When your goal is to improve the spatial accuracy of parcel corner points, you can use tools such as the least-squares adjustment to compute updated coordinate locations. The least-squares adjustment uses COGO dimensions on parcel lines together with control points to estimate best-fit coordinates for parcel fabric points. When using the least-squares adjustment tool, the end goal is to improve ground-truth coordinates for parcel corners using the best available land records.
The best available land records are those that most closely represent what's on the ground and are typically based on field surveys. The dimensions for parcel lines recorded on the original land records documents are often the closest and best representation of the parcel boundaries on the ground, because they are derived from field survey measurements. Bear in mind that older field surveys may be less accurate.
When running a least-squares adjustment on parcels, COGO dimensions should match those recorded on authoritative source documents, and control points should match the correctly identified ground locations.
If parcel lines have no COGO dimensions, the adjustment ignores them. You can use the Update COGO tool to generate the dimensions, but generated dimensions should only be used if it is known that the parcel line shape geometry closely matches record dimensions. An example of this would be accurate CAD lines that represent recorded dimensions. CAD lines can be used to build parcels in the parcel fabric.
The entire parcel fabric does not need to have COGO dimensions that match record dimensions in order to run a least-squares adjustment. You can run a least-squares adjustment on a portion of the parcel fabric by selecting connected parcel lines that have valid record dimensions. For example, you can run a least-squares adjustment on the lines created after entering one or more parcel subdivisions.
Adjusting large datasets
When your goal is to adjust a large area or an entire dataset of parcel lines, it is recommended that you perform least-squares adjustments on smaller blocks of data first. Once it has been confirmed that those smaller blocks of parcel lines are adjustable and have no adjustment failures, those blocks of data can be combined into a larger, single adjustment of the dataset.
Running least-squares adjustments on smaller blocks ensures that they are adjustable with no errors. Adjustable blocks must have lines with record dimensions and valid control points.
Data quality checks
Several data quality checks should be performed on parcel lines before running a least-squares adjustment. Fixing data quality issues will ensure the least-squares adjustment succeeds and outputs meaningful results.
Before running a consistency check or weighted least-squares adjustment, it is recommended that you perform the following data quality checks:
- Add the Distance Mismatch quality layer to the map to identify lines with COGO attributes that are different from the line shape geometry. Mismatches can indicate erroneous line geometries or incorrect COGO distances. If a COGO distance is correct (matches recorded distance) but does not match the line geometry, the least-squares adjustment will adjust the line geometry to represent the COGO distance. If the COGO distance is not correct (does not match the recorded distance), the distance should be corrected before the line is input into the least-squares adjustment.
- Add the Too Short quality layer to identify lines that are too short to represent a meaningful distance, for example 0.05 feet. Very short lines are most likely invalid and do not reflect COGO dimensions from the parcel record. These short lines must be rectified and should not be input into a least-squares adjustment. Furthermore, lines that are shorter in length than the default accuracies could cause the adjustment to fail. For example, the line length is 0.003 meters but the default distance accuracy is 0.15 meters.
- Use the optional DIRECTION MUST MATCH WITHIN attribute rule to identify COGO directions that do not match their line shape directions. Mismatches can indicate erroneous line geometries or incorrect COGO directions. If a COGO direction is correct (matches recorded direction) but does not match the line geometry, the least-squares adjustment will adjust the line geometry to represent the COGO direction. If the COGO direction is not correct (does not match the recorded direction), the direction should be corrected before the line is input into the least-squares adjustment.
- Use the Highlight command to detect gaps and overlaps between parcels. Parcel lines must be topologically correct and not have gaps and overlaps.
- Validate the parcel fabric topology to ensure there are no topology errors for the Endpoint Must Be Covered By and Must Not Have Dangles rules. Since parcel lines must be topologically correct and must have parcel fabric points at their ends, the Endpoint Must Be Covered By topology rule must have no errors. Additionally, the Must Not Have Dangles rule confirms that parcel boundary lines are connected to each other and must have no errors.
When to run a consistency check
A consistency check can be run whenever COGO dimensions have been entered for parcel lines. A consistency check uses a least-squares adjustment to evaluate how well the COGO dimensions on lines fit with each other, or in other words, how consistent they are with each other. For example, a consistency check will evaluate if the COGO dimensions of different parcel lines calculate the same coordinate location of the corner point they connect to. If the dimensions of a line compute significantly different coordinates for that point, that line is inconsistent with other lines connecting to the point and may have incorrect COGO dimensions.
