|
Post by Bruce Speirs on Sept 29, 2014 22:42:20 GMT
Alex has challenged me to comment on the RSC 2010 review due in 2015. It may be pedantic, but it is not a case of having rules to rigidly follow the letter of the law, it is more of a case of getting fair and flexible rules that are promulgated in terms of the Act. If the High Court action taught us anything, it is that you have to understand the structure of something before the detail will make sense. In that light it is clear that a new structure for the RCS is required, where these will be in three sections. 1. Cadastral Survey Rules - generally being the spatial elements of Cadastral Surveys in terms of s49(1)(a) & s7(1)(c) of the Act, 2. Cadastral Survey Dataset Rules - generally being non-spatial items related to Cadastral Surveys in terms of s49(1)(b) of the Act, and 3. Cadastral Survey Dataset Lodging Rules - being lodgement elements in terms of s9 of the Act. Please remember that what Cadastral Surveyors generally refer to as the CSD is in fact the dataset of Cadastral Survey (and which the Cadastral Surveyor certifies), and this is different from the legal definition of the CSD as set out in s4 of the Act. From the CSD diagram shown in the ICS Cadastral Surveyors certification project (under the resources tab of the ICS website) we know the structure of the Cadastral Survey and Cadastral Survey Dataset, and there is basic agreement with LINZ on the Survey and Tenure System information listed. After that, it is down to the specific rules required to achieve the purposes of the Act, and I will leave that to others at this time.
|
|
|
Post by Bruce Speirs on May 24, 2016 23:38:23 GMT
This project ports the RCS 2010 into a format that looks something like the requirements of the Cadastral Survey Act, with a number of alterations, additions, and omissions, to bring them into the real world environment. The documentation is under the legislation tab and open for comment as to structure and content.
(Check the "Legislation" tab RCS 2010 Project on the ICS website for the documents.)
|
|
|
Post by Pat Sole on Jun 12, 2016 7:38:29 GMT
Good stuff Bruce - at least you are giving this the priority it deserves. I have looked over your drafts and at least the rules in the format you have arrived at, make for much easier reading. However, I realise you have kept much of the content from the existing rules - I personally would like to challenge much of that content including accuracy standards, use of redundant terms (eg defined by survey), not showing old marks as old etc. I am finding that we are spending as much or more time on integration issues, as in capture of the dataset itself. Cheers
|
|
|
Post by Alex Liggett on Jun 12, 2016 20:33:37 GMT
What would you change about the accuracy standards?
|
|
|
Post by Pat Sole on Jun 13, 2016 5:56:41 GMT
Pretty much everything Alex. If we have learnt nothing about Cadastral Surveying after the Chch earthquakes, it is about the importance of survey marks and their accurate placement. (And analysing all the data, not just measurements/adoptions). When you can place rural boundary pegs out of the car window (even if they are only a metre apart, the nearest 0.1/0.2 will easily meet the rules requirements because of the large constants, even more for Class C) then boundary position is not readily preserved. Least squares is a great tool for analysing networks - but the reality is that most cadastral surveys dont benefit from it, and why make things more complex than neccessary. With the inaccuracies caused by constant earth deformation, earthquakes, human disturbance of marks, measurement errors etc, there are not that many cadastral surveys that benefit from such rigor (even though LINZ can produce it). The theoretical boffins have taken over! Bring back the simply RF if you ask me. Easy to think about/analyse in the field or office - is technology independent. The other thing the current accuracy standards produce is an apparent degradation of lots of quite good old work (eg good theodolite traverses for miles through bush boundaries 1:1500 close stuff end up Class C).
|
|
|
Post by Alex Liggett on Jun 13, 2016 9:36:19 GMT
The theoretical boffins have taken over in exchange for giving us Landonline. To some extent this gives the lie to the 'technology independent' RCS 2010, but the reality is that Landonline needs an RMS-based error standard, because like it or not it is a network.
Funnily enough, in 2010 or 2011 I would have agreed with you, but now I am reconciled to the new way. Having worked both in the earthquake-affected area and on very large rural surveys I pretty much don't have a problem with the form of the standard, although I would agree that the constant term of the Class B standard is a bit slack at 0.2m (I believe the logic was 'the size of the top of a post').
I'd be interested to know what the spread of views is on this, but in any case, I would offer pretty low odds that this will be on the table in the review, so you may end up having to just print yourself a relative accuracy standard 'ready reckoner' for the field and move on.
|
|
|
Post by Alex Liggett on Jun 13, 2016 9:46:26 GMT
....For example, section 3 reintroduces the requirement for cadastral surveys to be undertaken using good survey practice.... Further to this Bruce, do you think a definition for good survey practice might be required?
|
|
|
Post by Bruce Speirs on Jun 14, 2016 4:45:16 GMT
Pat, thanks for the favourable comment regarding the structure of RCS 2016 project.
Like Alex, I don't have a problem with the way the current accuracy standards are expressed. I am generally only concerned with the non-boundary accuracies in the field, and I have placed the graph of these in my field book (along with those for 1972 & 1998 / 2002 regs) for quick reference.
