LINZ has identified about 300 disturbed geodetic marks, including about 30 'intermediate' positions for marks that were observed in different positions between earthquake events.

1.1. Key Points

  • Where a geodetic control mark has been disturbed, for example by liquefaction, a new geodetic control mark is created in Landonline to reflect this disturbed position, where it has been assessed that the mark is still useable
  • A geodetic control mark is classified as disturbed if it has moved by more than the tolerance permitted by the Rules for Cadastral Survey 2010 (Rules), relative to nearby undisturbed geodetic control marks as determined by post earthquake geodetic survey observations
  • Some marks have several disturbed positions, to reflect movements after two or more of the Canterbury earthquakes
  • Newly measured observations should be linked to the most recent disturbed mark
  • Adopted observations should normally be linked to the original mark, unless adopting from a CSD which itself connected to a disturbed mark

1.2. Why We Need to Identify Disturbed Geodetic Control Marks

The Canterbury earthquake sequence resulted in the disturbance of a substantial number of geodetic control marks, due to shallow surface deformation effects such as liquefaction and lateral spreading. Landonline needs to account for this in a way that ensures the integrity of the geodetic system, complies with the Rules for Cadastral Survey 2010 (Rules) and provides clarity to users of the system, particularly cadastral surveyors. The focus here is on horizontal disturbance, as this is the primary concern in a cadastral survey context.

In addition to adding new surveyed coordinates, the December 2013 Landonline update added new calculated coordinates to existing Landonline nodes to reflect the modelled deep seated movement. Where there was evidence of disturbance, or in areas where the model was considered less reliable, the order of the coordinates was also downgraded. The update did not create new nodes, as would normally be the case for disturbed marks, as at that stage LINZ had not determined which coordinate changes were due to local disturbance, and which were due to deep seated movement.

Adding disturbed marks to Landonline and the Geodetic Database ensures that Cadastral Survey Datasets (CSDs) with a mixture of measured and adopted work can appropriately connect to geodetic control marks.

Disturbed marks are often are not suitable for future use as reference marks, however in this case the marks are deemed to be disturbed as post earthquake observations indicate that they have moved by more than the accuracy tolerance, after taking into account deep seated movement. In most cases they still appear sound and will still be useful reference marks in their new position. The record of the movement will also be useful for various survey purposes.

Failure to properly resolve the disturbed marks issue potentially leads to CSDs connecting to incorrect nodes, causing difficulties validating and integrating surveys. Potentially, inappropriate assessment of whether a mark is disturbed could lead to boundaries being defined incorrectly.

1.3. How We Identify Disturbed Geodetic Control Marks

The Rules state that "disturbed":

in relation to an old survey mark, means that the mark is in a position different from that originally placed and does not include a change of position due to deep-seated movement

Many of the geodetic control marks in Canterbury are used to provide control for cadastral surveys. In this context, it is the consistency of a mark with other nearby marks that is most important – which is assessed using local (relative) accuracy. This means we are interested in how a post-earthquake observed vector between two nearby marks compares with the vector calculated by applying the deformation model to the pre-earthquake vector. Where the observed and modelled vectors agree within tolerance, the mark would be assessed as reliable. Where they disagree, it would be assessed as disturbed.

To determine the appropriate tolerance between marks, LINZ has taken a pragmatic approach, which is to use the tolerance in the Rules (0.03m + 0.015m/100m). The rationale is that a cadastral surveyor would most likely use this tolerance if making an assessment of mark disturbance. If a vector difference exceeds this level, the surveyor would not be able to reconcile their new data with underlying data, without creating a new node for the disturbed position.

In addition to the tolerance in the Rules, a maximum permitted tolerance is useful to ensure that marks are not inappropriately judged reliable solely because the distance to the nearest marks is such that the local accuracy tolerance is large. For example, two marks separated by 500m would have a maximum tolerance of 0.105m. However, while such a discrepancy is technically acceptable in terms of the Rules, cadastral data integration is very challenging at these levels. Therefore, a cap of 0.05m has been used. This is broadly consistent with the absolute uncertainty of the deformation model, and is the value used in LINZ geodetic contract specifications to assess mark disturbance.

It is necessary to compare a mark against several near neighbours, to avoid undue influence of any individual neighbouring mark (which may itself be disturbed). The approach used is to compare each geodetic mark with its nearest five undisturbed neighbours.In some cases a mark will have been disturbed by more than one of the Canterbury earthquakes. For example, a mark might have been disturbed by the February 2011 and June 2011 earthquakes, but not the September 2010 or December 2011 earthquakes. In this case there would be three nodes associated with this mark: the original node (now in its post-earthquake modelled position after the December 2013 update), a node representing the intermediate February 2011 position and a node representing the final position.

Disturbed positions, including intermediate positions, are all re-measured positions resulting from post earthquake geodetic control surveys. They are not positions created by adjustments only.

Intermediate disturbed positions resulting from individual earthquakes are added to Landonline as new nodes. The tolerance to be used to assess disturbance of intermediate earthquake positions is 0.03m, the constant component of the accuracy standard in the Rules.

1.4. How the Naming Convention Works

LINZ has developed a naming convention which complies with the Standard for the New Zealand survey control system, the standard which governs naming of geodetic control marks. The naming convention will append the word "POSN" and the date of disturbance in the format YYYYMMDD to the existing mark name. Examples of this convention are shown in the table below:

IT VII SO 11487Original position (or position updated using deformation model)
IT VII SO 11487 POSN 201009044 Sep 2010 disturbed position
IT VII SO 11487 POSN 2011022222 Feb 2011 disturbed position
IT VII SO 11487 POSN 2011061313 June 2011 disturbed position
IT VII SO 11487 POSN 2011122323 Dec 2011 disturbed position
IT VII SO 11487 POSN ABT2011Disturbed position, exact date not known

Where new nodes that reflect disturbed positions have already been added to Landonline, for example by a CSD, these will not have used the naming convention above. To ensure consistency with the above naming convention, the mark will be renamed, with the name on the CSD recorded as an alternative name for that mark.

There should be no unproven geodetic control marks, where these marks have been surveyed for geodetic purposes after the last significant earthquake. They should either be "proven to have moved consistent with deep-seated motion", in which case the mark name and existing node is retained, or they are "proven disturbed", in which case a new node is created using the naming convention outlined. See Capturing Unproven Marks for further information about unproven marks.

1.5. Impact for Cadastral Surveyors

Cadastral surveyors will see a greater number of nodes in Landonline for the same physical mark, where the mark has been assessed as disturbed. It is important that adoptions are linked to the correct node, based on the date of the adopted observations. New measurements will need to be linked to the most recent disturbed position. The naming convention implemented for disturbed marks should ensure this is as simple is practicably possible. Additionally, previous positions will be downgraded to a lower order than the current position.

For example, an observation adopted to the above geodetic control mark BDVV from a survey dated 10 October 2011 shall be linked to node IT VII SO 11487 POSN 20110613. Similarly, an observation adopted from a survey dated 1 March 2006 shall be linked to node IT VII SO 11487. However, a new measurement shall be linked to node IT VII SO 11487 POSN 20111223.

When using software which automatically links to existing nodes using a distance tolerance, the automatic linking will need to be checked carefully. It is possible that the automated linking algorithm will link to the wrong mark where there are several in close proximity. The same care needs to be taken when using Landonline plan generation as this tool has the option to automatically link nodes.