The new station at Arisaig seems to have gone off line leaving a big gap in network coverage for the Highlands. It would appear Mallaig has gone completely. After correction to the RINEX header I have had excellent post process solutions when the Arisaig data has been included. It must be well positioned. I hope it gets switched on again soon.

Dear Stuart,

 Thank you for your email.

 Having checked we found that there was an error in the RINEX header coordinates for ARIS, which we have only discovered upon receiving your email.

 The correct coords for ARIS are:

 3472288.7161     -355730.7490     5320460.9994

 Please note, ARIS has an antenna height of 0.1804 metres

 RINEX data for times/dates from now will be correct, we will either remove or correct the files already sent to the website.

 We are very sorry for any problems caused by this error and trust you will find this information to be helpful.

 Regards

  Mrs Janet Keep

Customer Service Advisor

Customer Service Centre, Ordnance Survey

C454, Romsey Road, SOUTHAMPTON, United Kingdom, SO16 4GU

Phone: +44 (0) 23 8030 5030 |  Fax: +44 (0) 23 8079 2615

http://www.ordnancesurvey.co.uk/ | customerservices@ordnancesurvey.co.uk

I have been working 5 km south of Mallaig again today and I was looking forward to seeing what the RINEX data download presented this time. Alarmingly there is no data available for Mallaig so I wonder if the recent appearance of the Arisaig station is there as a permanent replacement for it. The position read in from the RINEX file and used by Trimble business centre is the same as last month and  remains about 460 mm out in easting, 940 mm in northing and 240 mm for level. There are no posts on the OS web site to hint at what may be happening and I haven’t heard anything back from my email enquiry when I first stumbled upon this issue.

Thankfully my own base station occupation time was of sufficient length to get a good solution from much longer vectors; Oban, Fort Augustus and Lochcarron.

24 March 2010
Following a small survey just south of Malaig I downloaded the appropriate geographic RINEX data for the exercise. I was surprised after Mallaig data there was data available for another very short base line to a previously unknown station at Arisaig. GPS Utopia I thought in an area of very sparse coverage historically. However post processing the vectors along with some other longer ones quickly cast doubt in the accuracy of the Arisaig station. I wonder if it has been put up on the OS site too early or by accident. The OS team, who are always very helpful on such matters, have not responded to my email as yet. Can anyone else shed any light?

Dear Mr Ewing

I feel compelled to write on the issue of how Land Registration is being handled in Scotland. In the course of my business I am being confronted with situations which beggar belief and in some cases wrecking the lives of victims.

I am based in Roy Bridge in Inverness-shire and work as a land surveyor. I am principally involved in gathering dimensional data ahead of the design of major civil engineering and building projects. We record the existing lie of the land in the form of complex three dimensional computer models so that engineers or architects can superimpose their design on it and analyse how will it interface in terms of cuttings and embankments for a road alignment or hydraulic profile for pipe lines. Projects include renewable energy schemes, associated infrastructure, industrial sites right down to the one off dwelling house. Our kit includes scientific quality GPS receivers and high end laser total stations (electronic theodolite with integrated distance measurement capability) and offers a level of fast measurement accuracy thought impossible two decades ago.

I apologise for labouring the point here but I want to emphasise just how accurately properly equipped land surveyors can work. If I measured the distance between a kerb nail in Roy Bridge and one in Fort William I would calculate not only the distance but the absolute position of each nail. Starting from first principles another surveyor repeating the exercise would achieve a solution within about 20 mm of my own. If the measurement was being presented to a client the distance and absolute coordinates would be given along with the grid system and some information about the effect of map projection and earth curvature.

As you would imagine we are occasionally involved in measuring land at the time of a change of ownership or plot subdivision. Here the surveyor leaves 21st century technology behind and unwittingly enters the fag packet sketch world of Registers of Scotland. Presented with a field which has been staked out by a farmer for a sub divided house plot our exercise could not be more straight forward. We can survey the key corner points and in addition pick up surrounding hard data to give context to the plan; typically adjacent tracks, walls and other buildings. The finished drawing would show the subjects outlined in red with qualifying dimensions and critically a coordinate table listing of each corner point. In addition, and only to give wider context, to the plot an inset plan may be included using OS map data. From the foregoing you will therefore appreciate that in the event of a boundary stake being removed or the garden fence falling into dilapidation 20 years hence a suitably equipped surveyor furnished with this plan could unambiguously restake the original boundary.

