Tutorial: Running a fabric least-squares adjustment

This topic applies to ArcEditor and ArcInfo only.

Complexity: Advanced Data Requirement: ArcGIS Tutorial Data Setup Data Path: \ArcGIS\ArcTutor\Parcel Editing Goal: Successfully complete a fabric least-squares adjustment

This tutorial describes a workflow for a parcel fabric least-squares adjustment. The fabric adjustment uses the stored bearings and distances on the parcel lines, together with the control point coordinates, to obtain a best-fit solution of all the spatial information. In this tutorial, the control points have already been entered and linked to their corresponding fabric points. After the adjustment completes, an adjustment summary is presented. The adjustment summary is used to better understand attribute information on the lines and to identify any problems with the data that may need to be corrected. Once the problems have been fixed, the fabric point data is adjusted. The adjustment is run repeatedly until the maximum displacement in the northings and eastings are close to zero, meaning that the adjustment has converged onto an optimal solution.

Steps:

1. Open ArcMap and load the Parcel Editing map document (ParcelEditing.mxd) from your \ArcTutor\Parcel Editing\ folder.
2. Zoom to the fabric adjustment area shown below or click the Bookmarks menu and click Fabric Adjustment Area.

Fabric adjustment area

3. Click the Parcel Editor arrow and click Start Editing.
4. Close the Create Features window. You will not work with this window in this tutorial.
Before running a fabric adjustment, you will do a check fit of the control network. The check fit process checks how well the parcel network fits with the control network.

Note:

It is always good practice to run a check fit on your control points before running a fabric least-squares adjustment. The check fit residuals indicate how well the parcel network will fit with its control network. Parcel lines are going to have to adjust at least by the amount of the check fit residuals in the least-squares adjustment. Therefore, check fit residuals can be used as a guideline for setting adjustment tolerances on the Adjust Coordinates dialog box. If check fit residuals are large and inconsistent with each other, it makes no sense to run a least-squares adjustment. Parcel lines are going to have to adjust at least by that amount, and the result will be an unstable adjustment. The reasons for the large and inconsistent residuals should be identified first, and the problems should be rectified. In most cases, very large check fit residuals are caused by control points that are linked to the wrong fabric point.

5. Click the Maintain Control Points tool on the Parcel Editor toolbar.
All visible control points in the map extent are listed on the Control dialog box.
6. Click the Check Fit button.
Check fit residuals are displayed next to each control point.

Check fit residuals

Since all residuals are 0, the parcel network fits very well with the control network. Check fit residuals don't need to be 0 to run a fabric adjustment but should be within an acceptable range. In this case, residuals of 2 feet and above would be investigated before a fabric adjustment is run.
7. Close the Control dialog box.
8. Click the Select Parcel Features tool on the Parcel Editor toolbar and drag a box to select all the parcels in the fabric adjustment area.

Select adjustment parcels

Tip:

You don't have to select the control points being used in the adjustment. Control points are automatically detected on selected parcels because they are linked to fabric points.

9. Click the Parcel Editor arrow and click Adjust.

When first doing an adjustment on a new set of parcel data, the initial step is to use the adjustment summary displayed in the Least Squares Adjustment Summary dialog box to understand how good the COGO attributes are and to understand if there are any potential problems in the data that need to be fixed. The adjustment process will compute coordinates that are the best-fit position for all the fabric points; this computation uses all the bearing and distance values on the lines.

Note:

The adjustment computation does not alter the original COGO attributes on lines.

The adjustment summary compares the new positions of the parcel lines with the original observed or recorded COGO values and reports the differences between the computed and observed values. These numbers are called residuals, or computed minus observed (c-o) values. Any residuals that are large compared with the check tolerance you specify are outliers, meaning that they fall outside your expected range. This means that the check tolerance you specify defines your tolerance for outliers in the data. If you choose a higher check tolerance, it means that you are willing to accept larger differences between the adjusted end points of lines and the associated COGO values. The report will only show (c-o) values that are greater than the distance or bearing check tolerance that you specify. Until you get to know the data, you will not know what to use as check tolerances. The approach described in this tutorial shows how to interpret results in the adjustment summary, allowing you to get to know your parcel data and choose appropriate check tolerances for the parcels that you are adjusting.