In the graphic below, the COGO dimensions on the line from point Sp2 are computing coordinates for point Sp5 that are inconsistent with the lines from the other points.
If COGO dimensions have been generated from the line shape geometry, running a consistency check on these dimensions will be meaningless. Generated COGO dimensions will match the fit of the line geometry and there will be no inconsistencies. For this reason, there is no need to run consistency checks on COGO dimensions generated from CAD lines, even if the CAD lines are highly accurate and reflect the dimensions of the parcel record. However if dimensions are edited, running a consistency check will be useful for detecting data entry mistakes.
When to run a weighted least-squares adjustment
A weighted least-squares adjustment is run when you want to improve the spatial accuracy of your parcel corner points. For example, you want your parcel boundaries to align better with your high-resolution imagery. A weighted least-squares adjustment uses the COGO dimensions on parcel lines together with weighted control points to derive updated, more accurate coordinates (x,y,z) for parcel points. To achieve improved spatial accuracy by running a least-squares adjustment, COGO dimensions on parcel lines must match recorded dimensions and control points must be valid control points (match correctly identified ground locations).
A weighted least-squares adjustment can be run on lines with unreliable or incorrect COGO but the results will be meaningless. Similarly, a weighted least-squares adjustment can be run with artificially created or invalid control points, but the results will be meaningless.
Before running a weighted least-squares adjustment, ensure that you have run a consistency check to identify erroneous COGO dimensions and have also performed the data quality checks described above.
Control points
A point can be considered a control point if it has the following characteristics:
- The point has a known x,y,z coordinates and matches a correctly identified ground location.
- The point has a higher accuracy than regular parcel corners. The higher accuracy point will act as a weight and will constrain the adjustment.
- The point is a parcel corner point or can be connected to a parcel point using a connection line.
Control points are often referenced on survey plats or plans and in some cases the coordinates are also provided. Control points can come from a variety of sources, however the most important thing to consider when obtaining control points for a least-squares adjustment is if the control points can be tied to parcel corners. This is why its ideal to, if possible, use control points referenced on a survey plat as they will either be located on parcel corners or be tied to parcel corners. Control points can be fully constrained, or weighted. The higher the accuracy of the control point, the more influence it will have on the outcome of the adjustment.
In some areas, locating control points to use in a least-squares adjustment can be difficult. If you have high-resolution imagery, points can be derived from the imagery and used as control points. However you must ensure that the control point can be tied to parcel corners (using a connection line with a direction and a distance).
When using control points derived from imagery, they will most likely will be less accurate than other control points. Set the point as a weighted control point and assign the point lower accuracy to distinguish it from more accurate control points.
How to use a priori accuracies
The a priori accuracy is an estimate of the measurement error in a dimension. Dimensions are derived from survey measurements taken out in the field and have an inherent measurement error associated with them. In general, the more recent the survey, the less the measurement error and the higher the accuracy is of the measurement. This is because the instrumentation used to take the measurements has become more technologically advanced over time.
There is a difference between measurement error and measurement mistakes. Dimensions based on measurements that are mistakes should be excluded from the least-squares adjustment and should not be allowed to affect the outcome of the adjustment. Dimensions based on field survey measurements (that have measurement error) can be assigned with estimated a priori accuracies to reflect the measurement error.
The estimated a priori accuracy will indicate how much weight the dimension will have in the adjustment. The lower the numerical value, the higher the a priori accuracy and the higher the weight of the dimension. In general, dimensions from more recent records should have higher a priori accuracies (lower numerical values) than older records. Thus, dimensions from more recent records will have more influence on the outcome of the adjustment than dimensions from older records.
A priori accuracies are an estimate, and they can be overestimated or underestimated. Overestimated a priori accuracies mean that the accuracies were estimated to be too high and need to be lowered. In other words, the numerical accuracy estimates need to be increased. Overestimated accuracies may occur if there are blunders or mistakes in the COGO dimensions of the lines being adjusted.
Underestimated accuracies mean that the accuracies were estimated to be too low and need to be raised. In other words, the dimensions are found to be statistically more accurate than the a priori estimates and the numerical accuracy estimates need to be decreased. Underestimated accuracies can occur when the lines being adjusted have COGO dimensions that were generated from the line geometry. In this case, the dimensions will all fit together almost exactly in the adjustment, indicating that the dimensions are highly accurate.
When a priori accuracies are overestimated or underestimated, a Standard Error(sigma) warning message is displayed under Messages on the Analyze By Parcel Least Squares Adjustment geoprocessing tool.