Tuesday, April 20, 2010

Cadastral/Parcel Mapping Quality Components

We are often asked “how good is that parcel mapping?” or “how accurate does my parcel mapping need to be?” Unlike data sets that are built from a single source, such as aerial imagery or topographic maps or lidar, the answer to the questions of how good or how accurate for parcel mapping is a case by case, parcel by parcel response. Certainly the results of an analysis could be rolled up to a statement about the overall data set but the quality of parcel data is much more than just spatial accuracy. What are the quality components for parcel or cadastral mapping?

Currency – How current is the parcel map?

The currency can be stated as simply as this parcel map includes all land transactions as of last Friday, or within one month or as of yesterday. The currency statement for the mapping might be different than the currency of the associated attributes or attributes that the map may be tied to. For example the assessment value information might be from last year. A parcel split might be current as of yesterday but the assessment values for the new parcels might be from the prior year until the year-end assessment information is brought current. The mapping currency should be stated separately from the related attributes.

Completeness – How complete is the parcel map?

At first glance most organizations might reply 100% complete, we map everything. But consider if the parcel map includes federally managed lands, are these mapped and indicated as federally managed land and does it reconcile with the federal agencies mapping of the same parcel? Are state managed lands included? Are right of ways included? Is this parcel mapping taxable lands? If it is taxable lands are all taxable lands on the roll represented on the map? Typically an ongoing parcel program will have 100% coverage for the parcels they are mapping, but it is always worth an extra look to be sure.

Lineage – Does every parcel polygon have a link to its source document?

An important part of the parcel mapping process is recognizing that it is a representation of the descriptions and content of the legal document from which it was generated. These documents can be survey plats, deeds, right of way plats or any number of types of conveyance documents. Providing a link to the source document that was used to define the geometry is an important and essential component of the parcel mapping. Note that this is the link to the document that defined the geometry and may not be the same as the document that contains information on the current owner.

Relative Positioning or Closure – Does every polygon have an indicator of the relative accuracy of the legal description?

For organizations that run coordinate geometry on their parcel it is possible to get a record of the linear error of closure (LEC) of the legal description, however not every organization uses coordinate geometry for parcel mapping and some legal descriptions are just not suited for this type of analysis. For example a parcel legal description may call out a river centerline or a physical feature such as a tree or road and the measurements to execute a linear error of closure analysis are not easily derived. To solve this it may be possible to develop a simple coding scheme that indicates the general quality of the associated legal description for the parcel polygon. For example the relative positioning codes could be something like the following.

1 – parcel closed, LEC at least 1 part in 1,000.

2 – parcel closed but the LEC was less than 1 part in 1,000 or one line of the parcel was a physical feature call that matched the imagery closely.

3 – parcel is drawn entirely from reference to imagery and physical references and these calls match the imagery.

4 – parcel metes and bounds description did not close and had to be best fit with surrounding parcel or imagery evidence.

5 – parcel closure or relative accuracy could not be ascertained or assigned.

6 – no attempt was made to determine the parcel relative accuracy.

These are just an example of some the things to consider in a relative accuracy coding. Another alternative might be to assign the LEC as the relative accuracy and indicate if no LEC was determined.

Absolute Positioning – How well odes the parcel legal description match control point location of the boundary?

This is the most traditional indication of “accuracy”. This is an indication of how well the parcel fits with other parcels and how well it ties to control measurements. For example if the corners of a subdivision have been located with global positioning system (GPS) how close is the corner location on the parcel map to that GPS location? Similarly if an organization has jurisdiction-wide orthoimagery, how close are the physical calls in the legal descriptions to the visual evidence on the imagery? F a parcel description runs to the middle of the road does the parcel mapping show the parcel line on the middle of the road, assuming the road on the imagery is the same road called for the legal description. Beware of the difference between the center of the right of way and the center of the road. These are often different. Even with the nuances of legal descriptions and visible evidence it is often possible to use the imagery to get a sense of the absolute positioning of the parcel maps.

Attributes Present – Are all of the attributes for the parcel map included?

This is a check of the parcel map database to assure that all of the parcel polygons have the attributes they are supposed to have. If the parcel has a tax identifier, is it entered into the parcel map database and is it correct? If the parcel map database uses annotation, are the annotation fields complete and correct? This evaluation would also apply to systems that use line types or any other codes for cartographic display or parcel classification.

The following two components are indications of the overall parcel mapping program in an organization. These components might be termed best practice as opposed to quality components, but these two components contribute to the overall quality of the parcel mapping.

