GNSS check shots conflict with 2009 cadastral coordinates
An 18-lot residential subdivision in Rolleston is being surveyed for CSD lodgement. Existing boundary marks were coordinated in 2009 in terms of NZGD2000 and NZTM2000. Your GNSS-RTK check on two existing marks gives differences of 0.045 m and 0.090 m from their Landonline coordinates. A closed traverse around the block miscloses 0.028 m over 1.4 km. The developer has a fixed-price contract and wants the CSD lodged this week. A neighbouring surveyor says the 0.090 m is 'just projection distortion' and should be ignored. The site is flat and there are no obvious mark disturbances. You need to decide whether to hold the old coordinates, adopt your GNSS positions, or run a local adjustment.
1. Assess the likely causes of the 0.045 m and 0.090 m differences and decide what evidence you need before adjusting.
Read the worked answer
Do not accept the 'projection distortion' explanation without testing it. NZTM2000 does have scale distortion, but it is systematic over the project area, not a random 0.090 m on one mark and 0.045 m on another. A systematic projection effect would affect both marks in a similar direction and would be handled consistently in the coordinate system. The more likely causes are mark instability, a coordinate error in the 2009 dataset, a control problem, or a GNSS setup/centring error. Check the marks physically for disturbance, check the antenna height and centring, repeat the observations at a different time, and connect to reliable NZGD2000 geodetic control rather than only to cadastral marks. Compare the 2009 survey report and the original CSD. Run a least-squares adjustment with realistic weights and examine residuals. If one mark has a large residual and the other does not, treat the suspect mark as disturbed rather than forcing a local shift that moves the whole subdivision.
2. How should the adjustment and CSD coordinates be handled to meet cadastral accuracy requirements?
Read the worked answer
Use a proper adjustment, not a simple shift. Hold the reliable control and any marks that pass the disturbance checks; weight the GNSS observations according to their actual precision; and allow the suspect mark to move only if the evidence supports that. If the 0.090 m mark is unreliable, do not hold it fixed. If it is reliable, investigate why your GNSS position differs before rejecting it. The CSD coordinates must meet the accuracy requirements of the Cadastral Survey Rules 2021 and the Surveyor-General's standards, and the survey report must explain the adjustment, the marks held, the residuals, and any mark treated as disturbed. A local adjustment that hides a 0.090 m problem is not acceptable. If the traverse misclose is within tolerance but the mark differences are not, the mark differences govern the investigation. Document the decision so that a later surveyor can follow the reasoning.
3. The developer demands lodgement this week. What do you tell them, and what contract and risk issues arise?
Read the worked answer
Tell the developer that the CSD cannot be lodged until the coordinate discrepancy is resolved to a defensible standard. A one-week delay is preferable to a dataset that may be challenged or require correction later. Explain the specific risk: if a boundary mark is held on a wrong coordinate, the new lots may be staked in the wrong place, affecting titles, fences, services, and building platforms. Check the contract for the scope of investigation, the client's obligations to provide information, and any agreed timeframes. If the extra work is outside the fixed price, issue a variation for the additional field and adjustment work. Record the client's instruction and your advice in writing. If the client refuses to allow the necessary work, decline to certify the CSD and explain the consequences. Notify your professional indemnity insurer only if a claim or dispute is threatened, but keep your file complete.