Accuracy Terms & Definitions

To make network accuracy values easier to describe they are grouped into successive bands called Tiers.

The accuracy of a coordinate can be measured in different ways. To ensure that quality information about coordinates is provided in a way that is helpful to a range of users LINZ describes accuracy using two methods: network accuracy and local accuracy.

Network accuracy (Tiers)

Network accuracy is a value that represents the uncertainty of a coordinate in relation to a datum. This can be thought of as the standalone accuracy of a coordinate. The network accuracy answers the question “How accurate is this coordinate?”

Network accuracy is quantified as a circular error around a point for a horizontal position (latitude and longitude or northing and easting) and a linear error for heights. It provides a method of describing the accuracy of a coordinate by a single value.

The maximum horizontal error (HE) and maximum vertical error (VE) for a coordinate (at the 95% confidence level) is computed by the following equations (σ is the standard deviation of the coordinate component):

To make network accuracy values easier to describe they are grouped into successive bands called Tiers. The LINZ Tiers are defined in LINZS25006, and can also be found as a summary.

Local accuracy (Classes)

Local accuracy is a value that represents the uncertainty of a coordinate in relation to other nearby coordinates. This is the accuracy measure that is of most benefit to users of coordinates in a relative sense (eg surveyors wanting to know the accuracy of a line between two control marks). It is also synonymous with observation accuracy.

Local accuracy is grouped into successive bands called Classes. Each Class has a maximum local accuracy value that is defined by a constant term (c) and a distance-dependent term (p). The maximum permissible error between two coordinates (at the 95% confidence level) is given by the following equation (where D is the distance between the coordinates):

The distance (D) used to calculate the local accuracy between coordinates depends on the type of coordinate that is being assessed:

• horizontal coordinate accuracy is assessed with the horizontal distance
• vertical (height) accuracy of control marks and non-boundary survey marks is assessed with the slope distance
• vertical (height) accuracy of boundary marks is assessed with the vertical distance.

For a coordinate to be assigned to a Class it must pass the local accuracy test with all other coordinates of the same or higher class within a specified radius. In the surveying context this means that the relationship is tested between all marks within the radius, regardless of whether there are direct observations between them.

The LINZ Classes are defined in LINZS25006, and can also be found as a summary.

Orders

Orders are classifications used in Landonline to combine the Tier and Class accuracy requirements into a single category. This enables the accuracy of all LINZ coordinates to be described in a consistent way.

For a coordinate to be assigned to an Order it needs to meet both the applicable Tier and Class standards. Different Orders are used for geometric coordinates (latitude, longitude, ellipsoidal height) and normal-orthometric heights.

Note that the Orders used by LINZ changed in May 2010. Find out more about changing to the new Orders.

The LINZ Landonline Orders are defined in LINZS25006, and can also be found as a summary.

The Coordinate Accuracy fact sheet - LINZG25706 explains coordinate accuracy, and shows the network accuracy and the local accuracy relevant for each Order.