Archived Data – Does the parcel mapping program include tracking historically mapped parcels?

Archiving data might be as simple as taking an annual snap shot of the parcel mapping database and archiving with the backup data. It could also be snap shots of the data periodically such as monthly or quarterly. In a transaction based parcel mapping system each parcel change will result in archiving the changed parcel to an historical parcel file or something similar so the parcel map can be used to “roll back” through time and visual observe changes. Parcel data that is archived in this manner typically keep the parcel number and linkages to source documents with the historical geometry.

Maintenance Process – Is there a documented and sustainable maintenance process in place?

Because parcel and cadastral data are so dynamic, potentially changing on a daily basis, it is important to have a defined and sustained maintenance process. A parcel map may be perfect at the moment it is delivered from conversion process, but it will soon be out of date if there is not a sustainable maintenance process. This means the updates and maintenance have to be consistently and on a regular basis. Even if that basis is once a year, it is a known maintenance cycle.

The reason organizations invest in parcel mapping is to use the resulting with related data such as the tax roll or deed recording or permitting or zoning or a plethora of other applications. The last three components of parcel mapping data quality categorize the data distribution or data publication. This is separate from the evaluation of the production data.

Publishing Schedule - Is the parcel map published on a regular schedule?

Do the customers for the parcel data know when to expect parcel mapping updates? Are updates provided or notification of updates made available to those who depend on the data. Notification might be that the most current information is pushed to a data service and when users add the data to their canvas the updates are reflected on their project without the need for additional notification.

Metadata – Does your published parcel data have essential metadata?

Metadata does not have to be onerous. If an organization focuses on establishing essential metadata for published data, then the metadata should be relatively stable and can be published with the data set as an XML file or as an html file or embedded in the distributed or published data set. With limited resources there is less demand for metadata for production data, data that is internal to the organization than there is for published data that will be likely be consumed by people who have little or no local knowledge. The FGDC Cadastral Subcommittee has developed a limited set of essential metadata and tools for creating the published version of this data, that can be found at this link http://www.nationalcad.org/showdocs.asp?docid=1153&navsrc=Standards&navsrc2=

Attributes Standardized – Are the attributes in the published data translated to standard naming conventions and content?

The FGDC Cadastral Subcommittee has developed a publication standard for core parcel data. This is a minimum set of information that meets a large audience of data requests. Some organization have added to this core level information but the goal is to encourage all published parcel data to be in or have a version of the published data that matches this standard. Added attributes are a bonus. The publication guideline can be found at this link.


So how good is your parcel map? Applying some or all of these quality measures and then developing an evaluation rating might be one way to answer that question. It should also be a way to measure the increasing quality of the parcel map as the maintenance progresses.


  1. Do you know of any org that has implemented anything like this? The jurisdiction I work with has asked for Parcel Map metrics and I developed a scale similar to what you mention. We are planning to move forward with it over the next couple of months.

  2. Stephen - I know counties that have incorporated components of this but i do not know any one who has holistically looked at this as a metric. I think it is great you are undertaking this, please keep us posted on your progress and let us know if we can be of any help. I am guessing if you are looking for a metric you'll be looking at the production or data set quality measures and then looking for changes in the overall status of the data set. sounds like you have a great start. - nancy

  3. I'd thought I'd send you an update on what we are doing.

    We are moving ahead with an external parcel map comparison test as the first step in developing a set of Land Records metrics. To develop the initial metric we decided to compare our parcel map against seven other Counties that are similar to ours in parcel density and general geography. The idea is to randomly select parcels from each of the County's and review them for aesthetic accuracy. Each parcel will be given a pass/fail rating and upon completion, the ratings will be used to create an accuracy rating for all eight Counties. We plan to use the seven other Counties as an aggregate measure against ours. We will then target the composite rating over the long term.

    To prepare for the test we have developed a viewer application that allows the reviewer to provide the pass/fail input. We have also gathered up the test parcel maps and divided them up into two groups. From each group of four Counties we will randomly select 70 parcels each. The reviewers will be students from a local university and I have asked our State Land Records Manager to act as a control reviewer in an attempt to add a little credence to the validity of the student’s judgment (and for impartiality).

    The plan is to perform the external review yearly and use it to test the progress of the internal QA/QC processes we are also developing. In theory, the internal improvements will eventually be reflected in the external measurement and will provide a reliable metric describing how accurate our parcel map is (or isn’t). - Stephen

  4. very helpful information for land surveyors in the cadastral sector..thanks for sharing this! http://landsurveyorsunited.com