OSBG36 V WGS84 big position error - what have I missed!

[ianj99]

I fitted an inexpensive tracking & alarm unit to the boat (a TK103b from Amazon). It texts me if the boat is moved, the engine started or the battery is low and is ridiculously cheap at £38.

The boat's position is correct to within 2metres when I check it using Google Maps. (the tracker has its own gps antenna)

The position given by both chart plotters is different - the boat is approx 10Miles south east of the tracker position according to Google Maps, but of course its correct on the UKHO & Cmap plotter charts.

Why?
( I thought the difference between WGS84 and OSGB36 was only metres so I must be overlooking something obvious - like units)

It is units: The tracker uses degrees and decimal degrees and the charts use degrees, minutes and decimal minutes but both are displayed in the same format on my laptop.

Tracker lat: 56.451537 decimal degress, plotter lat: 56.27.089' degrees. minutes.decimal minutes, (not seconds)
http://www.csgnetwork.com/gpscoordconv.html

 

[alanch]

I have never come across decimal degrees or decimal minutes in a lifetime spent at sea. Always degrees, minutes, seconds. Is it that common on small boat equipment to see these decimals quoted?
I can just imagine the confusion in a SAR situation.

 

[ianj99]

I have never come across decimal degrees or decimal minutes in a lifetime spent at sea. Always degrees, minutes, seconds. Is it that common on small boat equipment to see these decimals quoted?
I can just imagine the confusion in a SAR situation.

I agree it does cause confusion and might explain that often in a SAR event, those in need of help can't be found at the location they gave.

I think DMS is obsolete, but it would make sense to make D.M.M obsolete as well so there is no confusion. (all my nav software and apps use decimal minutes)

 

[RichardS]

I agree it does cause confusion and might explain that often in a SAR event, those in need of help can't be found at the location they gave.

I think DMS is obsolete, but it would make sense to make D.M.M obsolete as well so there is no confusion. (all my nav software and apps use decimal minutes)

But isn't DMM degree, minutes, decimal minutes which is what we all use these days?

Richard

 

[ianj99]

But isn't DMM degree, minutes, decimal minutes which is what we all use these days?

Richard

Yes, we do use D M.M but either this or decimal degrees, but not both. Google won't accept D.M.M or D.M.M'

 

[NorthUp]

I agree it does cause confusion and might explain that often in a SAR event, those in need of help can't be found at the location they gave.

..

I am told that reading the cursor position off the chartplotter, rather than the ship position has happened occasionally.

 

[NorthUp]

I have never come across decimal degrees or decimal minutes in a lifetime spent at sea. Always degrees, minutes, seconds. Is it that common on small boat equipment to see these decimals quoted?
I can just imagine the confusion in a SAR situation.

What charts are you using?

 

[RichardS]

Yes, we do use D M.M but either this or decimal degrees, but not both. Google won't accept D.M.M or D.M.M'

That's really odd as I use Google Earth which uses DMM - degrees, minutes, decimal minutes - which ties in with both my Navionics and OpenCPN.

I assume you're using Google Maps and if that won't display DMM I'm gobsmacked as DMM is surely the "standard" these days.

I'm still not sure why you want to make DMM obsolete for the sake of Google Maps as it appears to me that Google Maps is the odd-ball but never mind.

An interesting discussion though.

Richard

 

[ianj99]

That's really odd as I use Google Earth which uses DMM - degrees, minutes, decimal minutes - which ties in with both my Navionics and OpenCPN.

I assume you're using Google Maps and if that won't display DMM I'm gobsmacked as DMM is surely the "standard" these days.

I'm still not sure why you want to make DMM obsolete for the sake of Google Maps as it appears to me that Google Maps is the odd-ball but never mind.

An interesting discussion though.

Richard

I've tried Google maps with DMM and it won't recognise it, but does work with DD. What format did you use for DMM?

Clearly the tracker is the oddball and no way to change it from decimal degrees so in the unlikely event that I need to locate my stolen yacht, I'll have to remember to use an app to convert it to DMM. (My tablets gps is also DMM)

 

[RichardS]

I've tried Google maps with DMM and it won't recognise it, but does work with DD. What format did you use for DMM?

Clearly the tracker is the oddball and no way to change it from decimal degrees so in the unlikely event that I need to locate my stolen yacht, I'll have to remember to use an app to convert it to DMM. (My tablets gps is also DMM)

Just looked at Google Earth (which uses the same maps as Google Maps) and it displays 52° 10.307 N' 1° 52.255' W near my house.

Also looked at Google Maps and although it appears to accept DMM in this format as an input to locate a place on the map, it will not output this format when you select a location but only gives Degrees, Minutes, Seconds or Degrees, Decimal Degrees. Unless I'm missing something, that's ridiculous!

