The method of defining accuracy for traditional surveying methodologies (total station traversing and leveling), is through the specification of Order and Class, according to Horizontal/Vertical Accuracy Standards. The foundation for representing a survey product that meets these standards is by determining the relative accuracies of measurements between known and well established control points called misclosures. The Order and Class that a survey adheres to is not only defined by the relative accuracies achieved, but also by the equipment and field procedures applied to obtain those results. Generally, when performing a field survey, the results are only as good as the control and the procedures that are utilized or established.
For surveys performed utilizing Global Positioning Systems (GPS), there is not a directly measured closed network, and therefore relative accuracy cannot be represented in the same manner. In the late 1990s came the commercial prevalence of Aerial LiDAR applications which similar to GPS technology did not conform to the traditional method of defining surveying accuracy. There was a need to develop a method to represent accuracy of point data and digital mapping products. The National Standard for Spatial Data Accuracy (NSSDA) was implemented by the Federal Geographic Data Committee (FGDC) to define absolute accuracies not as a ratio, class or order, but as a unit of measure.
Following NSSDA procedures, a minimum of 20 evenly distributed points from the digital map product are compared to conventional survey points to compute the root-mean-square error (RMSE) at the 95% confidence interval. There are instances where client specific requirements provide guidance for applying this procedure FEMA requires validation in various land cover classifications. Utilizing rapid-static GPS, redundant real-time kinematic GPS (RTK), or digital leveling to establish survey control for Bakers Mobile LiDAR collections, NSSDA Vertical Accuracies below 0.1 can be consistently realized project-wide, and through the proper placement and configuration of survey control, vertical accuracies of 0.03 can be routinely achieved. Again, the resultant product accuracy is only as good as the control established to support the project.
The image shows a Mobile LiDAR target which is used to adjust individual "flight" lines to a known control point with Northing, Easting & Elevation.
Similarly tests for calculating NSSDA Horizontal Accuracies have been accomplished by reviewing hard targets collected with both traditional survey methods and Mobile LiDAR. The tests concluded there was virtually no difference in the horizontal position of the fixed features (usually man-made), between the two capture methods. Typically the ability to precisely determine the horizontal distance between control and LiDAR is predicated on ones ability to establish where the rod-person positioned the survey pole to measure a given feature. Whereas, the vertical component is determined by measuring the elevation difference of two points at a common coordinate.
When reviewing targets as large as a utility pole or manhole, it is my professional opinion that the redundancy of measurements generated on any one feature by Mobile LiDAR, as well as the high relative accuracy of the features (due primarily to density of data), trumps the single side-shot measurements from a total station. Leveraging the numerous data points provided by a Mobile LiDAR capture for any one feature additionally facilitates more robust quality assurance regimens that further validate, either statistically or by simple manual review, accuracy results.
After adjustment utilizing the Tie Lines, Ground Lines can be added to adjust strips which may not have been adjusted – primarily due to lack of visible control
Independent Evaluation
After adjustment utilizing the Tie Lines, Ground Lines can be added to adjust strips which may not have been adjusted – primarily due to lack of visible control
Independent Evaluation
On multiple occasions, our prime contract holder or client have performed independent evaluations on the Mobile LiDAR collections to validate the accuracy. Each of the evaluations presented similar results, in fact almost statistically identical within groupings, that were sectioned by the methodology and type of control utilized to constrain the LiDAR point-cloud.
Below we present the results of a collection performed on a four lane divided US highway. The horizontal control was established using GPS RTK, while the vertical control was established via digital leveling. This particular client performed an accuracy evaluation of the LiDAR data utilizing 110 check points throughout the project area; each surveyed with GPS and digital leveling similar to the method for establishing control.
Using the first 20 measurements in the collection, NSSDA accuracies were computed as shown below (values shown in feet):
Through comparison of all 110 check points the following statistics were obtained:
Average z: -0.015
Minimum z: -0.086
Maximum z: +0.093
Average magnitude: 0.028
Root Mean Square (RMS): 0.035
Standard Deviation: 0.032
Standard of Care
A primary factor for achieving survey-grade results is the standard of care followed by the collection crew(s). Simply starting the system and haphazardly performing a collection without consideration for proper mission planning, solution status or application of lessons learned puts you in a precarious position even before collected data are cycled to the processing team. By establishing and implementing best practices for collection of Mobile LiDAR data and creation of survey control, you create the foundation for acquiring reliable, repeatable, and defensible information to ensure project success.