Learn more about check tolerances on the Adjust Coordinates dialog box

The distance check tolerance is the best tolerance to use when getting to know the parcel record data.
10. On the Adjust Coordinates dialog box, change the Check Tolerance for Distances. Type a value of 0.100 US Survey Feet.
In this example, you start with the assumption that the data is very good, and you expect the adjustment to succeed with all residuals less than 0.3 feet (3 times the distance check tolerance).
Note:

A fabric adjustment fails if any residual is greater than three times the bearing or distance check tolerance.

11. Accept all other defaults on the Adjust Coordinates dialog box and click Run.

Adust Coordinates dialog box

The Least Squares Adjustment Summary dialog box appears and reports that the adjustment fails.
12. Scroll down the report to find the heading Parcel Lines Report - Difference between Computed and Observed/Recorded (c-o).
The screen shot below displays part of the parcel lines report.

Parcel lines report

Outliers are indicated in the report with a ## symbol. Note that there are five lines in the report with this symbol, and also that the last line of the report explains that the problem with these lines are that the difference between their recorded distance values and the new computed position exceed three times the tolerance that you specified. In other words, the differences between the computed distances and the recorded distances for these five lines are all greater than 0.30 feet.

Note:

The residuals are determined by a preliminary adjustment; none of the points have had their positions changed in the database at this stage.

These five outliers in the lines occur in two parcels, called LS 269 / LS 269 and ROS 6692/ ROS 6692. (These parcels have been stored to represent the plan boundaries, and so the plan name and parcel name are the same. The convention for naming in the report is [parcel name]/[plan name].)

The largest outlier of these five lines is 0.622 feet. This means that for the adjustment to succeed, the Distances check tolerance should be at least one-third of this value—in other words, after rounding to two decimals, at least 0.21.

One simple approach to have the adjustment succeed is to increase the check tolerance. However, if the check tolerance required for the adjustment to be successful is very large (10 feet or more), then this is a strong indication of a mistake in the data. In these cases, it is important to check that the values recorded for these reported problem lines match the original record values, as found on the legal description for the parcel.

Notice that there are also two points with the object ID’s 876 and 881 that are very close to each other. These two points are listed under the Close Points Report. Close points are fabric points that are closer than the specified Close Points tolerance and that do not have a line between them.

Close points report

Close points indicate that it is very likely the points should be the same point, that there is a topological inconsistency, and that these points should be merged. It is important to merge close points, because by doing so you are increasing the redundancy in the network; this is good for helping the adjustment succeed. For example, if there are two points close together that each have two lines defining their positions, you can merge the points and increase the redundancy by having four lines defining the position of just one point.

The least-squares adjustment summary has provided you with information about the actions needed to make this adjustment successful: first, to increase the check tolerance, and second, to merge some close points. It is important to note that increasing the check tolerance to 0.21 feet will make this adjustment succeed for this particular set of parcel data. However, after the points are merged, the data will have been changed, and different adjustment summary results will be expected. A good rule of thumb, if the previous adjustment was not successful, is to use three times the previously used value. In this case, use 0.3 feet. First, however, the close points need to be merged.

The fabric is not editable when the Least Squares Adjustment Summary dialog box is open; however, the information in the adjustment summary is also available in the results file.

13. Click the Results File button to access the results file of the adjustment summary.
The default text editor on your system opens the complete adjustment report for this adjustment.

Adjustment summary results file

14. Click Cancel to close the Least Squares Adjustment Summary dialog box.
15. Click Cancel to close the Adjust Coordinates dialog box.
16. Using the Select Parcel Features tool and drag an empty box on the map display to clear the selection of the parcels that were adjusted.
You will now merge the close points. To locate the points, you will use the point reporting tool located on the Parcel Editor toolbar.
17. Click the Point Reporting on 1,2 or 3 Points tool located on the Parcel Editor toolbar.
18. If necessary, move the Point Report dialog box such that all the adjustment parcels are visible.
19. On the Point Report dialog box, click the one point button and type 876 in the Point ID text box.
20. Click the Report button.
Coordinate information, point scale, and convergence are displayed for the point.