What’s the problem?

It is first necessary to understand the nature of OS digital mapping in Scotland. When OS mapping is interrogated in computer software it is possible to query dimensions and coordinates with millimetre accuracy and generally the relative position of roads and buildings “look” right if you have first hand experience of the area. However it is very easy for a surveyor to superimpose his own accurate survey data upon the OS line work for the same area. This can show a very different story. Buildings which looked okay at a cursory glance are in fact the wrong shape, size, position and orientation. Fence lines may be correct for orientation but may be displaced by four metres; yet another 20 metres away a corner in a roadside wall and road edge may be within a few inches of the true position. In other words -and you should never do this- a simple enbloc shift does not correct for a general displacement error. The OS linework can be spurious, errors inconsistent and can’t be generalised. In addition to the geometric surveying errors there are historic anomalies where long demolished features remain represented and conversely building extensions are not updated.

When our plan is passed to the client, to the solicitor and then to Registers of Scotland there is a black hole activity whereby the linework from our plan is transferred to the current OS map of that area. As I understand it the process is highly unscientific and involves more trying to make the dimensioned shape fit features on the OS map rather than relying on the surveyors infinitely more reliable coordinate data. (Its a bit like using a Stanley knife to make a jigsaw piece fit except in this case its the surrounding pieces which are wrong and should be adjusted.). The final product from Registers of Scotland is incredibly an A4 sketch at scale of 1 in 1250 at best with a thick red line representing the plot on an OS back drop. A simple house plot will be the size of a thumbnail and all the dimensional and coordinate data will have been stripped out. In engineering we would never scale from a drawing and would always rely upon figured lengths yet here we have tiny dimensionless shape where both the legal profession and Registers of Scotland are suggesting this very method of interrogation. A diligent solicitor or client may retain a copy of the surveyor’s drawing for their records should the need arise for boundary re-establishment in the future but in its absence it is almost impossible to reverse engineer the Register of Scotland sketch back into something meaningful.

I have just indicated the problem when a site has been surveyed properly. It gets worse. The above mentioned OS digital data is accessible to anyone with a cad program for a small fee. It is quite possible for an individual to procure OS data and import it into their own drawing and start applying notional boundary linework for house plots without any type of site measurement. This plan can find its way onto sellers brochures and sales particulars and without any real scrutiny to Registers of Scotland. Because the plan already looks like the OS backdrop they (RoS) have no problem digesting it and the familiar A4 sketch duly pops out of the system. All this can happen without a theodolite going near the site or a peg being put in the ground and there is no guarantee that the OS backdrop is remotely accurate. I have seen one such exercise recently with an architect adopting this desktop approach to subdivision resulting in existing house A’s septic tank being sold with the land belonging to new house B. I accept this is not a RoS issue but it does illustrate the sheer folly of drawing lines on OS maps rather than visiting site and doing an accurate feature survey. In this situation neither party was happy and were being faced with unnecessary remedial legal costs.

Elsewhere a client from Speyside required a land registration survey plan for a considerable area of decrofted land. I visited the site and the new fencing survey was going to be technically difficult because of the surrounding dense woodland and the resulting fee quote was high. She was however in possession of an agricultural department sketch showing the approximate position of the new fences. Because the 1979 Land Registration Act simply requires the plan to tie in to (look like) OS the client’s solicitor urged me to crudely plagiarise this sketch. I refused of course but I did delve into the origins of the agriculture sketch; I spoke to the man who did it; and I was able to determine that it was wholly unreliable in nature. I suggested to the client that if it was that simple then the solicitor should just do it. To the best of my knowledge this is what happened.

While the solicitor above was clearly trying to cut cost for the client I discussed the need in my view for a reliable measured survey and my reasoning clearly fell on deaf ears. Since then I have regularly had the opportunity to discuss the lack of traceability with RoS sketches with solicitors involved. Responses range from ” we know its rubbish but we’re stuck with it now” (the sentiment if not the exact words) to practitioners with an absolute lack of spatial awareness or the slightest concern about the integrity of RoS; or any other boundary plans crossing their desk. A cynic may observe that such a poor system creates work for the legal profession. The sheer apathy surprises me.