Richard

Just to clarify, to input into Google Maps you use N52 10.307 W1 52.255 in the Search box (or you can put the N and W after the numbers exactly as in Google Earth)

 

[AntarcticPilot]

To answer the OP's question, there is a substantial difference between OSGB36 and WGS84; the difference is up to (from memory) about 100m.

More detail!

OSGB36 is originally a datum entirely based on the survey network of the UK, as computed by a massive computing task completed in (surprise, surprise!) 1936. It was calculated using mechanical calculators, and the task was too large to be completed in one piece, so the country was in fact divided into two, roughly along the line from the Tees to the Exe. Not surprisingly, the lines of equal difference between OSGB36 and WGS84 often bend sharply across this boundary. There are also (rare) inconsistencies in OSGB36, such that if you compute a position starting at one survey node, you get a different answer than if you start from a different one.

The difference between OSGB36 abd WGS84 can a) be quite large and b) vary pretty irregularly.

OS maps and products such as charts based on OS datums were, until the mid 1990s all on this datum. The OS then moved to an ellipsoidal datum compatible with WGS84, but the process of converting from old to new was still continuing in the early 2000s. All modern OS based products are essentially based on a version of WGS84.

There were some interim solutions used, and OSGB36 MAY refer to one of these as well.

 

[ianj99]

To answer the OP's question, there is a substantial difference between OSGB36 and WGS84; the difference is up to (from memory) about 100m.

More detail!

OSGB36 is originally a datum entirely based on the survey network of the UK, as computed by a massive computing task completed in (surprise, surprise!) 1936. It was calculated using mechanical calculators, and the task was too large to be completed in one piece, so the country was in fact divided into two, roughly along the line from the Tees to the Exe. Not surprisingly, the lines of equal difference between OSGB36 and WGS84 often bend sharply across this boundary. There are also (rare) inconsistencies in OSGB36, such that if you compute a position starting at one survey node, you get a different answer than if you start from a different one.

The difference between OSGB36 abd WGS84 can a) be quite large and b) vary pretty irregularly.

OS maps and products such as charts based on OS datums were, until the mid 1990s all on this datum. The OS then moved to an ellipsoidal datum compatible with WGS84, but the process of converting from old to new was still continuing in the early 2000s. All modern OS based products are essentially based on a version of WGS84.

There were some interim solutions used, and OSGB36 MAY refer to one of these as well.

Thanks for the heads up.

Must have been a nightmare for the OS. Bet they didn't talk to the UKHO to try and get an interoperable system.

 

[AntarcticPilot]

Thanks for the heads up.

Must have been a nightmare for the OS. Bet they didn't talk to the UKHO to try and get an interoperable system.

For the UK, UKHO has always used whatever the current OS datum was. So interoperability was never a problem. The problems arise because different datums were used at different times, and of course the decade from 1995-2005 (approximately) was a transitional period where EVERYONE had problems. The worst difficulties were, in fact, in civil engineering and construction because existing site plans were on one datum and new maps on another. Despite the OS' best efforts to provide tools to allow migration of data from on datum to another, there always ended being up a level of manual reconciliation required at the end. I was once on the judging panel for a technical excellence award at a trade show, and we gave the prize to a company that had a good solution to exactly this problem! But it still required human intelligence in the loop. Sadly, people able to perform such tasks are also people who get bored very quickly doing them!

You get MUCH worse problems elsewhere in the world, where the local datum is an astronomical fix, with an accuracy of several hundred metres. And I know of at least one very nice map which is based on two different (and incompatible) datums in different areas of the map - the mismatch is lost in a a large glacier!

 

[richardbrennan]

Not quite the same problem, but having chartered a boat in New Zealand a couple of years ago, we could not understand why the handheld GPS was putting us in the middle of a car park in Auckland until we realised one had to count down rather than up the chart! Of course, I am sure no other forumites would have made such a stupid mistake!!

 

[ianj99]

Not quite the same problem, but having chartered a boat in New Zealand a couple of years ago, we could not understand why the handheld GPS was putting us in the middle of a car park in Auckland until we realised one had to count down rather than up the chart! Of course, I am sure no other forumites would have made such a stupid mistake!!

Yeah, right!

 

[AntarcticPilot]

Not quite the same problem, but having chartered a boat in New Zealand a couple of years ago, we could not understand why the handheld GPS was putting us in the middle of a car park in Auckland until we realised one had to count down rather than up the chart! Of course, I am sure no other forumites would have made such a stupid mistake!!

Imagine the potential for problems in the Ross Sea on a map of Antarctica in the conventional Greenwich meridian is UP orientation! South is up, and west is right, and longitude counts down in both directions from the middle if you forget the sign convention! Oh, and "West Antarctica" is East of you and "East Antarctica" is West! Just to put the cap on it, it also has one of the largest vertical deviations from the WGS84 vertical datum in the world, so you're apparently sailing about 70m below GPS sea-level