Point reporting on one point

Learn more about the point reporting tool
21. Click the Map Location button.
The map is panned to center on the point that is reported, and the point location flashes briefly in the map.
22. In the Table Of Contents window, right-click the points sublayer and click Label Features to turn on point ID labels for the points.
23. With the Point Report dialog box open, click the Zoom In tool located on the Tools toolbar and drag a box to zoom to the location of point ID 876.
You will need to zoom in several times to locate the point. As you zoom in to the point, you may need to click the Map Location button several times to get an idea of where the point is in the map.

Close points displayed in the map

24. Close the Point Report dialog box.
25. Click the Merges Unconnected Points Inside a Given Rectangle tool located on the Parcel Editor toolbar.
26. Drag a box around the close points.

Merge points

27. Click OK on the Mean Points dialog box to accept the default mean point tolerance.
The points are meaned (averaged) and merged into a single point.

Merged point

28. Close the adjustment results file.
29. Zoom back to the extent of the adjustment parcels. Click the Bookmarks menu and click Fabric Adjustment Area.
30. Open the job book. Click the Parcel Editor arrow and click Job Book.
Notice that there is a job with a "Mean Points" description. Each edit is tracked automatically in the job book, and the job book provides an audit of edits on the fabric.
31. Close the job book.
32. Click the Select Parcel Features tool on the Parcel Editor toolbar and drag a box to select all the parcels in the fabric adjustment area as you did in step 6.
33. Click Parcel Editor > Adjust.
Since the adjustment data has been edited by the merging of two points, you should expect the behavior and results for the adjustment summary to be slightly different. It was previously determined that a distance tolerance of 0.21 feet should be used for the adjustment to succeed. To be assured that the adjustment succeeds, you will use three times the previously used number, and use 0.3 as the tolerance.
34. Type 0.3 for the Distances check tolerance on the Adjust Coordinates dialog box.

Change the Distances check tolerance.

35. Click Run.
The adjustment completes successfully. Notice that there is now information displayed about the maximum shift of the adjustment.

Maximum shift

Maximum shift, which is the largest coordinate shift in the adjustment, is displayed above the Accept button for quick access. The same maximum shift information is displayed in the adjustment summary with the point ID. In this adjustment, the point with an ID of 884 moved more than any other point in the adjustment and had a shift in easting of -0.55 feet and a shift in northing of 0.139 feet.
In the fabric least-squares adjustment, the goal is rerun or iterate the adjustment until the points are no longer changing and the maximum shift becomes 0.00 feet for both easting and northing. The fabric adjustment needs to be manually iterated—iteration is not automatic.
36. Click Accept on the Least Squares Adjustment Summary dialog box.
37. On the Adjust Coordinates dialog box, click Run to perform the second iteration of the adjustment.
Notice that the maximum shift is now at point ID 1018, and it has moved less than one hundredth of a foot.

Maximum shift

Since the shift value approached closer to zero after the second adjustment, the fabric adjustment is converging. If the shift values become higher after the first iteration, this indicates that the adjustment is diverging. A diverging solution happens when one of two things has occurred:
1. There is a large blunder in the dataset, and no solution is possible until the blunder is corrected.
2. The best solution for the given dataset has already been reached, and the solution converges only to diverge on the next iteration. This occurs when 0.00 easting and 0.00 northing convergence is too stringent for the quality and precision of the dataset.
If the data is good quality and there are no major blunders, a fabric least-squares adjustment will usually converge after three iterations.
38. Click Accept and run the adjustment for the third iteration.
The maximum shift of 0.001 easting and 0.000 northing shows that the adjustment has converged.
39. Click Accept for the third time on the Least Squares Adjustment Summary dialog box and click OK.
40. Click OK on the Adjust Coordinates dialog box to apply and complete the adjustment.
41. Click Parcel Editor > Save Edits to save the edit session and complete the tutorial.

Tip:

Once a fabric adjustment has converged, you can run the adjustment one more time to perform the postprocessing options under Plan Structure Constraints on the Adjust Coordinates dialog box. An adjustment should be converged first; plan structure constraints are performed afterwards.