There is a second issue in that historic documents from Sasine days are inherited and have to be used at the time of first RoS registration. Here again we have potentially good data disappearing into the black hole and it is a concern that someone at a computer screen in Edinburgh can somehow assemble the sketch without viewing the boundary dimensions in site context. Instead of looking on site for evidence of the boundaries the operator comes up with an artist’s impression by trying to force the sizes to fit the OS plan which we already know could be metres wrong. He, in all probability, has just created his next boundary dispute. I have recently had cause to query such a sketch and unfortunately the subjects had been conveyed twice in RoS times. Because of this it appears even more difficult to investigate exactly how the RoS sketch had been assembled. No notes are recorded and there is no traceability.

I have obviously spoken to RoS people several times and have written at least twice recording my concerns. I explain what we (any land surveyor) can do with our survey equipment and about the clearly annotated unambiguous plans that we produce. I express my frustration at reliance on woeful OS mapping and the tiny dimensionless sketches that they produce from our data afterwards. Sadly some of them just can’t grasp what I am getting at. The line that comes up repeatedly is that if they match it up with the OS map they have complied with the 1979 Land Registration Act. Beyond that they don’t care. From their sketch I ask them how I can tell what the width of a plot is or what the included angles measure around a site or fundamentally tell me the coordinates of the corners. They suggest scaling where the line thickness is the width of a car. I ask then for them to refer to the data that was used behind the scenes to create the sketch in the first place; “if you write in we may have something but there would be a fee”. Sitting side by side with this ambiguity is the RoS indemnity scheme guaranteeing title to the land on the sketch. I can only compare this to a bank offering security of your savings where they say your savings are safe with us but can’t actually say how much you have in your account.

Fundamental in engineering and surveying is traceability. If I open up a plan from a total station survey carried out four years ago and had a question about one observation taken to an inspection cover I could quickly retrieve the station (precise position) that the instrument had been set up on, the physical height of the telescope above the station, the horizontal angle, vertical angle, slope distance and pole height at the prism, the name of the surveyor, the date and time, to the second, that the reading was taken. This is the raw data that goes on behind the scenes to build up a survey model and it is preserved and archived for quick reference. If a client wanted this data perhaps simply to test my own traceability or for other technical reasons they could have it that day free of charge. Despite living in a digital age there is no attempt to preserve traceability at RoS, quite the opposite in that if you want information you will write in and pay for it and there is no assurance that they will find anything. On the input side it is also clear that RoS will accept anything that that looks like it has been drawn on an OS backdrop and do not press for any survey data to support the integrity of what they are about to digest.

Until it bites them hard clients live in blissful ignorance thinking that their land registration certificate gives them perfect and unambiguous peace of mind of exactly what they own. Sadly the ambiguity of the RoS sketch can also be used by unscrupulous neighbours (and former friends) to block a house sale effectively using a flawed plan as a ransom note. I have seen such victims in tears.

In all my negative comments above you would hope that I have some suggestion and it is very simple. The OS digital mapping, in spite of promises of impending improvement, will never come close to the public’s expectation of how tightly their boundary should be recorded. Rather than being used as the absolute basis for the recording of subjects it can still be used to give broad context to a property location. The current style RoS sketch can be used to illustrate that the property is say opposite the boat yard in Mallaig.

As described earlier this should be an inset plan on a much larger scale detail plan, say at 1 in 200 scale, properly executed by GPS and Total Station. Corner coordinates should be annotated, perimeter and bracing dimension should be given and descriptions of what each boundary is. The name of the surveyor, date, technical notes on instruments used including calibration certificates should all be recorded.

In other words it should be precise, transparent and traceable. Everything that the current system is not.

In a Utopian state all submissions could be merged, paper free, onto one digital master plan and be left in the public domain. Data recorded this way handed to our counterparts in the future could be re-established easily avoiding the nonsense which currently prevails.

Until you are exposed to the RoS system a few times it is difficult to believe how unsatisfactory it is; possibly the worst in the developed world. It must be overhauled and brought into the Twentyfirst Century. Right now we have the survey resources and technical ability to make this happen but apparently no political will.

I look forward to your comments.

Yours sincerely

Stuart Ross