If you want an approximate RF for the non-boundary accuracies, I suggest you use; 0.025m for distances up to 100 m, 0.020 m + 0.005 m per 100 m from 100 - 300 m, and 0.005 m + 0.010 m per 100 m from 300 - 1000 m.
Your comment on integration is interesting. Integration is not in fact part of a cadastral survey (nor indeed a CSD) and is subject to different cost benefit analysis to the cadastral survey itself. Costs of integration should be spread to a wider grouping than our clients, and this raises the possibility of our clients being recompensed for this work.
Your comments regarding the Christchurch earthquakes goes to the heart of the witnessing requirements of regulations (both number and distance). Personally I am glad the RCS 2010 had not been in effect for very long before the earthquakes, and the non-boundary mark density they allowed had not yet decreased the number of witness marks significantly.
Alex, regarding your query do we need a definition of good survey practice. I think the time has probably come when we do, as both LINZ and a number of cadastral surveyors in Christchurch forgot what this means when the going got tough. Perhaps the NZIS / ICS working party would consider generalising their recent practice guidelines for that purpose.
|
|
|
Post by Brent George on Jun 14, 2016 22:44:55 GMT
Of course any accuracy standard (or value) does not need to be considered a target. If it is considered as an indicative limit it should not restrict the ability of the surveyor to complete their work with greater precision and leaving a record that work accordingly.
Like the mantra on speed limits - "the speed limit is a maximum - not a target"; a comparable maxim in this case could be: "the accuracy standard is a limit - not a target"....
|
|
|
Post by Alex Liggett on Jun 14, 2016 23:14:45 GMT
Of course any accuracy standard (or value) does not need to be considered a target. If it is considered as an indicative limit it should not restrict the ability of the surveyor to complete their work with greater precision and leaving a record that work accordingly. None of us is worried about our own work, just others taking advantage of the slack tolerances.
|
|
|
Post by Pat Sole on Jun 20, 2016 23:47:45 GMT
All good comments above. In terms of accuracy standards though Alex, you disagree but then question the "constant" size - get rid of the constant size and you are left with a distance related term which approximates an RF. Bruce, you are comfortable with them but do not use all the boundary (only traverse) standards in the field - I suggest that when you are miles from anywhere it would be good to know (and have very simple tests) that your survey was compliant before you leave the field. I largely ignore all the standards in the field knowing that good survey practice and well adjusted gear should easily satisfy them - but not for old work. I know LoL is a network, but we have to be careful that what we set as rules for cadastral survey are not dictated or a requirement of the system/process, rather than that required of a survey. I'm comfortable that LINZ can run tests on my work in whatever manner they wish but do not want LoL to further dictate what and how we capture data - this leads to the burden of more and more integration work (residue parcels/node capture etc) and maintenance of the cadastre which is a government function (not our clients).
|
|
|
Post by Alex Liggett on Jun 21, 2016 3:19:04 GMT
All good comments above. In terms of accuracy standards though Alex, you disagree but then question the "constant" size - get rid of the constant size and you are left with a distance related term which approximates an RF. Yes, I said that the Class B constant term was a bit slack. Actually I had it wrong, as it is 0.3m not 0.2m - the perils of working from memory. I would sooner see it reduced from 0.3m to 0.15m. There is 0 chance of going back to an RF based standard, and in my view a minimal chance that any change to the accuracy standards will be considered.
|
|
|
Post by Pat Sole on Jun 22, 2016 2:23:55 GMT
Heres a few more issues to keep the discussion going: (1) Why do we have to capture two vectors for every point? (discourages good survey practice by hooking onto additional work, otherwise clutters the data with meaningless calculated lines, but satisfies least squares rules) (2) Why is it not mandatory to lodge field records/calc sheet etc of water (natural) boundary fix?(these boundaries can never be plotted with any accuracy other than the accuracy from which they are digitised into Landonline otherwise)(3)There is an important distinction between marks that are disturbed and are unreliable (we can no longer show this and make the jobs of surveyors more difficult) and why show an old mark as "new" just because it is "disturbed" (fine with relabelling it but should be shown as old)(4)Why should we have to recapture easements every time!
|
|
|
Post by Brent George on Jun 22, 2016 2:37:46 GMT
I agree that the distinction disturbed vs unreliable is important and should not be lost. For goodness sake we have been using those terms successfully for a much longer time than LandonLine has been in existence, yet they were sin-binned in order to accommodate their application within LoL.
Be thankful Pat that you don't have to deal with UNPROVEN as well. This term and usage/mis-usage is a blight on surveys all over Christchurch these days!!!
|
|
|
Post by Alex Liggett on Jun 22, 2016 3:36:52 GMT
....For example, section 3 reintroduces the requirement for cadastral surveys to be undertaken using good survey practice.... Further to this Bruce, do you think a definition for good survey practice might be required? Good Survey Practice - in response to my own question to Bruce above, how about this as a starting point for discussion: From "Cadastral Survey Guidelines V4.2" LINZ OSG - April 2004.
|
|