From the above you may think I don't like electronic charts.
Far from it.
I have Nobeltec Admiral, SeaClear and OpenCPN to name a few.
You probably got to this page from my page that has links to where you can find free electronic charts (here)
My point is that electronic charts can only be used safely if you have a good working knowledge of how to use paper charts, and you know what the additional problems are that creep in when adding electronics into the mix.
I STRONGLY believe this!
I won't go into the possibilities of electronics failure, etc. We'll leave that for another discussion.
Another topic - Other chart errors and where they came from. Raster/Vector/Both.
Some people point out that the information on electronic charts came from the paper charts originally, and others say the paper charts are now all produced from digital information. Each statement trying to prove some point about chart accuracy.
So, let's demystify some of this, mystify some other stuff, and basically just veer off into some wild speculation.
I don't know what the different information flows are, so mainly I'm just speculating on things, but I think overall it will show how complex the situation is and how, no matter what, you can't blindly trust any of it. Errors can creep into one format that doesn't creep into the other. Neither format avoids all the errors that can occur in the other. And there are many errors that can occur in both.
I'm starting this in the sextant era.
Captains got positions by sextant and kept notebooks. Charts were drawn using these notes. Sometimes a mapmaker came along. They may have taken these positions and plotted them right there. In some cases, they went ashore. There they could get better fixes and do averaging, sometimes getting quite accurate. Then, using standard surveying techniques, they could very accurately plot the relation between various land features. They might have a dinghy hold a position over and underwater feature and plot its position from land. In general, the relation between points in an area would be more accurate than the positioning of the overall chart on the earth. Local magnetic changes could have them skew the chart. Inaccurate sextant readings could have them plot the location off as much as a coupla miles.
Along the way, governments started collecting these notebooks and charts and created their own database. Then some mapmakers might go through this to create a more uniformly drawn and notated set of charts. When they found multiple sources of information that disagreed with each other, how they resolved those differences is something I'm not even going to try to guess at.
Various governments might form alliances and share information.
In all this combining of information you'd hope it resulted in an increase of accuracy, but I'm sure errors crept in and politics played a part.
Then along came computers.
The first digital information surely came from someone manually plotting the points and typing them in. Then came digitizers. Along the way, finally, you'd have a database of points. But these would first be for the most used charts. So, charts would have been digitized at different time periods and using different methods.
New areas being charted might have been dealt with in different ways. Early on, some may still have been done as paper first. As time went by, they shifted to going on a computer first, then hand plotted onto paper charts. Eventually we get to the point where it's totally digital. GPS positions fed directly to a computer and travels through various processes and end up on charts.
Now, in what order did the following happen?
*Paper charts are created from the digital data.
*Paper charts are scanned for use on computers.
*Vector charts are used.
*Paper charts are created directly by the computer
I don't know, but none of it was where a switch was flipped and everything started happening the same way. It was an organic process with multiple methods being used simultaneously.
So, what do we have now?
We have charts in various formats. Some of the features on those charts came from modern times. Some came from the 1800's notebooks. Some came from GPS plots.
Some had errors introduced when they were plotted by hand. Some had errors introduced when some computer conversion routine mishandled an overflow condition when sending it to the plotter. Some errors are introduced by the software that displays it.
Some of these errors are common to both formats (raster/vector). Some only happened when the paper chart was produced, and others only when the vector data was compiled.
So, what is this diatribe is meant to communicate?
Saying all the data came from paper charts is way over simplifying the situation. And saying paper charts now all come from computers also way over simplifies it. They are basically meaningless statements and in general, don't prove what the person was trying to prove.
It also shows that the data has made convoluted, error prone routes to your screen, no matter what the format they end up. And that some errors may be different between the formats.
Once again, it boils down to not trusting one single method of navigation. In electronic navigation, I want multiple sets of charts, the more the merrier. I want multiple sets (and types) of hardware. I want multiple sets of software and operating systems. Now, this may simply mean a computer and a chartplotter, both with their own sets of charts. For myself, I also want a set of paper charts. But I can totally respect someone who is fine relying on a good electronic setup.
When it comes to navigation, many many things cannot be said as an absolute. And this includes many things that on first glance you may think can be said as an absolute.
Please email me with any comments.